31ST Bled eConference Digital Transformation: Meeting the Challenges June 17 – 20, 2018, Bled, Slovenia Conference Proceedings Editors: dr. Andreja Pucihar dr. Mirjana Kljajić Borštnar dr. Pascal Ravesteijn dr. Juergen Seitz dr. Roger Bons June 2018 Title: 31ST Bled eConference Digital Transformation: Meeting the Challenges Subtitle: June 17 – 20, 2018, Bled, Slovenia, Conference Proceedings Editors: Andreja Pucihar, (University of Maribor, Faculty of Organizational Sciences, Slovenia), Mirjana Kljajić Borštnar, (University of Maribor, Faculty of Organizational Sciences, Slovenia), Pascal Ravesteijn (HU University of Applied Sciences Utrecht, The Netherlands), Juergen Seitz (Baden-Wuerttemberg Cooperative State University, Germany) Roger Bons (FOM Hochschule, Germany) Review: Albrecht Fritzsche, (Friedrich-Alexander University Erlangen-Nuremburg, Germany), Alenka Baggia, (University of Maribor, Slovenia), Ana Meštrović, (University of Rijeka, Croatia), Anand Sheombar, (HU University of Applied Sciences Utrecht, Netherlands), Andrej Škraba, (University of Maribor, Slovenia), Andreja Pucihar, (University of Maribor, Slovenia), Anna Sell, (Åbo Akademi University, Finland), Arjen Maris, (HU University of Applied Sciences Utrecht, Netherlands), Arthur Tatnall, (Victoria University, Australia), Benny De Waal, (HU University of Applied Sciences Utrecht, Netherlands), Casandra Grundstrom, (University of Oulu, Finland), Catherine Han Lin, (Deakin University, Australia), Christer Carlsson (Åbo Akademi University, Finland), Christian Bremser, (University of Applied Sciences Mainz, Germany), Christian Kittl, (evolaris next level Research Centre, Austria), David Murungi, (Bentley University, United States), David Phillips, (Deakin University, Australia), Davorin Kofjač, (University of Maribor, Slovenia), Doug Vogel, (Harbin Institute of Technology, China), Edvard Tijan, (University of Rijeka, Croatia), Eeva Kettunen. (University of Jyvaskyla, Finland), Florian Maurer, (Vorarlberg University of Applied Sciences, Austria), Francesco Pisani, (Frankfurt School of Finance & Management, Germany), Friedrich Holotiuk, (Frankfurt School of Finance & Management, Germany), Gregor Lenart, (University of Maribor, Slovenia), Guido Ongena, (HU University of Applied Sciences Utrecht, Netherlands), Hans Allmér, (Linnaeus University, Sweden), Hans-Dieter Zimmermann, (FHS St. Gallen, Switzerland), Harald Vranken, (Radboud University, Netherlands), Harry Bouwman, (Delft University of Technology , Netherlands), Helen Cripps, (Edith cowan University, Australia), Indrit Troshani, (University of Adelaide, Australia), Jakobus Smit, (HU University of Applied Sciences Utrecht, Netherlands), Jeroen Van Grondelle, (HU University of Applied Sciences Utrecht, Netherlands), Johan Versendaal, (HU University of Applied Sciences Utrecht, Netherlands), Jonna Järveläinen, (University of Turku, Finland), Juergen Moormann, (Frankfurt School of Finance & Management, Germany), Juergen Seitz, (Baden-Wuerttemberg Cooperative State University, Germany), Julian Szymanski, (Gdansk University of Technology, Poland), Katri Kerem, (Estonian Business School, Estonia), Kirsten Liere-Netheler, (University of Osnabrück, Germany), Kristin Vogelsang, (University of Osnabrück, Germany), Lauri Frank, (University of Jyvaskyla, Finland), Leif Marcusson, (Linnaeus University, Sweden), Liam D. Turner, (Cardiff University, Wales), Luuk Simons, (Delft University, Netherlands), Marijn G. A. Plomp, (VU University Amsterdam, Netherlands), Marja Exalto, (HU University of Applied Sciences Utrecht, Netherlands), Marjeta Marolt, (University of Maribor, Slovenia), Mark de Reuver, (Delft University of Technology, Netherlands), Marko Urh, (University of Maribor, Slovenia), Marta Aranyossy, (Corvinus University of Budapest, Hungary), Matt Glowatz, (University College Dublin, Ireland), Matthijs Smakman, (HU University of Applied Sciences Utrecht, Netherlands), Michael Meyer, (Braunschweig University of Technology, Germany), Milan Frederik Klus, (University of Münster, Germany), Mirjana Kljajić Borštnar, (University of Maribor, Slovenia), Mladen Jardas, (University of Rijeka, Croatia), Nilmini Wickramasinghe, (Epworth HealthCare and Deakin University, Australia), Pascal Ravensteijn, (HU University of Applied Sciences Utrecht, Netherlands), Patrick Helmholz, (Braunschweig University of Technology, Germany), Robertas Damasevicius, (Kaunas University od Technology, Lithuania), Roger Bons, (FOM University of Applied Sciences for Economics and Management, Germany), Rok Piltaver, (Jožef Stefan Institute, Slovenia), Ronald Spanjers, (Roessingh, Centrum voor Revalidatie, Netherlands), Sanda Martinčić Ipšić, (University of Rijeka, Croatia), Siw Lundqvist, (Linnaeus University, Sweden), Slobodan Beliga, (University of Rijeka, Croatia), Stefan Stieglitz, (University of Duisburg-Essen, Germany), Tiina Kemppainen, (University of Jyväskylä, Finland), Todor Stefan Lohwasser, (University of Münster, Germany), Tuomas Kari, (University of Jyvaskyla, Finland) Committee: Andreja Pucihar, (University of Maribor, Slovenia), Mirjana Kljajić Borštnar, (University of Maribor, Slovenia), Iztok Podbregar, (University of Maribor, Slovenia), Christian Kittl, (evolaris next level Research Centre, Austria), Pascal Ravesteijn, (University of Applied Sciences Utrecht, Netherlands), Hans-Dieter Zimmermann, (FHS St. Gallen, Switzerland), Helen Cripps, (Edith Cowan University, Australia), Nilmini Wickramasinghe, (Epworth HealthCare and Deakin University, Australia), Juergen Seitz, (Baden-Wuerttemberg Cooperative State University, Germany), Christer Carlsson, (Åbo Akademi University, Finland), Pirkko Walden, (Åbo Akademi University, Finland), Christian Kittl, (evolaris next level Research Centre, Austria), Matt Glowatz, (University College Dublin, Ireland), Roger Clarke, (Xamax Consultancy Pty Ltd & The Australian National University, Australia), Roger Bons, (Professor, FOM Hochschule, Germany), Doug Vogel, (Harbin Institute of Technology, China), Johan Versendaal, (HU University of Applied Sciences Utrecht and Open University of the Netherlands, Netherlands), Marijn Plomp, (Vrije Universiteit Amsterdam, The Netherlands) IVx=Independent variables DVx= Dependent variables User’s decision making about whetherto adapt components Assessment feature IV3 improves DV Technical editors: current busine of the Idea-generation on how to change different components Templates with solution-based pattern IV2 improves User’s understanding of the components of the current DVbusiness model. Pre-filled business model template IV1 facilitates Borut Slabe (Unversity of Maribor, Faculty of Organizational Sciences) and Jan Perša (University of Maribor Press) IoT Business Model Tooling functions Post- test survey Performing scenarios with the developed prototype Performing scenarios with the use of CANVAS online tool Pre- test sur Treatm ent Group Control Introd uction Computer Room 2a 2b 2c 3a 3b 3C F 1a 2b 3c Na Nb Nc XA,B,C=Participant of group X F=Facilitator =Computer =Control artefact =Treatment artefact Cover design: Jan Perša (University of Maribor Press) Conference: 31ST Bled eConference: Digital Transformation: Meeting the Challenges Conference date: June 17 – 20, 2018 Co-publisher/Izdajateljica: University of Maribor, Faculty of Organizational Sciences Kidričeva cesta 55a, 4000 Kranj, Slovenia http://www.fov.um.si, dekanat@fov.uni-mb.si Publisher/Založnik University of Maribor Press Slomškov trg 15, 2000 Maribor, Slovenia http://press.um.si, zalozba@um.si Edition: 1ST Available at: http://press.um.si/index.php/ump/catalog/book/343 Published: Maribor, June 2018 © University of Maribor Press All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publisher. CIP - Kataložni zapis o publikaciji Univerzitetna knjižnica Maribor 659.23:004(082)(0.034.2) BLED eConference (31 ; 2018 ; Bled) Digital transformation [Elektronski vir] : meeting the challenges : conference proceedings / 31st Bled eConference, June 17 - 20, 2018, Bled, Slovenia ; editors Andreja Pucihar ... [et al.]. - 1st ed. - El. zbornik. - Maribor : University of Maribor Press, 2018 Način dostopa (URL): http://press.um.si/index.php/ump/catalog/book/343 ISBN 978-961-286-170-4 (pdf) doi: 10.18690/978-961-286-170-4 1. Gl. stv. nasl. 2. Pucihar, Andreja COBISS.SI-ID 94707457 ISBN 978-961-286-170-4 DOI: https://doi.org/10.18690/978-961-286-170-4 Price: Free copy For publisher: full. prof. Žan Jan Oplotnik, Ph.D., Vicerector (University of Maribor) Digital Transformation: Meeting the Challenges ANDREJA PUCIHAR, MIRJANA KLJAJIĆ BORŠTNAR, PASCAL RAVESTEIJN, JUERGEN SEITZ & ROGER BONS 1 CORRESPONDENCE ADDRESS: Andreja Pucihar, Ph.D. Program Committee Chair, 31st Bled eConference, Associate Professor, University of Maribor, Faculty of Organizational Sciences, Kidričeva cesta 55a, 4000 Kranj, Slovenia e-mail: andreja.pucihar@fov.uni-mb.si. Mirjana Kljajič Borštnar, Ph.D., Associate Professor, University of Maribor, Faculty of Organizational Sciences, Kidričeva cesta 55a, 4000 Kranj, Slovenia e-mail: mirjana.kljajic@fov.uni-mb.si. Pascal Ravesteijn, Professor, HU University of Applied Sciences Utrecht, Bolognalaan 101, 3584 CJ Utrecht, The Netherlands, e-mail: pascal.ravesteijn@hu.nl. Juergen Seitz, Professor, Baden-Wuerttemberg Cooperative State University, Marienstraße 20, 89518 Heidenheim an der Brenz, Germany, e-mail: seitz@dhbw-heidenheim.de. Roger Bons, Professor, FOM Hochschule, Herkulesstraße 32, 45127 Essen, Germany, e-mail: roger.bons@fom.de DOI https://doi.org/10.18690/978-961-286-170-4 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract Bled eConference, organized by University of Maribor, Faculty of Organizational Sciences, has been shaping electronic interactions since 1988. Bled eConference is the oldest, most traditional and well renowned conference in the field with more than 30 years of tradition. The theme of this year’s conference is dedicated to “Digital Transformation – Meeting the Challenges”. 1 Current State of Enterprise Architecture 2 Future State of Enterprise Architecture Dynamic organizationalcomplexity Dynamic technologicalcomplexity The evolution of digital technologies and solutions has significantly impacted the way in which business is conducted and have big implications on our lives. Nowadays digital economy calls for transformation of businesses, governments, education and societies as whole. It also calls for enabling policies and politics for cross border and global digital business. In this year’s conference, we address various aspects of digital transformation and provide directions and guidelines for organizations to meet and overcome these challenges on their way towards successful digital transformation. Themes covered in the papers of these proceedings are focused on: digital transformation; business model innovation; blockchain and social media; big data, data science, and decision support systems; e-health, digital wellness and wellbeing; new applications and organizational models; and novel approaches and cases in education in digital economy. Keywords: digital transformation • business model innovation • digital technology • innovation • digitalization • Table of Contents CONFERENCE PROCEEDINGS PAGE Integrating Maritime National Single Window with Port Community System – Case Study Croatia Edvard Tijan, Mladen Jardas, Saša Aksentijević & Ana Perić Hadžić 1 How to start with big data - a multiple case study Christian Bremser, Gunther Piller & Franz Rothlauf 13 Citizen adoption of e-government services – Evidence from Hungary Marta Aranyossy 27 Studying Complex IT Challenges? Discuss Real Cases Janis L. Gogan & David M. Murungi 47 Business Model Innovation and Firm Performance: The Role of Mediation and Moderation Factors Mohammad-Ali Latifi & Harry Bouwman 67 Towards Responsible Data Analytics: A Process Approach Roger Clarke & Kerry Taylor 85 Examining the PoC System Implementation and Adoption: A FVM Perspective Imran Muhammad & Nilmini Wickramasinghe 109 Influence of Digital Coaching on Physical Activity: Motivation and Behaviour of Physically Inactive Individuals Tuomas Kari & Petriina Rinne 127 Perceived Intrusiveness of Personalized Marketing Katri Kerem & Monika Ulla 147 Digital Wellness Services’ Servicescape for Young Elderly Hans Allmér 159 Developing an E-learning Course on Serious Games: A Multidisciplinary Approach Tanja Korhonen, Teija Ravelin & Raija Halonen 171 An Explorative Study Into Situational Artefact Construction in Business Rules Management Sam Leewis, Koen Smit & Martijn Zoet 187 Digital Transformation in Healthcare – the case of a Chinese Medicine Inquiring System Catherine Han Lin, David Phillips & Nilmini Wickramasinghe 205 eHealth WhatsApp Group for Social Support: Preliminary Results Luuk P.A. Simons, Wouter A.C. Van Den Heuvel & Catholijn M. Jonker 225 Critical success factors in the design of suitable oncology platforms Muhammad Nadeem Shuakat & Nilmini Wickramasinghe 239 Digital Wellness Services: Key to Better Quality of Life for Young Elderly Christer Carlsson & Pirkko Walden 249 Improving Smart City Design: A Conceptual Model for Governing Complex Smart City Ecosystems Mervi Hämäläinen & Pasi Tyrväinen 265 Transforming operating rooms: factors for successful implementations of new medical equipment Navin Sewberath Misser, Joris Jaspers, Bas van Zaane, Hein Gooszen & Johan Versendaal 279 Digital Transformation in Retail: Can Customer Value Services enhance the Experience? Michael Meyer, Patrick Helmholz & Susanne Robra-Bissantz 291 Motives to Form Alliances for Digital Innovation: The Case of Banks and Fintechs Friedrich Holotiuk, Milan Frederik Klus, Todor Stefan Lohwasser & Jürgen Moormann 301 Evaluation of Push Notifications for Social Media Applications Emma Fahlman, Thomas Mejtoft & Helen Cripps 317 Learning Success: A Comparative Analysis of a Digital Game-Based Approach and a Face-to-Face Approach Linda Eckardt & Susanne Robra-Bissantz 331 Blockchain-based Smart Contracts in Waste Management: A Silver Bullet? Guido Ongena, Koen Smit, Jarno Boksebeld, Gerben Adams, Yorin Roelofs & Pascal Ravesteijn 345 Boosting Innovation for the Development of Smart-Service Factories of the Future: The Cases of the Federal State of Vorarlberg and its Neighbouring Regions Florian Maurer & Albrecht Fritzsche 357 Business Customer eXperience Alignment Framework: Improving Customer Satisfaction Felix Gras, Pascal Ravesteijn, Marlies van Steenbergen & Roland Bijvank 375 Adaptation of enterprise architecture efforts to an agile environment Robin Duijs, Pascal Ravesteijn & Marlies van Steenbergen 389 Willingness to Use Smartphone Application Assistant to Support Migraine Treatment Hanna-Leena Huttunen & Raija Halonen 401 ICT supporting healthcare for elderly in China: A systematic mapping study Zihao Liu & Raija Halonen 415 A case for participatory practices in the digital transformation of insurance Michael Persson, Casandra Grundstrom & Karin Väyrynen 429 How a flexible collaboration infrastructure impacts healthcare information exchange Rogier van de Wetering & Johan Versendaal 441 Uncovering the digital “x” phenomena in the IS field: A text analysis approach Joaquin Rodriguez & Gabriele Piccoli 455 Can Sport and Wellness Technology be My Personal Trainer? – Teenagers and Digital Coaching Eeva Kettunen & Tuomas Kari 463 Designing business model tooling for business model exploration: An experimental design for evaluation Alexia Athanasopoulou & Mark de Reuver 477 Identifying the Sales Patterns of Online Stores with Time Series Clustering Markus Makkonen & Lauri Frank 491 The Role of Competition in the Adoption of Mobile Payment among Merchants Francesco Pisani & Jürgen Moormann 507 Co-creation Model to Design Wearables for Emotional Wellness of Elderly Muhammad Usman Warraich, Irum Rauf & Anna Sell 515 Determining the Required Capacity of Business Information Management: Towards a Concise Instrument Frank Van Outvorst, Mart Meijnen, Pascal Timens, Sjoerd Walenbergh, Benny De Waal 531 Business Model Innovation in European SMEs - Descriptive analysis of quantitative survey and case survey data Marikka Heikkilä & Harry Bouwman 543 Means to Survive Disruption: Business Model Innovation and Strategic Continuity Management? Jukka Heikkilä, Marikka Heikkilä, Marko Niemimaa & Jonna Järveläinen 561 SMEs Perspective on Business Model Innovation Marjeta Marolt, Gregor Lenart, Mirjana Kljajić Borštnar, Doroteja Vidmar & Andreja Pucihar 577 Comparative Study of Sharing Economy Business Models in Accommodation Sector Rimantas Gatautis†, Eglė Vaičiukynaitė & Elena Vitkauskaitė 593 DOCTORAL CONSORTIUM What do we Know about Personality Traits, Technology Use, Intention, Reputation and Brand Credibility? Cross Cultural Issues in the Aviation Industry’s User-Generated Content (UGC) Ukamaka Onyenucheya 607 A model for smart technology adaptation in the case of urban infrastructure development with the purpose of improving the quality of citizens’ well-being Vasja Roblek 619 Managing Complexity of Digital Transformation with Enterprise Architecture Fabian Gampfer 635 Energy security as an indicator of the efficiency of strategic energy management in the European Union Bojana Vasić 643 Obtaining and Evaluating Digital Competences of Slovenian General Gymnasium Students Srečko Zakrajšek 651 Technology Enabled Sustainable Business Model Innovation Doroteja Vidmar 661 Literature Review on Root Cause Analysis Jaka Toman 679 Designing an Adaptive and Sustainable ITG model for SMEs Aleš Levstek 697 Integrating Maritime National Single Window with Port Community System – Case Study Croatia EDVARD TIJAN, MLADEN JARDAS, SAŠA AKSENTIJEVIĆ & ANA PERIĆ HADŽIĆ 2 CORRESPONDENCE ADDRESS: Edvard Tijan, Ph.D., Associate Professor, University of Rijeka, Faculty of Maritime Studies, Studentska 2, Rijeka, Croatia, e-mail: etijan@pfri.hr. 2nd Author Mladen Jardas, MSc., assistant, University of Rijeka, Faculty of Maritime Studies, Studentska 2, Rijeka, Croatia ,e-mail: mjardas@pfri.hr. 3rd Author Saša Aksentijević, Ph.D., Assistant Professor, Aksentijević Forensics and Consulting, Gornji Sroki 125a, Viškovo, Croatia, e-mail: sasa.aksentijevic@gmail.com. 4th Author Ana Perić Hadžić, Associate Professor, University of Rijeka, Faculty of Maritime Studies, Studentska 2, Rijeka, Croatia, e-mail: ana@pfri.hr DOI https://doi.org/10.18690/978-961-286-170-4.1 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract A single point of data entry for documentary requirements and procedures in maritime transport - Maritime National Single Window (MNSW) includes process integration of all stakeholders in the entire seaport system. The primary goal of the MNSW is to eliminate data redundancy in a way that the entered data is instantly visible in other systems, according to the set level of authorization and authentication. In many seaports, the administrative MNSW is connected to the commercial Port Community System (PCS), an information system for the exchange of cargo related commercial data. The linking of the MNSW and the PCS connects the administrative with the commercial business aspect, making seaport business processes more efficient and more effective. Both interfaces can only be developed by using process reengineering and presume significant investments in underlying information technologies. Such implementation requires in-depth analysis of all stakeholders’ processes in the seaport system, in order for both systems to complement each other. Keywords: • Maritime National Single Window • Port Community System • business integration • reengineering • 1 Introduction According to the Directive 2010/65/EU of the European Parliament and of the Council (Directive 2010/65EU, 2010) each Member State should implement the Maritime National Single Window (MNSW) in order to optimize and facilitate the process of announcement and registration of ships which arrive to ports and/or depart from ports of the Member States. Single Window is defined as a facility that allows parties involved in trade and transport to lodge standardized information and documents using a single entry point to fulfil all import, export, and transit-related regulatory requirements (Moïsé, E. et al., 2011). MNSW is the place where according to the Directive, all information is entered only once, and becomes available to various competent authorities and the Member States. The aim of the Directive is to simplify and harmonize the administrative procedures applicable to maritime transport by electronic data exchange prescribed by standards and rationalization of formal reporting (European Commission, 2015). According to the new directive, which should be published mid-2018, a better cooperation is required between national authorities in the development of the MNSW and the implementation of the eManifest. The main objective of the eManifest is to demonstrate the way in which cargo-related information required by both maritime and customs authorities can be submitted together with other reporting formalities required by Directive 2010/65/EU in a harmonised manner, and using a European Maritime Single Windows environment (EMSA, 2015). The Republic of Croatia, as an EU member, is obligated to invest in the development of information technologies and thus in the development of the MNSW (EMSA, 2015). The development of MNSW should accelerate business processes, which also means increasing competitiveness in relation to neighboring ports. Croatia is currently using an integrated maritime information system (CIMIS) (Čičovački, 2013) for electronic delivery and exchange of data about ships, cargo and passengers in the official processing of announcement, arrival and departure of the ships, their stay in ports, as well as the related official procedures. The problems occur in subsequent official procedures due to multiple entries of the same data in several diverse applications where errors often occur during repeated data input. Data is still duplicated using paper and electronic means (inputting the same data in different systems). It is therefore necessary to develop a Port Community System (PCS) which represents a neutral platform for electronic data exchange between all stakeholders of the port process (Bezić et. al., 2011). By harmonizing MNSW and PCS, the ultimate goal would have been achieved, which is the elimination of data duplication in a way that once entered data is visible to all stakeholders of the seaport system, resulting in more efficient and more effective business processes. 2 Maritime National Single Window and Port Community System One of the most acute problems of seaport processes is the lack of automatic data exchange between existing systems. Processes are being slowed down due to the administrative load, which refers to the entry of data related to the announcement and registration of the arrival of ships into the port. Processes are being slowed down even further in the commercial part of the port activities (cargo-related data) because of the similar reasons. Stakeholders involved in seaport processes are still burdened with the manual inputting of data into paper documents, documents that contain almost the same information as already electronically submitted data. Therefore, data and/or document distribution to the competent authorities creates confusion caused by data redundancy. These administrative tasks significantly increase shipping costs by utilizing excessive waiting time for cargo processing, also causing delays in the next port of call. This archaic method of doing business reduces the competitiveness of maritime transport. Therefore, the EU has adopted a Directive 2010/65/EU about reporting formal reports, whose goal is to harmonize administrative procedures applicable to maritime transport. To achieve this, it is necessary to develop the MNSW that would be connected with the SafeSeaNet, eCustoms and other regional and supranational electronic data exchange systems. Safeseanet, a system of EU for the exchange of maritime information to promote efficient maritime transport (SafeSeaNet, 2016) provides almost real-time information for about 17 thousand ships that operate on a daily basis throughout the EU (EMSA, 2018). The data that are exchanged in SafeSeaNet are identity of the ship, position and condition of the ship, time of departure and arrival, incident reports, details of dangerous cargo, waste and ship safety. eCustoms is defined as an application of IT technologies in public administration. It is connected with the organizational changes and new abilities of public services, which aim to improve quality of provided services by the government (Granqvist et al., 2012.) MNSW are systems that combine data from all ports of a state and have B2G features (Single Window Directive for formalities in maritime traffic, 2015). Usually a smaller set of data is shared through these systems because this data is important on the national level. For the past several years, the EU has been working on the development of the European Maritime Single Window (EMSW) with the aim to fully harmonize interfaces available to operators of ships in order to provide required information all across the EU. The purpose of EMSW is to standardize information needed for port management so that the submitted data can be publicly available to all relevant stakeholders. (EMSA, 2018) To improve business processes, EMSW and MNSW should be synchronized and fully integrated with the PCS. The PCS combines and exchanges data among all the stakeholders involved in the operation of a particular port cluster (van Oosterhout et al., 2008). Many countries do not have a developed MNSW, but they have very diverse and well developed PCS systems instead. There is no global and standardized PCS model applicable to all ports as it may be the case with the MNSW. Each PCS is specific, each country has specific legal regulations and therefore each port community develops its own PCS according to its needs. Each PCS should communicate through the interface of MNSW so that the data would not be duplicated and in order to speed up business processes (Kapidani et. al., 2015). PCS helps stakeholders of the port processes to reduce logistics costs through faster information flow, deliver the cargo faster, enable the flow of goods, and finally, boost economic growth. As a secondary result it helps in reduction of externalities such as pollution and harmful emissions (IPCSA, 2018). Therefore, conclusion can be derived that there is no unique way to define MNSW. Each Member State will have its own requirements and conditions, for example, should the MNSW be associated with other systems of the authorities or will the authorities only collect information from the MNSW’s user interface; should the national maritime transport system (in Croatia - CIMIS) be part of the holistic MNSW solution or will it be a separate system; which legal regulations will be included in the development of the MNSW environment. 3 Current State of MNSW Development in the Republic of Croatia Beginning with 2013, an integrated maritime information system (CIMIS) has been in development in the Republic of Croatia for electronic delivery and data exchange about ships, cargo and passengers in the official proceedings of announcement and registration of arrival and departure, as well as related official procedures except in one segment - customs supervision of ship`s cargo. Using announcement of the ship’s arrival, the agent receives the documents that are entered into the CIMIS system (The regulation on forms, documents and data in maritime traffic, their collection and exchange, and the issuing of Free Pratique, 2017). In the relevant documents, some fields are redundant, but in the CIMIS system they are entered only once, and the entered data is automatically propagated into the fields of other documents, where applicable. Documents entered into CIMIS are Notification of arrival, Ballast Water Reporting Form, Notification of Ship-Generated Waste, ISPS Code Arrival Notification, Dangerous Goods Manifest, PortPlus Announcement - PSC MEI, IMO FAL forms (Tijan et al., 2017). By entering data related to the announcement of the ship’s arrival, the data becomes visible to all users of the process and they can use them for planning and implementation of their own activities. During the announcement of ship’s arrival, the data about the ship, the voyage, passengers, crew, cargo and documents that are submitted during the procedure of the ship’s arrival are entered into the system (Tijan et al., 2017). At the time when the foreseen operations are completed during stay of the ship in the port, be it commercial activities related to loading or unloading of cargo, the ship’s agent (or ship`s captain) submits required documents through the CIMIS system that are then used by all stakeholders involved in the administrative reception of the ship such as the Harbor Master Office, Port Authority, Ministry of Health, Ministry of Internal Affairs and Customs. Data provided through the CIMIS system are the crew and passengers list, stability calculations of the ship, the permit to complete all customs procedures, ETD, actual quantity of the cargo and exact number of passengers on board. The port authority defines location of the mooring or berthing through CIMIS and confirms the ETA. SHIP AGENTCIMISCUSTOMSFREIGHT AGENT- POLICE- PORT AUTHORITY- HARBOUR MASTER`S OFFICE Scheme 1: Current procedures in Croatian ports The problem occurs after the ship’s arrival in the port at the cargo registration through Cargo Declaration - IMO FAL 2 form, (IMO, 2018) which in the current system does not have the possibility of uploading so the data must be manually entered which requires quite a lot of time. Often, multiple errors occur when entering the data. At this point, it is important to distinguish the ship’s agent (usually one) from the freight agent (possibly multiple). Several freight agents can sometimes handle the cargo formalities for different cargo aboard the same ship. The freight agent takes the cargo list from CIMIS and must enter the cargo data in detail in the IMO FAL 2 form, which is after completion sent to the Customs administration. It should be emphasized that the Customs interface is functioning independently of the CIMIS system and that is the reason why it is necessary to constantly retype the data manually (Čičovački, 2018). The Customs administration communicates directly with the freight forwarder and the freight agent during cargo processing (Tijan et al., 2017). The same procedure is applicable when the ship leaves the port. The freight agent manually fills out the data (Export manifest) and only then the ship can leave the port. This problem occurs also because the whole procedure does not use electronic signature resulting in processes being performed both in electronic and in paper form (Port of Rijeka Authority, 2018). The procedure is even more complicated in the case when the ship is moored outside the customs gate (Tijan et. al., 2017). Then the ship agent needs to submit the request to the Ministry of Internal Affairs and to the Customs Administration that approves the request. The Harbor Master`s Office gives the consent to such request which is previously approved by the Ministry of Internal Affairs and the Customs administration. In terms of delivering real and accurate data of the ship, the ship’s agent announces the ETA via CIMIS (Čičovački 2018). This data can be used by the Harbor Master Office, pilot service, tug service, and the number of available tugs can be checked if they are necessary for the ship's entry process. The Harbor Master's Office then issues the formal permit which enables the start of port operations that are not directly under the jurisdiction of the Port Authority. This permit may only be issued after the following services have ascertained that there are no obstacles in their domain of jurisdiction: consent must be given by the Sanitary Inspection, the Border Phytosanitary Inspection, the Ministry of the Internal Affairs - Maritime Police and the Customs Administration. The port authority shall provide (in the CIMIS system) a certificate of the quantity of the ship’s really delivered (unloaded) waste to the Harbor Master`s Office for inspection so that the ships can receive the permit to navigate to the next port without the delivery of the waste and/or cargo residues (Tijan et. al., 2017). If a ship is travelling in an international voyage from the mooring or berthing that is outside of the customs gate, the agent must submit a request. Upon receipt of the request, the Ministry of Internal Affairs can make the necessary checks and issue a decision that is available to the Customs, or to the Harbor Master Office through the CIMIS. On the basis of the received data, documents related to the international voyages departing from the last Croatian port, the port authority issues the permission to the ship to leave the port when all conditions for departure are met in accordance with special regulations. 4 Development of MNSW and PCS with EMSW in Croatia At the beginning of 2017, the EU ministers agreed on a declaration that would harmonize the data between ports of the Member States (European Commission, 2018). The European Maritime Single Windows (EMSW) is a place where all information (except eManifest) is registered and accessible to all various competent authorities (EMSA, 2018). The data would be available to all providers of the ship-related data (agents, ship masters, ship owners), relevant public authorities associated with the seaport and other Member States of the SafeSeaNet. The data that is entered in the EMSW are port arrival and departure, ship particulars, cargo and dangerous goods, ship`s stores, crew and passengers, crew`s effects, security, waste and health data. The ship agent registers the described information in previously described CIMIS system, which in Croatia currently serves as the MNSW. In the development of the single interface, MNSW would represent an information bus established between the CIMIS and the EMSW. By linking these systems, administrative procedure of the ship’s arrival and departure would be significantly accelerated, which would also accelerate other administrative and commercial procedures between the stakeholders of the seaport system. Data redundancy would also be avoided. In a real life scenario, the ship agent inputs the ship identification data, crew list and other relevant data in the CIMIS system, which could be avoided because of established connection to the EMSW through the MNSW bus. For example, if the ship is leaving the port of Trieste for the port of Rijeka, all data related to the ship would be visible through the EMSW in the next phase(s) of the voyage. CIMISCUSTOMSFREIGHT AGENT- POLICE- PORT AUTHORITY- HARBOUR MASTER`S OFFICEPCSEMNSWMNSW (shared bus)SHIP AGENT Scheme 2: System orchestration in Croatian Maritime National Single Window Further problems that slow down the seaport processes are caused by an underdeveloped or non-existent PCS system (current exception is the port of Ploče, where PCS is partially introduced and still in development), thus making the customs procedures and the work of freight agents more difficult. By introducing the eManifest into the EMSW, the work of the freight agent and customs would be considerably simplified. The eManifest project started in February 2017 and will encompass the Arrival Manifest, using the existing Maritime Single Window prototype which was developed by the European Maritime Safety Agency (EMSA) in compliance with the Directive 2010/65/EU on reporting formalities (EMSA, 2016). The eManifest, or electronic cargo list, contains information about the status of transported goods and is considered to be a practical solution by which this could be achieved. The eManifest would assume the form of a harmonized electronic cargo list and as such would represent an instrument for further facilitation of the maritime transport which is done by ships that operate in seaports of the EU and third countries. The eManifest must be available in electronic form to the customs authorities at the next port of call where the cargo will be unloaded, whereby quick transit would be enabled to cargo from the EU (European Commission, 2018). By referring to the information about the cargo that is collected in the previous port in the eManifest, an additional factor of compliance with the fiscal and safety requirements of the EU would be provided. By introducing the electronic signature within the PCS system, administration procedures would be additionally validated and secured. 5 The Example of MNSW and PCS Integration The integration of MNSW and PCS would greatly contribute to the development of seaport processes, while at the same time the cost to the ship owners or shippers would be reduced. Therefore, it is of a common interest to speed up and facilitate the procedures required for the arrival, stay and departure of the ship, throughout the business processes. For example, when a ship departs from one port, the ship's port of entry would convey the EMSW data regarding the arrival of the ship to the next port. The data entered in the EMSW would be synchronized to national platforms or the NMSW. The next step is establishing the connection of the NMSW with the PCS. The data entered in NMSW (synchronized data from EMSW) would be visible in the PCS (an integration of the entire port community is required as shown in Scheme 3.). Scheme 3. Communication between individual seaport stakeholders (a) without the PCS and (b) with the PCS (Irrannezhad et. al., 2017) Using the data related to the type and size of the ship, the port authority could plan in advance or define the place of mooring or berthing, so all stakeholders of the port process could be optimally organized such as pilots, Harbor Master Office, customs, etc. (scheme 4). The described communication would greatly accelerate and facilitate the ship's agent work through the announcement and registration procedures of the ship, as well as other operational activities during the ship’s stay in the port. MNSWPCSEMSWPORT AUTHORITYPOLICEHARBOUR MASTER`S OFFICECUSTOMSAGENTS Scheme 4: Integration between MNSW and PCS 6 Conclusion According to the presented research, Croatia will have to develop the MNSW to comply with the present EU recommendations. For the development of the MNSW, it is necessary to analyze the existing systems that will participate in the exchange of data, their ability to communicate through exchange of the messages, receiving, storing and exchanging the documents and information. It is important to emphasize that there is no unified or universally applicable method of establishing the MNSW. It is necessary to establish a broader integration infrastructure of seaport stakeholders. In Croatia, the MNSW should be integrated with the CIMIS system, which proved to be successful and well accepted by stakeholders. Therefore, the MNSW should represent a “government service bus” that will be used for the exchange of data, ensure the compliance with business processes and conversion of the various data formats entered by stakeholders. This means that the MNSW system should not be used for the active entry of documents and data in the sense of a single data entry point. The relevant data exchange will be achieved through the existing or future systems (the new PCS systems of Croatian seaports). The construction of the PCS system is essential for the stakeholders of the port processes to whom business operations would considerably be facilitated through faster information flow and the reduction of errors that occur through multiple data entry. In order to achieve this, it is primarily necessary to invest in information and communication technologies that represent the basis for the described development. This should substantially accelerate business and port operations, both in the seaport itself, and among stakeholders. In addition to the acceleration of the processes, it would also simplify data entry procedures, resources would be used more efficiently, redundant data would be eliminated, cargo would be processed faster (which especially applies to the Customs part) and security, integrity and transparency of the process would be greatly increased. Acknowledgement This work has been financially supported by University of Rijeka under the Faculty of Maritime Studies projects. References Bezić H., Tijan E., Aksentijević S. (2011). Port Community Systems – Economic Feasibility Evaluation. Review of Contemporary Entrepreneurship, Business, and Economic Issues. J.J. Strossmayer University of Osijek, Faculty of Economics in Osijek. 2/2011 247-256 - 2011. Čičovački, L. (2018). Maritime National Single Window – today and tomorrow, MNSW and maritime traffic digitalization conference. Ministry of Sea, Transport and Infrastructure of Croatia, Maritime Safety Directorate Čičovački L. (2013). Electronic announcement of vessel departure in domestic voyage via Croatian Integrated Maritime Information System - CIMIS, Ministry of Sea, Transport and Infrastructure of Croatia, Maritime Safety Directorate Directive 2010/65 EU of the European Parliament and of the Council, Official Journal of the European Union, 2010 EMSA (2015). National Single Window Prototype - An electronic solution for simplifying administrative procedures EMSA (2016). eManifest Pilot Project, Version 0.8. EMSA (2018.) European Maritime Single Window (EMSW), http://emsa.europa.eu/infographics/item/2832-european-maritime-single-window-emsw.html, (05.03.2018.) EMSA (2018). European Maritime Single Window Environment, http://www.emsa.europa.eu/related-projects/emsw.html 05.03.2018.) EMSA (2018). Vessel traffic monitoring in EU waters, http://www.emsa.europa.eu/ssn-main.html (05.03.2018.) European Commission (2015). National Single Window Guidelines European Commission, Mobility and Transport, https://ec.europa.eu/transport/modes/maritime/digital-services/e-maritime_nl, 05.03.2018) Granqvist, M., Hintsa, J., Mannisto, T. (2012). Future Development of e-Customs: A Survey Study with Swiss Companies, International Journal of Electronic Government Research, Vol. 8, No. 4, pp. 49-59. IMO, FAL forms and Certificates, http://www.imo.org/en/OurWork/Facilitation/FormsCertificates/Pages/Default.aspx, 05.03.2018. International Port Community System Association (IPCSA), Port Community Systems, http://www.epcsa.eu/pcs, 05.03.2018. Irannezhad E., Hickman M., Prato C. Modelling the Efficiency of a Port Community System as an Agent – based Process (2017). The 6th International Workshop on Agent-based Mobility, Traffic and Transportation Models, Methodologies and Applications Kapidani N., Kočan E. (2015). Implementation of National Maritime Single Window in Montenegro. Telecommunications Forum Telfor (TELFOR), 2015 23. Moïsé, E., Orliac, T., Minor P. (2011), Trade Facilitation Indicators: The Impact on Trade Costs, OECD Trade Policy Working Papers, No. 118, OECD Publishing, http://dx.doi.org/10.1787/5kg6nk654hmr-en Port of Rijeka Authority (2018), Maritime National Single Window and maritime traffic digitalization conference The regulation on forms, documents and data in maritime traffic, their collection and exchange, and the issuing of Free Pratique (Croatian Official Gazette 70/2013, 55/2015 and 103/2017) SafeSeaNet, (2016) Interface and Functionalities Control Document, Version 1.1.2. Srour, van Oosterhout, van Baalen, Zuidwijk (2008). Port Community System Implementation: Lessons Learned from International Scan, Conference: Conference: Transportation Research Board 87th Annual Meeting, Volume: 08-2041 Tijan E., Zec D., Jugović A., Aksentijević S., Jardas M., (2017), The development of the Single Window for maritime traffic formalities (NSW). University of Rijeka, Faculty of Maritime Studies. Single Window Directive for formalities in maritime traffic (Croatian Official Gazette 119/2015). How to start with big data - a multiple case study CHRISTIAN BREMSER, GUNTHER PILLER & FRANZ ROTHLAUF 3 CORRESPONDENCE ADDRESS: Christian Bremser, University of Applied Sciences Mainz, Lucy-Hillebrand-Straße 2, 55128 Mainz, Germany, e-mail: christian.bremser@hs-mainz.de. Gunther Piller, University of Applied Sciences Mainz, Lucy-Hillebrand-Straße 2, 55128 Mainz, Germany, e-mail: gunther.piller@hs-mainz.de. Franz Rothlauf, Gutenberg University Mainz, Saarstraße 21, 55122 Mainz, e-mail: rothlauf@uni-mainz.de. DOI https://doi.org/10.18690/978-961-286-170-4.2 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract As part of an advancing digitalization, many enterprises feel the need to explore the possibilities big data may provide for their business. However, only a few companies use big data applications productively, despite its high expected potential. How companies examine the possibilities of big data, is therefore a highly interesting and relevant question. Based on a multiple case study we identify three different approaches: Companies either initially focus entirely on business aspects, or on a systematic build-up of a big data technology and data platform. Innovation adoption research is used as a theoretical basis. Keywords: • big data • digitalization • innovation adoption •¸ 1 Introduction The potential benefits and challenges associated with big data are an important topic for companies in all industries – in particular in the current era of an increasing digitalization. Big data promises new data-driven services to improve processes and enable innovative products and business models (Sivarajah et al., 2017). Against this backdrop, a growing number of companies are investing in big data, combined with the hopes of competitive advantages (Constantiou & Kallinikos, 2015). Nevertheless, companies seem to have difficulties with the productive implementation of big data applications. According to a Gartner study, only 14% of enterprises have put big data projects into production (Kart, 2015). Therefore research on the adoption of big data applications are important and of scientific and practical interest. The introduction of innovations is described by adoption theories. The process of innovation adoption typically involves two phases (Rogers, 2003): initiation and implementation. Within these phases, new technologies have to overcome several hurdles before being used productively. For technology-driven innovations, like big data (Nam, Kang, & Kim, 2015), the initiation phase, where companies search for valuable use cases for different big data technologies, poses a first serious obstacle. This initial step towards the exploration of big data potentials is the focus of our study. In particular we address the following research question: Which generic approaches can be identified when companies explore the potentials of big data in the initiation phase of innovation adoption? Despite the high relevance, there are no specific studies on the initiation phase of big data adoption. Current research mainly investigates general influencing factors and hurdles during the implementation of big data technologies. In contrast, this paper analyses current approaches for the exploration of new big data potentials in the initiation phase. For this purpose, a multiple case study with ten companies from different industries was conducted. The organizational adoption process of Rogers (2003) has been used as a theoretical starting point. This report is organized as follows: The current research on big data adoption is summarized in the next section. Section 3 presents our conceptual framework. Section 4 introduces the research design. Section 5 presents the findings from our cases. A discussion of the results in section 6 and a summary of the main points in section 7 complete this work. 2 Current research Big data is defined by the TechAmerica Foundation (2012) as "a term that describes large volumes of high velocity, complex and variable data that require advanced techniques and technologies to enable the capture, storage, distribution, management, and analysis of the information." Obviously, big data is a bundle of new technological and methodological possibilities that allow to process and analyse large, complex and rapidly growing data sets (e.g. stream analytics, in-memory data processing, NoSQL databases). Enterprises want to take advantage of these opportunities and promise a wide range of benefits through the introduction of big data applications (see, for example, (Brown, Chui, & Manyika, 2011; Davenport, Barth, & Bean, 2012; Kiron, Prentice, & Ferguson, 2014)). One approach to describe the introduction of technology-driven innovations is the adoption theory. On the one hand, this theory covers the identification of factors that influence the decision-making process of innovation adoption (Rogers, 2003). On the other hand, it describes the process which innovations have to go through, ranging from an initial awareness in companies to its productive use (Fichman, 2000). Previous work in the context of big data adoption mainly focuses on the investigation of general influencing factors through the Technology-Organization-Environment Framework (TOE) (see, for example, (Agrawal, 2015; Malaka & Brown, 2015; Sun et al., 2016)). The TOE describes the impact of technological, organizational and environmental aspects on organizational decision making with respect to technology innovations (Tornatzky, L. G., Fleischer, M. & Chakrabarti, 1990). As a result, it has been shown (Agrawal, 2015; Malaka & Brown, 2015; Sun et al., 2016) that the protection and integration of data are considered as important technological challenges. Organizational aspects, such as, unclear processes, lack of analytical skills and indistinct prioritization of use cases are further obstacles to the successful adoption of big data. However, the adoption is positively influenced by company size and competition intensity. Nam et al. (2015) have investigated the change of influencing factors during the adoption process. As a result, Nam et al. (2015) prove that existing information systems (IS) competence has a positive impact in the beginning of the adoption process, while competitive intensity and financial readiness significantly support the successful implementation of use cases. Bremser et al. (2017) have used the TOE to identify factors that influence the approaches companies use to explore big data potentials. IS competence, perceived complexity of the big data technologies, as well as the financial and strategic readiness of companies were identified as major influencing factors. So far an investigation of the big data adoption process has been carried out only by Chen et al. (2015). They use a multiple case study to describe the implementation phase and corresponding influencing factors. TOE (Tornatzky, L. G., Fleischer, M. & Chakrabarti, 1990), diffusion of innovation (Rogers, 2003) and the IT fashion theory (Wang, 2010) were used as sources for influence factors. The diffusion theory describes the spreading of an innovation among members of a social system (Rogers, 2003). The IT fashion theory highlight the social settings of emerging IT trends, e.g. the influence of consultants and technology analysts (Wang, 2010). According to Chen et al. (2015) the implementation phase involves far-reaching organizational changes that are necessary for the productive implementation of big data applications. As a result, they present a "limbo stage", where companies continuously experiment with big data technologies for a long time and do not proceed to deployment, despite their intent to adopt. This overview of current research shows that influence factors on the adoption process for big data have already been extensively discussed. An investigation of the adoption process was conducted by Chen et al. (2015), but only with focus on the implementation phase. Despite its high relevance, no scientific investigation exists for the initiation phase when companies start to approach the potentials of big data. 3 Conceptual Framework In order to address this research gap by investigating the initial exploration of big data applications, Rogers’ adoption process (2003) is used as a theoretical starting point which is illustrated in figure 1. According to Rogers (2003), the process of innovation adoption is described by two major phases: initiation and implementation, with both phases being separated by an adoption decision. The initiation phase consists of two stages: The first stage agenda-setting is triggered by an organizational problem or by the perception of an innovation. The organizational problem manifests itself through a perceived performance gap, which is the result of internal inefficiency or altered environmental conditions (Damanpour & Schneider, 2006). The observation of an innovation is achieved by continuously scanning the business environment (for example, monitoring competitors or technological developments). Both triggers force enterprises to consider the potentials of innovations. Within the agenda-setting stage the entrepreneurial range of possible reactions with respect to innovations is weighed. For example, a company decides how to react on the availability of a new technology. Figure 1: Innovation adoption process by Rogers (2003) The second stage matching includes all activities that proof whether an identified innovation is suitable for fulfilling the organizational needs in context of the current situation of a company. Typically, some members of an organizational unit explore the capabilities of the innovation to make a prediction on its potential for specific use cases. If this forecast is positive, the implementation phase is triggered in the adoption process. This phase consists of the stages redefining, clarifying and routinizing, and includes all the activities and decisions that are necessary to put the innovation into production. The present research focuses on the initiation phase of the adoption process. The implementation phase is not the subject of this work. It has been investigated already by Chen et al. (2015). 4 Research Design Phenomena around big data adoption are complex and certainly not well understood so far. For this reason, a case study approach is suitable (Dubé & Paré, 2003; Yin, 2003). Our main information sources are in-depth expert interviews with key-informants. Interviewees were heads of business and IT divisions, chief architects and chief strategist. In the sense of a strict implementation of the research design, four established quality criteria were used (Yin, 2003): external validity, internal validity, construct validity and reliability. The external validity focusses on the generalizability of the results. This is ensured by replicating the case studies. In the context of big data adoption, it was decided to conduct a multiple case study. The case studies were selected according to the “literal replication logic” (Dubé & Paré, 2003). In order to ensure a comparable organizational and technological context pure internet companies were excluded in the selection phase of the case studies. In addition, the reference to big data has been validated by scientific big data taxonomies (see, e.g. (Kune et al., 2016)). Big data taxonomies represent a classification scheme and show which technologies, methods and data are typically used in the context of big data. In order to ensure internal validity, the interview guide was developed on the basis of the conceptual framework described in chapter 3. The expert interviews were semi-structured and we kept our questions open to allow interviewees freely to speak. The first part contains general questions about the role and responsibility of the interviewee, the current strategic and tactical challenges of the company and their influence upon dealing with new possibilities of big data. The second part of our questions concentrates on the current use of data, methods and technologies for data-driven decision making as well as corresponding organizational structures and processes. For example, we asked about the relevance of data and data-driven decision making in different organizations and inquired which kind of analytical applications were currently in use. The third and most extensive set of questions was directed upon “why” and “how” organizations explore the potentials of big data. These questions concerned the trigger of big data initiatives, their focus and their organizational setup. Also we inquired the process for the evaluation of big data potentials and the criteria applied therein. Yin (2003) suggests triangulation to ensure construct validity. Within the case studies, different data sources were therefore used. In addition to the key-informant interviews, public and - if available - internal documents of big data initiatives and strategies of the investigated companies were used. Furthermore, interviews with other organizational members, consultants and software vendors specialized on big data adoption were conducted to triangulate the data. In order to minimize errors and biases, the reliability of the case studies was ensured by establishing a case study database. There, we stored all information about the data collection process, the data itself and the case study results. According to Yin (2003), this helps to provide the same results in repeated trials. The data collection started in June 2016 and stretched over a period of seven months. Each interview lasted approximately 90 minutes and was conducted on site or by telephone conference. The conversations were recorded and transcribed. Shortly after each interview, the main points and key findings were recapitulated in a contact summary sheet (Miles, Huberman, & Saldana, 2013). The interviews were then analysed and coded. We used first-level coding (Miles et al., 2013) to identify in particular all statements related to company’s procedures for the initiation phase of big data adoption. Table 1 presents an overview of the participants of the case study. In the case selection, we focused on companies that have more than 10,000 employees and their headquarters in Germany. The investigated companies operate business-to-consumer as well as in business-to-business segments. The interviewees had roles in business and IT and were responsible for big data activities within their organizations. Table 1: Participating companies Industry Number of employees business segment Role of Interviewee 1 Transport >50,000 B2C, B2B Head Domain Architecture 2 Banking >50,000 B2C, B2B Head IT Architecture 3 Insurance >10,000 B2C, B2B Head Group Strategy 4 Manufacturing Vehicle >50,000 B2B IS Chief-Architect 5 Retail Trade >50,000 B2C Head Business Intelligence 6 Utilities >50,000 B2C, B2B Chief Digital IT Strategist 7 Manufacturing Vehicle >50,000 B2B Head Analytics Lab 8 Manufacturing Apparel >50,000 B2C Head Data Analytics Lab 9 Manufacturing CPG >10,000 B2C Head Marketing & Analytics 10 Manufacturing Chemicals >10,000 B2B Head BI Architecture The analysis of these cases was carried out in a twofold way. First, we have used a within-case analysis (Yin, 2003) to extract all characteristic content related to the agenda-setting and matching stages of individual cases. The corresponding results are shown in chapter 5. For data presentation we apply the process definition of Hammer and Champy (1993) , showing in particular input, activities and results. In the second step, a cross-case analysis (Yin, 2003) was conducted and the cases were compared to each other. The result of this comparison is discussed in chapter 6. 5 Results from cases The trigger for a company to deal with the possibilities of a technological innovation is, according to Rogers, a performance gap of the company or the perception of new possibilities (Rogers, 2003). In the case of a technology-driven innovation, a company examines in the subsequently launched agenda-setting phase, how it shall react upon the availability of a new technology (Rogers, 2003). The result is a so-called agenda, which defines the goals of the next steps in the adoption process. The following matching phase examines the extent to which an innovation could be used to address entrepreneurial needs. If a promising forecast can be given for a specific application scenario, the innovation will be proposed for implementation (Rogers, 2003). Table 2 shows an overview of the agenda-setting and matching phase. Table 2: Agenda Setting and Matching by Rogers (Rogers, 2003) Agenda-Setting Input Performance gap or perception of an innovation Activity Consideration of possible organizational reactions Result Agenda, containing company-specific goals for the next activities in the adoption process Matching Input Agenda Activity Exploration whether organizational needs can be addressed with the innovation Result Decision-making document for the possible implementation of an use case In all observed cases, the big data adoption process was initiated by senior management. Always the perception of the hype surrounding big data was decisive, not a search for ways to solve existing requirements. This is described by an interviewee from case 8 as follows: "It was our former CIO [...], who said that big data is a megatrend, which I will definitely not miss ...". The long-term expectations companies relate to big data span from possible improvements of existing processes to entirely new business services or business models. A quote from case 5 confirms this: "... [we hope to use big data] either for rationalization or for other value-creation opportunities, which are not necessarily connected to rationalization, but where really new fields are opened up." In order to determine the direction of the next big data activities, the topic was discussed at senior management level in all companies. As a result, first objectives and corresponding steps have been defined. Personnel responsible for big data initiatives and resources for project teams were named. Table 3 shows the short-term goals, i.e. the agenda, for the different big data initiatives. Table 3: Agenda of the interview participants 1 Portfolio for innovative data-driven products, services, business models 2 Possibilities for the cost-neutral minimization of technological hurdles for future big data applications 3 Roadmap for a systematic development of internal capabilities to use big data technologies and to provide data appropriately 4 Consistent data basis for company-wide analyses, initially for the identification of potential efficiency enhancements within the existing value chain 5 Opportunities for a future-oriented development of a data and technology platform for analytical applications 6 Portfolio of innovative digital products for public, commercial and private customers 7 List of necessary requirements with respect to technologies and organizations for future data-driven product innovations as well as a consistent data base 8 Potentials for innovative products and process optimizations along existing value chains 9 Possibilities of data analyses to increase the efficiency of existing processes with focus on marketing and sales 10 Data basis out of existing and new data for future data-driven services As Table 3 shows, the aim of companies in cases 1, 6, 8, 9 was directed towards application use cases. For the remaining enterprises (cases 2, 3, 4, 5, 7, 10), the creation of a good technological starting point and a solid basis of data was at the forefront. Following their agenda, the examined companies have carried out different activities in the matching phase. The main activities of the investigated cases are shown in Table 4. Table 4: Main activities of the matching phase 1 - Search for new, data-driven products and customer services carried out by business departments - Evaluation of use cases in terms of their economic value - Implementation of prototypes in cross-functional project teams partly with external support by consultancies - Validation of prototypes in selected market segments 2 - Analysis of existing requirements conducted by the IT architecture organization to identify opportunities for implementing big data technologies - Search for opportunities to replace existing IT components cost-neutrally with big data technologies (for example Hadoop file system instead of Oracle cluster) 3 - Gap analysis for big data capabilities and corresponding pre-requisites through industry-specific big data use cases - Roadmap definition for the development of missing capabilities focussing on big data technologies and data management through a work group guided by middle management 4 - Definition of performance key performance indicators (KPI) through top management for cross functional business processes - Validation of KPIs and identification of weaknesses in the data architecture via data lab - Planning of a cloud data lake to provide a consistent and central database throughout the enterprise 5 - Continuous evaluation of technology innovations for the evolution of the central data platform through employees of the internal IT department - Identification and integration of new data sources for future big data use cases 6 - Search for new data driven customer services in the B2C and B2B segment by cross-functional teams staffed with employees from business and IT departments - Development of potential services as prototypes using agile project methodologies - Evaluation of the big data use cases from an economic perspective as a proof of concept (PoC) within a time-limited test phase in selected market segments 7 - Reduction of data silos and creation of a central data basis - Exploratory implementation of industry-specific big data uses cases in a laboratory environment - Evaluation of typical big data requirements and their impact on the technology landscape and organization 8 - Search for use cases in business divisions with focus on potential growth fields and customer benefits - Establishing a laboratory within the IT department for exploratory data analysis; recruitment of data scientists and data engineers - Identification of lighthouse use cases and implementation as PoC in a laboratory environment 9 - Identification of new opportunities for data driven personalized customer services in marketing and sales departments - Use case driven extension of the enterprise-wide data platform through external IT providers focussing on the integration of external data, e.g. media data from external agencies 10 - Development of a cloud data lake and integration of enterprise data by the business intelligence and advanced analytics departments - Design and implementation of data governance processes - Recruitment of Data Scientist to investigate future big data use cases The matching phase aims at the provisioning of appropriate information to decide upon the adoption of specific big data use cases. As Table 4 shows, some enterprises (cases 2, 3, 4, 5, 7, 10) develop a data platform and capabilities to deal with big data technologies, in line with their agenda. Investigations of company-specific big data use cases were planned by these companies after these activities. Companies (cases 1, 6, 8, 9) which implement and evaluate prototypes for specific use cases, take these into a project or innovation portfolio in case of a positive adoption decision. There, the use cases compete with other enterprise projects for the resources required for a productive implementation. 6 Discussion A comparison of the case studies shows that three different approaches can be identified. Figure 2 illustrates this in a schematic way. The first approach (Business First) focuses entirely on a business perspective. Companies in this category search for use cases with high expected business value. These use cases span from possible improvements of existing processes to entirely new business services or business models. A quote from case 1 emphasizes this: "... on the one hand you have to adapt the existing business to the current market [...], i.e. to improve existing processes [...] On the other hand you have to invent new processes or tackle new business models”. A technical evaluation of the use cases with regard to their possible integration into existing IT landscapes is not carried out during the initiation phase. This is shown by a quotation from case 6: "At the beginning, the integration is actually subordinated [...] If a product that has been developed in a sandbox is successful on the market, of course, we want to integrate it." Typically, big data use cases are developed as stand-alone IT systems with necessary big data technologies and data in a laboratory environment. Developed prototypes are then tested with selected market participants. If this phase is successful, the use cases are suggested for productive implementation. For example, case 6 states: "[the goal is to] test use cases in 6-12 months with several thousands, maybe even ten thousands of customers in real use. And then there's the decision: go or no-go." In case of a positive decision, applications will initially operate as independent IT systems, being then integrated step by step into existing IT landscapes. Companies with this approach can be found in cases 1, 6, 8 and 9. Figure 2: Approaches to the analysis of big data potentials: A) Business First, B) Platform Building, C) Data Integration The second approach (Platform Building) aims upon the development of a technology and data platform for big data. For this purpose requirements derived from industry-specific use cases are analyzed for orientation. In addition, existing business demands were used to introduce new technologies, for example in case 2: "... it is the strategy [...] to use new technologies for existing demands that we are obliged to do, in order to [...] reduce the hurdle for hardly-calculable [big data] use cases.". Companies with this approach want to keep the monetary expenses for future use cases as low as possible – mostly due to financial restrictions. Therefore, e.g. in case 2, existing data management technologies are successively substituted by big data technologies (e.g. hadoop for archiving bank account movements). To keep pace with current developments – although business departments have not identified convincing big data use cases yet – is another observed motivation for platform building, e.g. in case 5: "In memory technology, [...] in the next 10 years, this will be used in all areas. Then I can already invest now, without a concrete use case in mind". The big data platform resulting from this approach forms the basis for the subsequent identification and evaluation of big data use cases. This approach can be found in case 2, 3, 5 and 7. The primary objective of the third approach (Data Integration) is to provide a consistent basis of data for future analyses. Consistent and integrated data are considered as a fundamental basis for all further developments in the area of big data. For instance, case 10 emphasizes: "... this is our approach [...], we want to build up an enterprise data repository, [...] step by step, to integrate and organize all data there, to build a semantic network". In the cases following this approach, the created data platform is first used for traditional analyses. If they proof successful, companies plan to deal with big data use cases and technologies. Case 4 and 10 follow this approach. A comparison of the three approaches shows that Business First focuses on the realization of business potentials and initially neglects the integration of new technologies and data sources. This is in contrast to the approaches Platform Building and Data Integration. There, the integration of technologies and data is of primary interest and seen as a necessary step towards the successful adoption of big data. In Business First, investigations of the required efforts for implementing use cases into the productive IT landscape are postponed to a later stage. Also investigations of constraints from a technical or data management perspective are ignored here during the initiation phase. If a lack of necessary technological or methodical competences arises, organizations procure external resources to bridge these gaps, e.g. through external IT service provider or consultancies with specialist knowledge. For instance, case 9 states: “Our IT is a profit center. They do not expose employees to innovative topics; they only want to put employees into projects. But that’s a problem. People are not trained and further educated [..] Therefore, I have to hire Accenture or any other consultancy, I pay the double day rate, but it can realize my use cases in half of the time.” Differences can also be found in the organizational set ups of the three approaches. In Business First, the identification of possible use cases is carried out by business departments, while the evaluation and implementation of prototypes is mainly conducted within cross-functional teams staffed with employees from business and IT. In Platform Building and Data Integration, traditional IT projects are carried out in IT departments introducing and integrating new technologies and data sources. IT-driven laboratory environments are often used to identify requirements for a future big data infrastructure. This is emphasized by a statement from case 7: “We [the data lab] do not necessarily need to solve the Autonomous Driving. We [the company] have enough engineers for that. Our goal is to provide the technical possibilities that this can work.” 7 Summary and outlook In this paper we have investigated how companies start exploring big data potentials and how companies proceed to an adoption decision. The theoretical framework was the adoption process of Rogers (2003), whose first stages of the agenda setting and matching phases were examined in detail based on a multiple case study. As a result, three different approaches could be identified. Companies either initially focus on business aspects, or on a systematic build-up of a big data technology and data platform. The next step is to examine in detail the influence factors that are responsible for these different strategies and the advantages and disadvantages associated with them. The insights gained from current and upcoming studies will be then used to construct a method kit for the identification and evaluation of potential big data use cases, catering to the specific needs of individual enterprises. References Agrawal, K. P. (2015). Investigating the Determinants of Big Data Analytics ( BDA ) Adoption in Asian Emerging Economies. Academy of Management Proceedings, 2015(1), 11290. https://doi.org/10.5465/AMBPP.2015.11290abstract Bremser, C., Piller, G., & Rothlauf, F. (2017). Strategies and Influencing Factors for Big Data Exploration. In Proceedings of the American Conference on Information Systems. Boston, USA: AIS Electronic Library. Brown, B., Chui, M., & Manyika, J. (2011). Are You Ready for the Era of “Big Data”? McKinsey Quarterly, 4(1), 24–35. Chen, H.-M., Kazman, R., & Matthes, F. (2015). Demystifying big data adoption: Beyond IT fashion and relative advantage. In DIGIT 2015 Proceedings. Fort Worth, USA: AIS Electronic Library. Constantiou, I. D., & Kallinikos, J. (2015). New Games, New Rules: Big Data and the Changing Context of Strategy. Journal of Information Technology, 30(1), 44–57. https://doi.org/10.1057/jit.2014.17 Damanpour, F., & Schneider, M. (2006). Phases of the adoption of innovation in organizations: Effects of environment, organization and top managers. British Journal of Management, 17(3), 215–236. https://doi.org/10.1111/j.1467-8551.2006.00498.x Davenport, T. H., Barth, P., & Bean, R. (2012). How “Big Data” is Different. MIT Sloan Management Review, 54(1), 22–24. Dubé, L., & Paré, G. (2003). Rigor in Information Systems Positivist Case Research: Current Practices, Trends, and Recommendations. MIS Quarterly, 27(4), 597–636. https://doi.org/10.2307/30036550 Fichman, R. G. (2000). The Diffusion and Assimilation of Information Technology Innovations. Framing the Domains of IT Management, 105–127. https://doi.org/10.1.1.24.4539 Hammer, M., & Champy, J. (1993). Reengineering the corporation. A manifesto for business revolution. Business Horizon, 36(5), 90–91. Kart, L. (2015). Big Data Industry Insights 2015. Retrieved March 30, 2017, from http://public.brighttalk.com/resource/core/80421/september_29_industry_insights_lkart_118453.pdf Kiron, D., Prentice, P. K., & Ferguson, R. B. (2014). The Analytics Mandate. MIT Sloan Management Review, 55(4), 1. Kune, R., Konugurthi, P. K., Agarwal, A., Chillarige, R. R., & Buyya, R. (2016). The Anatomy of Big Data Computing. Software - Practice and Experience, 46(1), 79–105. https://doi.org/10.1002/spe.2374 Malaka, I., & Brown, I. (2015). Challenges to the Organisational Adoption of Big Data Analytics : A Case Study in the South African Teleco. Industry. In Proceedings of the South African Institute of Computer Scientists and Information Technologists Conference (p. 27). Stellenbosch, South Africa: ACM Digital Library. Miles, M. B., Huberman, A. M., & Saldana, J. (2013). Qualitative Data Analysis: A Methods Sourcebook (3rd ed.). Los Angeles: SAGE Publications. Nam, D. W., Kang, D., & Kim, S. H. (2015). Process of Big Data Analysis Adoption: Defining Big Data as a New IS Innovation and Examining Factors Affecting the Process. In Proceedings of the Hawaii International Conference on System Sciences (pp. 4792–4801). Kauai, HI, USA: IEEE. Rogers, E. M. (2003). Diffusion of Innovations (5th ed.). New York: Free Press. Sivarajah, U., Kamal, M. M., Irani, Z., & Weerakkody, V. (2017). Critical Analysis of Big Data Challenges and Analytical Methods. Journal of Business Research, 70, 263–286. https://doi.org/10.1016/j.jbusres.2016.08.001 Sun, S., Cegielski, C. G., Jia, L., & Hall, D. J. (2016). Understanding the Factors Affecting the Organizational Adoption of Big Data. Journal of Computer Information Systems, 1–11. https://doi.org/10.1080/08874417.2016.1222891 TechAmerica Foundation. (2012). Demystifying Big Data: A Practical Guide to Transforming the Business of Government. Retrieved March 30, 2017, from https://bigdatawg.nist.gov/_uploadfiles/M0068_v1_3903747095.pdf Tornatzky, L. G., Fleischer, M. & Chakrabarti, A. K. (1990). Technological Innovation as a Process. In Processes of Technological Innovation. Lexington: Lexington Books. Wang, P. (2010). Chasing The Hottest IT: Effects of Information Technology Fashion on Organizations. MIS Quarterly, 34(1), 63–85. Yin, R. K. (2003). Case Study Research: Design and Methods. Case Study Research: Design and Methods (3rd ed.). New York: SAGE Publications. Citizen adoption of e-government services – Evidence from Hungary MARTA ARANYOSSY 4 CORRESPONDENCE ADDRESS: Marta Aranyossy, Ph.D., Assistant Professor, Corvinus University of Budapest, Corvinus Business School, Budapest, Fővám tér 8., Hungary, marta.aranyossy@uni-corvinus.hu DOI https://doi.org/10.18690/978-961-286-170-4.3 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract In a citizen centric approach – which became increasingly popular in the last decade – e-government success begins with citizens starting to use e-government systems, solutions, services. In line with this our paper investigates the factors – presented by the technology acceptance literature – influencing e-government service usage, on a large representative Hungarian sample concerning a wide range of B2C public administration services. Our results imply that the Hungarian government can further increase the usage of e-government services by influencing effort expectancy, trust of internet, facilitating conditions, user experience or habits. Keywords: • e-government • technology adoption • UTAUT • e-government adoption • 1 Introduction Some current streams of e-government research focus on customer centric service development and performance assessment (see Scott et al. 2009; Alomari 2012; Barbosa 2013; Nica 2015; Carter et al., 2016). This perspective acknowledges the key role of citizens as end users of e-government services (Clarke – Pucihar 2013). One of the main questions of citizen centric e-government research – and also e-government research in general – is: why do or do not citizens use e-government services, what motivates technology acceptance on the field (Bannister – Connolly, 2012; Rana et al., 2013). In order to develop citizen-centred electronic services that create value for citizens and society alike, government agencies must explore and understand the factors that drive adoption and usage of these digital innovations first (Carter – Belanger 2005). The theories of technology acceptance are relatively often used in e-government literature: amongst the most cited 15 e-government articles 5 used TAM (Technology Acceptance Model) or DOI (Diffusion of Innovation) (Belanger – Carter 2012, 371). Rana et al. (2013) identified 54 articles using TAM, 20 using DOI and 13 based on UTAUT (United Theory of Acceptance and Use of Technology) in the e-government literature. So technology adoption by individuals has been the subject of several studies, as we will show in a literature review in the next section. Although this field is not underresearched most of the studies fall into the following categories:  analysing technology adoption of e-government services under hypothetical circumstances (e.g. Alomari et a. 2014; Nemeslaki et al. 2016 );  analysing only intent to use not actual usage of an e-government service (e.g. Carter – Belanger 2005, Lin et al. 2011);  analysing actual usage of only one system or service (e.g. Hung et al. 2006, AlAwadhi – Morris 2008). In this paper we aim to present a more comprehensive research on e-government adoption: we explore adoption factors of 12 different Hungarian e-government services, focusing on actual usage, using a large and representative sample. These 12 service areas represent the whole palette of currently available e-government services in Hungary. Also, our research is broad from a theoretical point of view: we examined many of the possible factors of technology acceptance suggested by the literature, this way presenting results independent of the different models. We believe that this rich evidence will give researchers and practitioners a more detailed view of factors driving citizen adoption of e-government services. 2 Literature review In the end, the success of e-government initiatives depends significantly on whether users – public employees or citizens – are willing to accept and use the innovation, the new tool, system or service. One of the most utilised models for the exploration of information technology innovation acceptance is the TAM (Technology Acceptance Model – see Davis 1989; Venkatesh – Davis 2000). The essence of the model is rather simple: attitudes regarding use, intention to use, and actual use are defined by two variables: perceived usefulness of the system and its perceived ease of use. In the IT literature, more than 100 empirical studies tested these simple relationships of the TAM model. The effect of perceived usefulness was supported in 74% of these studies, while ease of use often proved to be a necessary, but not sufficient condition (Lee et al. 2003). Regarding e-government innovation adoption, a considerable number of empirical research papers utilised the TAM model, or its expanded versions. Carter and Belanger (2005) also performed their research using a modified version of TAM, finding that regarding e-government services, three primary factors define citizens’ intention to use: perceived ease of use, compatibility (congruency with and similarity to citizens’ normal way of communication or transactions) and reliability (whether users judge the service to be reliable and safe, and trust it). Many also used the TAM model regarding the acceptance of other e-government innovation, for example, e-voting technology (Schaupp – Carter 2005; Chiang 2009; Choi – Kim 2012, Nemeslaki et al. 2016), and found its explanatory power to be strong. The original TAM model was extended by many, and the UTAUT model (United Theory of Acceptance and Use of Technology, Venkatesh et al. 2003; see Figure 1) attempted to unify these improvements into a single model. In the last decade, the use of the UTAUT model gained acceptance in the e-government literature as well (e.g. Gupta et al. 2008; Powel et al. 2012). Here, along with the original two TAM variables, the moderating effects of social environmental influence, other workplace driving factors, and individual characteristics (users’ age, gender, experience, and voluntary nature of use) are also present in the model’s context. Hung et al. (2006) used an expanded TAM and UTAUT model to research factors influencing willingness to use a Taiwanese online tax system, and found the following significant factors that effected attitudes regarding use: perceived usefulness, perceived ease of use, perceived risk, trust, compatibility, as well as external influence, interpersonal influence, self-efficacy (one’s mental image of one’s own efficiency) and facilitating conditions. Along with external factors, factors such as media influence should be mentioned, while interpersonal influence may take the form of colleagues and friends’ positive opinions on e-government services, and their encouragement to use it. Finally, the easy accessibility and availability of the necessary devices, hardware, and software is a must among the facilitating conditions. Perceived UsefulnessPerceived Ease of UseBehavioral IntentionUse BehaviorSocial InfluenceFacilitating ConditionsModerators:Age, Gender, ExperienceHedonic MotivationPrice ValueHabit Figure 1: The TAM (elements in bold; Davis 1989) and the expanded UTAUT model (elements underlined, Venkatesh et al. 2003) and UTAUT 2 (the entire figure; Venkatesh et al. 2012) While UTAUT is a popular model, Venkatesh et al. (2012) proposed a further extended version, UTAUT2. With three new constructs – hedonic motivation, price value, and habit – the predictive power of their model increased significantly. Diffusion of Innovation (DOI, Rogers 2003) is also a widely used model in IT adoption research, but e-government researchers claim that its core constructs are similar to and substitutable by TAM factors: relative advantage with perceived usefulness and complexity with perceived ease of use (Carter – Bélanger 2005; Colesca – Dobrica 2008). Rana et al. (2013) compared the explanatory power of the five most used IT adoption models in a meta-analysis based on 87 studies of citizen centric e-government services. Based on their results, all of the basic relationships of TAM could be confirmed. This was the most widely used model in research focused on e-government adoption, and this also seemed the most appropriate one for studies that focus on citizens. Although DOI was the second most common research framework, only a small number of its relationships were validated, and empirical research only concentrated on three of its explanatory variables (compatibility, complexity, relative benefits). The most important of the new variables of the UTAUT model is the social effect, while the effect of facilitating conditions was under-researched. This meta-analysis also highlighted that factors – that all central models lack – such as trust, safety, privacy, and risk – appear rather often in empirical studies on e-government adoption and seem to have significant effects. Naturally, along with using the theories of the scientific mainstream, independent e-government acceptance theories have also been constructed. Ziemba and co-authors (2013, 2015), for example, examine factors of a successful e-government in one such model. In their model of e-government adoption at the local and state levels, they attempt to explain the three factors of e-government adoption (ICT availability, ICT competence and awareness, ICT use) with economic, socio-cultural, technological, and organisational factors. In Table 1 we summarise a number of further e-government technology adoption models, to demonstrate the diversity of theoretical approaches. Table 1: Special e-government technology adoption models (partially based on Panda – Sahu 2013) The model of Ziemba et al. (2015) is one of the few e-government adoption researches in Central and Eastern European context. While the usage of TAM is not without exception in Hungarian technology adoption research as well (e.g. Keszey – Zsukk 2017), Hungarian e-government adoption research is still rare (e.g. Nemeslaki et al. 2016; Molnár et al., 2017). 3 Research model and methodology The empirical basis of or research was the Good State Public Administration Opinion Survey (henceforward referred to as Survey; Kaiser 2017) carried out in Hungary 2017. The data collection was planned, tested and carried out by Szociometrum Social Science Research. The survey questions were tested on a representative sample for the adult (age 18+) Hungarian population. The sampling method was multistage, proportionally stratified probability sampling, while the database was also corrected ex post with matrix weighting procedure concerning age, gender, region, settlement type and education. (See descriptive statistics in Table 2.) Table 2: Descriptive statistics of the Good State Public Administration Opinion Survey sample The Survey contained 70 questions, some with many sub-questions, of which we are only using some (see Appendix 1) to explore the influencing factors of e-government service usage. The Survey provided the opportunity to use a large (n=2506) representative database, with data about citizens’ usage and experience of 12 different areas of e-government services (see list in Table 3) and the citizens general background as well. For these 12 administrative areas citizens were asked whether they had to use these public services in the past 3 years, if yes, whether they did that themselves, and finally if yes, in what way did they gather information (in person, via phone, via e-mail, via website) or handled the necessary transaction (in person, via phone, online or via app, via post). This data gave us the opportunity to compare citizens using online and traditional channels in the information or the transaction phase of public administration on different fields. Table 3: List of government service areas and frequency of online usage (11. Administration of construction affairs was omitted from further analysis due to the low subsample size) To test the different factors of e-government adoption we used as many variables from the literature (summarized in the previous section) for which we had the corresponding relevant data in the Survey. This gave us the opportunity to test the effect of the following factors:  Effort expectancy (Perceived ease of use)  Trust of internet  Trust of government  Facilitating conditions  Experience  Habit  Age  Gender The basis of this factor selection was Venkatesh et al. (2012). We analysed most of the factors and moderators included in UTAUT2 (presented in Figure 1). Notice that we did not use one of the key variables of TAM, perceived usefulness, as the survey question concerning this was too distant and biased. We also did not include price value and hedonic motivation, as by definition these concepts are less relevant in e-government settings and also did not have a real history in e-government adoption research. Our list of factors under examination is also similar to e-government adoption research carried out by Carter – Belanger (2005) in two respects: we tested many potential factors from different theoretic models, and we included the factor of trust. Trust related concepts are amongst the most common extensions of TAM in e-government research and their significance is shown in many studies (e.g. Schaupp – Carter 2005, Powell et al. 2012). Just like these previous studies we included both trust in the internet and trust in the government in our research. For measurement of the factors we used direct survey variables in case of age and gender, and factor analysis in case of the other six. The original questions, their measurement, factor component matrices and KMO statistics can be found in Appendix 1. Although answers to the key construct questions were measured on a Likert scale, if a Likert scale is equidistant it behaves more like an interval-level measurement and therefore, can be viewed as an interval scale and used for factor analysis (see Carifio – Perla 2007). The operationalization methods of the Good State data collection did not allow us to build regression or structural equation model – common in the TAM literature – for the analysis. So to test the relationships between the potential influential factors and actual usage we used traditional association metrics and statistical tests to identify significant differences between online and offline e-government service users. Being aware of the limits of the measurement scales used in the data collection not only ANOVA-based mean tests, but also nonparametric Mann Whitney U test and median tests where employed to identify significant differences of distributions and medians. The limitations of the database is also the reason behind our decision that age, gender and experience were also be tested as potential influential factors and not as modifiers. Although the database served as a source of many limitations, but it was also advantageous from another point of view: it allowed us not only to test adoption factors of actual usage, but we could distinguish between informational and transaction level of e-government usage as well. In summary, the guiding research question of our study was: Which factors – presented in the technology acceptance literature – are associated with informational or transactional e-government service use in comparison to traditional forms of public service use concerning a wide range of B2C public administration services in Hungary? The research model is represented in Figure 2. 4 Results A summary of our results is presented in Table 4 and Figure 2. We indicated all the connections, where we found statistically significant differences between mean or median values of online and offline users of administrative government services. The sample for construction affairs administration (11.) was too low to carry out in depth statistical analysis or find significant results – so in the further discussion we focus on the other eleven service areas. Table 4: Results by government service areas where factor values for online users are significantly higher than factor values for offline users (*: p<0.05, **: p<0.01, bold: difference in factor is larger than 25% of the range of its value) Figure 2: Model Results (the identification number of the government service area – see numbered list in Table 3 – is indicated above the arrows where factor values for online users are significantly higher than factor values for offline users, *: p<0.05, **: p<0.01) We see that the most common relationships related to informational e-government usage are with habit, facilitating conditions and internet trust: the level of these factors was significantly higher for online users than for offline ones in at least nine of the eleven observed service areas. These factors are similarly important in case of e-government transactions as well, although here the level of experience is also significantly higher for online users. On the other hand, trust in the government, age or gender does not seem to differentiate between citizens using e-government services and citizens that are choosing offline administrative options (only in some rare, specific service areas). 5 Discussion Concerning the original factors of TAM our results seem to be less decisive. The effect of performance expectancy was not measurable based on our database and effort expectancy seems to be an important determinant in only four out of the eleven Hungarian e-government service areas. These areas are: income and other tax administration, government issued document and motor vehicle administration – most of these being the larger subsamples and still showing significant relationships only at p<0.05 levels. So effort expectancy seems not to be the main factor behind Hungarian citizens’ decision of using online or offline e-government platforms. One of the reasons behind this result could be that we tested effects on actual e-government usage only, and not on behavioural intention – while according to the original TAM, effort expectancy has a direct effect on intent to use and not on actual usage. If we look at the extended UTAUT2 factors we see more significant results. Habit seems to be the most important differentiator between offline and online usage in all e-government service areas. The relationship with habit is significant in all 22 cases (all 11 analysed service areas, regarding both online information and transaction), and in 17 instances the level of habit for online users is more than 25% higher than others (as a percentage of the range of this variable). We conceptualized habit as prior behaviour (Kim – Malhotra 2005), so our results suggest that prior general administrative behaviour and platform choices influence electronic government adoption of Hungarian citizens the most. Habit has been one of the two factors in UTAUT2 with assumed direct effect on actual usage – this direct effect has been found significant originally by Venkatesh et al (2012) and in case of e-government usage here as well. The other factor of UTAUT2 (and UTAUT) with hypothesised direct effect on usage is the factor of facilitating conditions. Facilitating conditions – measured by the accessibility of devices and internet connection – was the second most important factor in our study of Hungarian e-government adoption as well: found significant in 18 out of 22 cases. According to a study of the Hungarian Central Statistical Office (2015) 76% of adult Hungarian citizens are using computers and the internet – only 1-2% less than the European average (although some differences between rural areas might still exist; Csótó – Herdon, 2008). This means that the significant effect of facilitating conditions for some citizens is not caused by a nationwide lagging of ICT penetration. The generally high level of connectivity might also be a reason why there is a significant but never above 25% difference in the value of the facilitating conditions factor when comparing online and offline government service users. We mentioned that internet and government trust are frequent extensions of the TAM/UTAUT in e-government literature. The phenomenon that only internet trust has a significant effect on e-government adoption is also not uncommon. Trust in the government seemed not to be a significant influencer of adoption in international studies (e.g. Powell et al. 2012) or in other Hungarian studies (NEMESLAKI et al. 2016) – and that is what we see in our case as well. Examining a broad spectrum of e-government areas we found a statistically significant relationship between citizens’ trust in the government and e-government adoption only in four cases. Interestingly enough three of these cases are tax and pension related transactions, which have serious financial components – these seem to be the cases where higher level of government trust leads to more online transactions. Conversely, trust in the internet seems to have a significant effect in a wide range of Hungarian e-government areas (in 16 analysed cases). In 8 instances the trust level of online government service users is above 25% higher than the offline ones. This is not entirely surprising in Hungarian context, as the low level of internet trust among Hungarians was one of the factors why Hungary lagged behind in terms of e-commerce and especially e-payment market developments (e.g. Aranyossy – Juhász 2013; Fehér – Varga 2017). Concerning Hungarians’ e-voting attitude Nemeslaki et al. (2016) found that internet trust is the second most important factor, and the statistical data collection of HCSO (2015) also stated that 11% of Hungarian citizens are not using e-government transactions because they do not trust the systems enough to give their personal data. Our results also confirm this important role of trust in the internet in Hungarian e-government adoption. Although we handled the variables of age, gender and experience differently than the original UTAUT2 by analysing them as factors and not moderators, our results here are also noteworthy. While based on our analysis the age and gender of the citizens are not differentiating factors in terms of administrative channel choice (offline vs. online) – experience is. Citizens choosing six of the informational and nine of the transactional e-government services have significantly higher level of experience in other, non-governmental online transactions than offline citizens. This also suggests, that experience is more important when citizens have to choose a transactional channel, and less for collecting information online – so to administer online people rely more on prior experiences of e-transactions. 6 Conclusions, limitations and implications for theory and practice In this paper we tested six factors of the UTAUT2 model extended with trust to analyse e-government adoption on a large Hungarian sample. One of the novelties of our research was that we examined factors of actual e-government usage and not only the behavioural intent. Also the robustness of our findings is increased by the fact that not only one but eleven, a near total spectrum of Hungarian G2C administration service areas were examined. We found that the key factors differentiating e-government users are habit, trust in the internet and facilitating conditions, while in case of online transactions prior e-commerce experience is also important. Practitioners might also be interested in some of the detailed results of analysis on the level of the eleven individual government service areas. The service with the highest proportion of online users (27% vs. the average 9,6%) was income tax administration. This is not surprising knowing that the Hungarian tax authority informs about the possibilities of the online income tax administration option every citizen yearly, strongly arguing for the usage of the online platform. Also, at the beginning of 2017 a new income tax administration service was launched making the tax declaration process faster and easier and the usage of the platform more user friendly. While the Hungarian online income tax administration platform seems to be a success it is still true that all of the examined factors except for age and gender show a significant effect on adoption (see Appendix 2), so there are still possibilities to increase the number of online users by influencing citizens’ internet trust or facilitating conditions for example. On the other hand, the most frequently used administrative service seems to be the administration of government issued documents: more than 41% of the citizens had to deal with this process. While many people search for information online regarding this process only 6% of them chose online transactional options. This gives the government an opportunity to have a significant impact on e-government penetration based – partly – on our results as well: by influencing effort expectancy, trust of internet, facilitating conditions, experience or even habit if possible. Methodological limitations of our study could also guide future research. A more model-specific data collection method – including direct questions regarding performance expectancy and more detailed Likert-scales to measure – would have supported more complex analysis, potentially even with PLS method. On the other hand our data and analysis could be used to draw more in depth conclusions regarding the individual e-government areas – the length of this paper did not allow the publication of these details. Acknowledgments This paper has been written with the support and within the framework of: KÖFOP 2.1.2 – VEKOP – 15-2016-00001 Public Service Development for Establishing Good Governance: Digital Governance and Digital Government Research Program. References AlAwadhi, S. & Morris, A. (2008): The Use of the UTAUT Model in the Adoption of E-government Services in Kuwait. Proceedings of the 41st Hawaii International Conference on System Sciences. 1530-1605/08 IEEE https://doi.org/10.1109/HICSS.2008.452 Alomari, M. K., Sandhu, K., & Woods, P. (2014): Exploring citizen perceptions of barriers to e-government adoption in a developing country. Transforming Government: People, Process and Policy, 8(1), 131-150. https://doi.org/10.1108/TG-05-2013-0013 Aranyossy, M., & Juhász, P. (2013): Értékteremtés e-kereskedelemmel–kitörési lehetőség a recesszió sújtotta magyar kiskereskedelemben?/Value creation of e-commerce–Chance to break out for the recession-hit Hungarian retail sector?. Vezetéstudomány/Budapest Management Review, 44(11), 16-33. Azadegan, A. – Teich, J. (2010): Effective Benchmarking of Innovation Adoptions: A Theoretical Framework for e-Procurement Technologies. Benchmarking: An International Journal, 17(4), 472–490. https://doi.org/10.1108/14635771011060558 Barbosa, A. F., Pozzebon, M., & Diniz, E. H. (2013): Rethinking E‐Government Performance Assessment From A Citizen Perspective. Public Administration, 91(3), 744-762. https://doi.org/10.1111/j.1467-9299.2012.02095.x Bannister, F., & Connolly, R. (2012): Forward to the past: Lessons for the future of e-government from the story so far. Information Polity, 17(3, 4), 211-226. Bélanger, F. – Carter, L. (2012): Digitizing government interactions with constituents: an historical review of e-government research in information systems. Journal of the Association for information Systems, 13(5), 363. https://doi.org/10.17705/1jais.00295 Carifio, J., & Perla, R. (2007): Ten common misunderstandings, misconceptions, persistent myths and urban legends about Likert scales and Likert response formats and their antidotes. Journal of Social Sciences, 3(3), 106-116. https://doi.org/10.3844/jssp.2007.106.116 Carter, L. – Bélanger, F. (2005): The utilization of e‐government services: citizen trust, innovation and acceptance factors. Information Systems Journal, 15(1), 5–25. https://doi.org/10.1111/j.1365-2575.2005.00183.x Carter, L., Weerakkody, V., Phillips, B., & Dwivedi, Y. K. (2016): Citizen adoption of e-government Services: Exploring citizen perceptions of online services in the United States and United Kingdom. Information Systems Management, 33(2), 124-140. https://doi.org/10.1080/10580530.2016.1155948 Chiang, L. (2009): Trust and security in the e-voting system. Electronic Government, an International Journal, 6(4), 343–360. https://doi.org/10.1504/EG.2009.027782 Choi, S. O. – Kim, B. C. (2012): Voter intention to use e-voting technologies: security, technology acceptance, election type, and political ideology. Journal of Information Technology and Politics, 9(4), 433–452. https://doi.org/10.1080/19331681.2012.710042 Clarke, R., & Pucihar, A. (2013): Electronic interaction research 1988–2012 through the lens of the Bled eConference. Electronic Markets, 23(4), 271-283. https://doi.org/10.1007/s12525-013-0144-4 Colesca, S. E., & Dobrica, L. (2008): Adoption and use of e-government services: The case of Romania. Journal of applied research and technology, 6(3), 204-217. Csótó, M., & Herdon, M. (2008). Information technology in rural Hungary: plans and reality. in: Rural Futures: Dreams, Dilemmas, Dangers. University of Plymouth, UK Davis, F. D. (1989): Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Ouarterly, 13(3), 319–339. https://doi.org/10.2307/249008 DeLone, W. H., & McLean, E. R. (1992): Information systems success: The quest for the dependent variable. Information systems research, 3(1), 60-95. https://doi.org/10.1287/isre.3.1.60 Fehér, P. & Varga, K. (2017): Using Design Thinking to Identify Banking Digitization Opportunities–Snapshot of the Hungarian Banking System. BLED 2017 Proceedings. 39. http://aisel.aisnet.org/bled2017/39 (24.01.2018) Guha, J. – Chakrabarti, B. (2014): Making e-government work: Adopting the network approach. Government Information Quarterly, 31(2), 327–336. https://doi.org/10.1016/j.giq.2013.11.008 Gupta, B. – Dasgupta, S. – Gupta, A. (2008): Adoption of ICT in a government organization in a developing country: An empirical study. The Journal of Strategic Information Systems, 17(2), 140–154. https://doi.org/10.1016/j.jsis.2007.12.004 HCSO (2015): Infokommunikációs (IKT-) eszközök és használatuk a háztartásokban és a vállalkozásokban, 2014 / ICT devices and their usage in households and enterprises, 2014. Hungarian Central Statistical Office, http://www.ksh.hu/docs/hun/xftp/idoszaki/ikt/ikt14.pdf (23.01.2018) Hung, S. Y., Chang, C. M., & Yu, T. J. (2006): Determinants of user acceptance of the e-Government services: The case of online tax filing and payment system. Government Information Quarterly, 23(1), 97-122. https://doi.org/10.1016/j.giq.2005.11.005 Kaiser, T. (2017): A Jó Állam kutatások bemutatása: koncepciók és perspektívák. Summary of the Good State research at National University of Public Service on 18.10.2017, Budapest, http://akfi-dl.uni-nke.hu/jo_allam_jelentes_2017/prezentaciok/plenaris/KaiserTamas_JoAllamJelentesBemutatas.pdf (18.01.2018) Keszey, T., & Zsukk, J. (2017): Az új technológiák fogyasztói elfogadása. A magyar és nemzetközi szakirodalom áttekintése és kritikai értékelése. Vezetéstudomány/Budapest Management Review, 48(10), 38-47. Kim, S. S., & Malhotra, N. K. (2005): A Longitudinal Model of Continued IS Use: An Integrative View of Four Mechanisms Underlying Post-Adoption Phenomena, Management Science, 51(5), 741-755. https://doi.org/10.1287/mnsc.1040.0326 Lee, Y. – Kozar, K. A. – Larsen, K. (2003): The technology acceptance model: Past, present, and future. Communications of the Association for Information Systems, 12(50), 752–780. Lin, F., Fofanah, S. S., & Liang, D. (2011): Assessing citizen adoption of e-Government initiatives in Gambia: A validation of the technology acceptance model in information systems success. Government Information Quarterly, 28(2), 271-279. https://doi.org/10.1016/j.giq.2010.09.004 Molnár, T., Kő, A., & Mátyus, B. (2017): Exploring Usability and Acceptance Factors of m-Government Systems for Elderly. In International Conference on Electronic Government and the Information Systems Perspective, Springer, 175-188. https://doi.org/10.1007/978-3-319-64248-2_13 Nemeslaki, A., Aranyossy, M., & Sasvári, P. (2016): Could on-line voting boost desire to vote?–Technology acceptance perceptions of young Hungarian citizens. Government Information Quarterly, 33(4), 705-714. https://doi.org/10.1016/j.giq.2016.11.003 Nica, E. (2015): Sustainable Development and citizen-Centric E-government Services, Economics, Management, and Financial Markets, 10, 69-74. Panda, P. – Sahu, G. P. (2013): Critical success factors for e-Gov project: A unified model. IUP Journal of Supply Chain Management, 10(2),19–32. Powell, A. – Williams, C. K. – Bock, D. B., Doellman, T. – Allen, J. (2012): e-Voting intent: A comparison of young and elderly voters. Government Information Quarterly, 29(3), 361–372. https://doi.org/10.1016/j.giq.2012.01.003 Rana, N. P., Dwivedi, Y. K., & Williams, M. D. (2013): Evaluating alternative theoretical models for examining citizen centric adoption of e-government. Transforming Government: People, Process and Policy, 7(1), 27-49. https://doi.org/10.1108/17506161311308151 Rogers, E. M.(2003): Diffusion of Innovations. 5th Edition. New York, NY: Free Press Schaupp, C. L. – Carter, L. (2005): E-voting: from apathy to adoption. Journal of Enterprise Information Management, 18(5), 586–601. https://doi.org/10.1108/17410390510624025 Scott, M. – DeLone, W. H. – Golden, W. (2009): Understanding net benefits: A citizen-based perspective on e-government success. ICIS 2009 Proceedings, Paper 86. Tas, B. K. O. – Genis-Gruber, A. (2008): E-Procurement Savings and the Competition Effect: Analysis of Cultural Differences Through a Unified Model. TOBB University of Economics and Technology, Department of Economics Working Paper, 8–15. Venkatesh, V. – Davis, F. D. (2000): A theoretical extension of the technology acceptance model: Four longitudinal field studies. Management Science, 45(2), 186–204. https://doi.org/10.1287/mnsc.46.2.186.11926 Venkatesh, V. – Morris, M. G. – Davis, G. B. – Davis, F. D. (2003): User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425–478. https://doi.org/10.2307/30036540 Venkatesh, V., Thong, J. Y. L., & Xu, X. (2012): Consumer acceptance and use of information: extending the unified theory of acceptance and use of technology. MIS Quarterly 36(1), 157-178. Ziemba, E. – Papaj, T. – Żelazny, R. (2013): A model of success factors for e-government adoption–the case of Poland. Issues in Information Systems, 14(2), 87–100. Ziemba, E. – Papaj, T. – Jadamus-Hacura, M. (2015): Critical success factors for adopting state and local egovernment–Polish insights. Proceedings of the 13th International Conference on e-Society 2015, International Association for Development of the Information Society, 95–102. Appendix Good State Public Administration Opinion Survey – selected questions, measurements and results of the factor analyses Appendix 2 Studying Complex IT Challenges? Discuss Real Cases JANIS L. GOGAN & DAVID M. MURUNGI 5 CORRESPONDENCE ADDRESS: Janis L. Gogan, DBA, Professor, Bentley University, Information & Process Management, 75 Forest Street, Waltham MA 02452-4705 USA, jgogan@bentley.edu. David M. Murungi, Ph.D., Assistant Professor, Bentley University, Information & Process Management, 75 Forest Street, Waltham MA 02452-4705, USA, dmurungi@bentley.edu. DOI https://doi.org/10.18690/978-961-286-170-4.4 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract Field-based case studies are uniquely suited to exploring the complex challenges of digital transformation. We propose that real discussion cases (developed from rigorously-researched field-based case studies) can trigger useful discourse that helps scholars improve theories addressing complex digital transformation challenges. To advance this argument we undertook an extreme-case study to examine the practices and theoretical contributions of an exceptionally impactful researcher in a non-IT domain (Sumantra Ghoshal, an international management scholar who conducted many field- base case studies and produced many real discussion cases). Based on our extreme-case study findings we consider implications for research on digital transformation. Keywords: • case research • case study • discourse • teaching cases • 1 Introduction Field-based case studies are uniquely suited to exploring complex challenges organizations face in adopting and implementing emerging technologies (Eisenhardt 1989). We propose that real discussion cases, developed from rigorously-researched field- based case studies, can trigger useful scholarly discourse. Because case discussants can interpret real cases richly, holistically, and freely, discussions can help the researcher appreciate new perspectives on study findings, particularly if disscussants vary in their backgrounds, experience, and other dimensions. New perspectives, in turn, may help the researcher identify useful new questions for further study. Thus, when real cases are critiqued through discussions among scholars and with practitioners, scholars should develop ideas that lead to stronger theories. To advance this argument, we present an extreme-case study which examined the work and research practices of an exceptionally impactful scholar in a non-IT domain: Sumantra Ghoshal. As explained in Yin (2018) and Mills et al. (2010) an extreme-case study is an appropriate research method if the objective is to learn from a rare or unusually positive or negative example. In a 20-year academic career (1985-2004) Ghoshal published more than 70 papers, 12 books, and 36 (or more) discussion cases1 before his death at age 55 in 2004. One of the most influential management scholars of the 20th century (Rugman 2002), Ghoshal is best known for conceptualizing the multinational corporation (MNC) as an inter-organizational network confronting the dual challenges of integration and differentiation (first articulated by Lawrence and Lorsch, 1967 as intra-organizational challenges). Later, Ghoshal's papers considered why human resources management practices were not keeping pace with socio-economic and technical forces. He also published much-cited critiques of rational-actor management theories which, he asserted, harm both business education and management practice (Ghoshal 2005). In 2010 the Sumantra Ghoshal Conference was established at London Business School (LBS) in his honor; its annual Ghoshal Award for Research Relevance and Rigour has gone to Kathleen Eisenhardt, Michael Tushman, Ranjay Gulati, David Teece, Robert Sutton, Ron Adner, Laurence Capron, and Amy Edmondson. Thus, the extreme case of Sumantra Ghoshal is that of a scholar who made an exceptionally strong impact on the field of international management, as well as more generally on theories of management, leadership, and organization design and business education. His documented enthusiasm for translating field-based case studies into both scholarly publications and practitioner publications – including real discussion cases -- provides an opportunity for us to consider whether and how case research and discussion cases influence research outputs and practices. In this paper we show that Ghoshal's case research was complemented by vigorous case discussions with co-authors, business leaders, and students. Consistent with ideas advanced by Vermeulen (2007), we contend that case discussions were important to a virtuous cycle of communication that improved both the relevance and rigor of his research. The extreme-case study evidence presented here supports our proposition that real discussion cases can trigger a process of mutually-informing collaboration among scholars and practitioners. In this paper, we first define and discuss foundation concepts: discourse, case research, complex IT challenges. Next, we report on our extreme-case study (of Ghoshal as an extreme case of exemplary double-impact research). After discussing how cases informed Ghoshal's work, we consider implications for research on digital transformation. We conclude with a broader consideration of how rigorously researched real discussion cases contribute to theories addressing complex IT challenges. 2 Foundation Concepts: Discourse, Complex IT Challenges, and Discussion Cases 2.1 Discourse Ideas are socially constructed in discourse among an “invisible college” of scholars (Paisley 1972), within and across disciplinary and geographic boundaries. Impactful research may contribute to the invisible college, to practice, or to both, and scholarly research and teaching can be synergistic: “Theory surely leads to practice, but practice also leads to theory. Teaching, at its best, shapes both research and practice. …” (Boyer 1990, p. 16). Vermuelen (2007) proposes that two communication loops help a scholar produce rigorous and relevant work (Figure 1). Consistent with this view, Tushman and O’Reilly (2007) report that doing case interviews and discussing real cases in executive programs are mutually-reinforcing activities. Thus, a real discussion case can trigger virtuous cycles of reflection and feedback. While Figure 1 emphasizes executive education discussions, we propose a broader view of the Relevance communication loop: case discussion with undergraduate or MBA students, PhD students, colleagues, and/or people in various jobs at various levels in various industries may give a researcher new ideas that trigger new studies in the Rigor communication loop. Figure 1: Loops of Communication (Vermeulen 2007, p. 757) Vermeulen’s rigor-relevance communication loops are supported by discourse theory, which explains how arguments are translated from one domain to another. “Argumentation” -- broadly construed as a way to justify knowledge claims (Habermas 1984; Toulmin 2003) – is categorized in two complementary types: Dialectical and Rhetorical (Hohman 2000; Leff 2002). Dialectical argumentation, corresponding to Vermeulen’s Rigor communication loop, is an abstract and structured form of propositional logic that tests arguments by applying formal rules (Rowland 1987). Rhetorical argumentation, corresponding to Vermeulen's Relevance loop, tests the plausibility of a proposition relative to a particular audience, rather than in relation to alternative propositions (Jacobs 2000). Some theorists view the two forms of argument as antagonistic, since the tendency of dialectical argument is to transcend, while the tendency of rhetorical argument is to situate (Leff 2002). Aristotle saw the two forms of argument as complementary. He argued that rhetoric is the necessary counterpart of dialectic, since rhetoric is needed to defend decisions (you may be right, but you still need to convince others (in Krabbe 2000). Leff (2002) also sees rhetoric and dialectic as complementary, in that dialectic depends on rhetoric to “close and define the situations in which it can operate.” In this view, rhetoric can provide provisional, local closure when conclusive agreements are not reached through inference. However, the addition of dialectical rationality to an argument helps achieve the goal of effective persuasion. In the IS subfield of systems design and development, Peter Checkland's engaged scholarship (action research) resonates in ways similar to Vermeulen's Loops of Communication, particularly in the Inquiring Learning Cycle of SSM (Checkland 1999, p. A9). He states optimistically that “as long as the interaction between the rhetoric and the experienced 'reality' is the subject of conscious and continual reflection, there is a good chance of recognizing and pinning down the learning which has occurred.” Yet, he cautions, “The process of learning by relating experiences to ideas is always both rich and confusing.” (Checkland 1999, p. A7). The field of IS is grateful that Checkland and colleagues did not give up; their ability to rethink classic systems engineering methodology led to the important “distinction between 'hard' and 'soft' systems thinking”: the world may be 'hard' but the essential 'process of inquiry' is the 'soft' and all-important 'learning system' (Checkland, 1999, p A7) 2.2 Complexity and Case Research Complex problems (and especially-complex “wicked” problems) arise in many contexts, including government policymaking (Nickerson and Sanders 2014), corporate strategy (Camillus (2008 and 2016), software development (DeGrace and Stahl 1990), and other domains (e.g., Conklin 2005; Ritchey 2011). A wicked problem is “a social or cultural problem… that is difficult or impossible to solve for as many as four reasons: 1) incomplete or contradictory knowledge, 2) the number of people and opinions involved, 3) the large economic burden, and 4) [being] interconnected … with other problems.” (Kolko 2012) Methods of engaged scholarship, including action research (Checkland 1999) and case research (the topic of this paper) are well suited to the study of complex problems, including emerging-IT challenges. Several recent case studies show the benefit of using case research to study complex emerging-IT challenges:  A recent multiple-case study aimed to understand how “big old” companies achieve digital transformation. That study homed in on two key capabilities: building and continuously improving an organization's “operational backbone” for highly reliable systems and data, and building and refining a flexible and responsive “digital services platform” (Sebastian et al., 2017).  Another multiple-case study (Ma and McGroarty 2017) considered how three disruptive technologies – high- frequency trading, social network analytics and smart mobile applications – change financial markets and introduce important societal implications. Their case findings demonstrate that while innovations led to improvements (e.g., harnessing crowd wisdom, leading traders to produce more accurate price estimates) the increased transaction velocity gave rise to new or exacerbated challenges (e.g., misinformation due to complex information networks, speculative trading behavior, increased market volatility).  An ethnographic case study (Niemimaa and Niemimaa 2017) investigated how universal best-practice prescriptions for information systems security are translated into actual organizational practices. While the literature had shown that best practices should be contextualized, little was known about how organizations actually translate these into situated practice. This case study both illuminates important translational mechanisms and reveals hurdles which an organization faced during this translation process. Case researchers disseminate their findings via three routes: scholarly journals (e.g. EJIS, MISQ), practitioner outlets (e.g., MISQe, Harvard Business Review), and discussion cases (distributed by CaseCentre, HBS Publishing and others). We argue that these varied publication outlets make it possible for case researchers to produce stronger and more useful theories about complex IT challenges, by triggering discourse among scholars, practitioners and future leaders (e.g., MBA students). Real discussion cases are particularly well suited to this process, as we discuss next. 2.3 Discussion Cases An ideal high-quality real discussion case (per Naumes and Naumes 2012) is based on primary data gathered in the field (via interviews and observation, data from an organization’s web site, annual reports, internal memoranda, etc.). The case provides truthful evidence about multiple facets of a focal situation (that is, it is real). It takes the point of view of a particular protagonist at a point in time and (if written well) it has a story-like tone that helps readers empathize with the protagonist. Skilled case discussion facilitators encourage discussants to “stand in the shoes” of this manager at this organization, facing this particular challenge. The problem the manager faces “has not yet been solved, and may even remain to be identified” (Naumes and Naumes, 2012 p. 33). As is true of most complex problems, there is no single “right” answer, and the discussion case does not explicitly state the author's interpretation (thus freeing discussants to offer their interpretations). In discussion, participants usually come to appreciate facets they did not consider during individual case preparation (Andersen and Schiano 2014). These characteristics make the preparation of a real discussion case a useful (maybe optimal) early step in a program of research on a complex emerging-IT problem. Emerging IT challenges involve equivocal technologies: IT appplications for which “information is incomplete, hyperbolic, or highly ambiguous” (Berente et al. 2011). Most emerging IT challenges are equivocal because of the essential flexibility of software; its potential is limited primarily by the imagination and skill of its developers. Similarly, the potential of a flexible IT-enabled platform is limited primarily by the imagination and skills of its end users. Consider the many emergent forms that an Excel spreadsheet or Word document (examples of end-user platforms) can take through use. ERP and other enterprise and inter-enterprise software may be tightly restrictive, but other IT platforms offer great flexibility -- so much so that Cash et al. (2008) observe that every IT organization faces a continual challenge of effectively managing known technologies and applications (with low or manageable equivocality), while exploring a continuous stream of new equivocal technologies (via separate processes, structures and controls that are conducive to experimentation). Relevant individual and collective sensemaking about equivocal technologies is supported by collective discourse, in which varied interpretations of use and value are aired. Reporting on a case study of sensemaking in a virtual world (Second Life), Berente et al. (2011, p. 705) concluded: “From rational argumentation around conceptual capabilities to the metaphorical association with earlier waves of innovation, individuals draw upon a rich tapestry of sensemaking strategies to confront the equivocality that they encounter. ” Having laid out our arguments for the importance of discourse for making sense of equivocal technologies in general, and for how real discussion cases support flexible sensemaking that may improve theory, we next report on the extreme case of an exemplary double-impact researcher: Sumantra Ghoshal. Our findings will reveal how real discussion cases support the dual communication loops of rigor and relevance (or of dialectic and rhetoric). 3 The Extreme Case of Sumantra Ghoshal 3.1 Case Study Overview This study is part of a broader investigation of how exemplary double-impact scholars in several management disciplines use case studies to investigate complex problems. Our investigation began with an exploratory single-case study, in which we examined the extreme case of an influential international management scholar (Sumantra Ghoshal's home discipline was management, not IS management). We drew primarily on archival sources (his papers, discussion cases, and books, as well as presentations, articles and books about him). We also conducted a one-hour telephone interview with his main collaborator, Christopher Bartlett, who co-authored four books and ten Financial Times (FT)50 journal publications with Ghoshal. Born in 1948 in Calcutta, Ghoshal studied physics at Delhi University, and worked for Indian Oil for about a decade, before moving to the United States to pursue graduate studies. His MIT Master’s thesis (1983) asked 32 US managers in Fortune 500 multinational corporations (MNCs) about their external information needs. Similar to Aguilar (1967) Ghoshal reported that competitive and market information were important to these managers; however, respondents’ rankings of other factors differed from prior studies. Ghoshal built on this initial work while earning two doctorates: a PhD from MIT in 1985 and a DBA from Harvard in 1986. His rise was meteoric. His scholarly career began at INSEAD in 1985 and he also was founding dean of IBS- Hyderabad. He joined the LBS faculty in 1994 and remained at LBS until his death in 2004. Our case study time period covers his doctoral studies until his death. Exhibit 1 summarizes Ghoshal's work over his two-decade scholarly career. His Top Ten most highly-cited journal publications for his career (four with Bartlett) yielded more than 38,800 citations. 3.1 Sumantra Ghoshal's Two-Decade Scholarly Career, 1985-1995 Our purpose is to elucidate how Ghoshal successfully blended rigorous case research, story-like discussion cases, and practitioner articles and books, leading to strong and useful new theories of international management. However, this section should not be approached as a literature review, since our purpose is to illustrate the value of conducting rigorous case research, producing discussion cases, and discussing them -- not to fully inform the reader about international management research. Ghoshal's MIT PhD dissertation reference list shows Ghoshal (1985) was influenced by many scholars in the “invisible college” (175 works cited). This was a multi-method study (surveys, cases, non-case interviews) of environmental scanning at both individual and organizational levels of analysis. It aimed to answer two questions: RQ1) What environmental, organizational, and individual attributes affect the way managers scan their business environments? RQ2) How might a firm organize the environmental scanning function? Ghoshal saw organizations as complex systems and noted that organizational complexity is second only to a “transcendent” level of complexity in general systems theory (Boulding, 1956), similar to a wicked problem (Rittell and Webber 1973). To learn how managers allocated time to various informational tasks, Ghoshal surveyed 55 managers at six Korean global trading companies and 56 managers at 10 Korean manufacturers (similar to Mintzberg, 1978 and 1990). To learn why, he interviewed 36 managers in three Korean firms, and visited environmental scanning departments at two firms which each used a scanning methodology that was pioneered by a former Samsung employee. Ghoshal also reported in detail on a Samsung case study that, he stated, led him to believe that organizational capabilities are heavily affected by individual competencies/behaviors. This theme pervades much of his later work. For his Harvard DBA Ghoshal worked under the supervision of Christopher Bartlett, who had already launched what would be a nine-case study in three industries and three countries (US, Europe, Japan): Electronics Telecom Packaged Goods GE ITT Unilever Philips Ericsson P&G Matsushita NEC Kao Bartlett’s project aimed to create teaching cases and a book for a new course, Managing International Business (MIB). “Everything we worked on showed up in that MIB course,” Bartlett told us. The nine cases (of which, the Matsushita case was led by Ghoshal) are the “core” of Managing Across Borders (Bartlett and Ghoshal 1989). Before being made available via HBS Publishing2, they were taught in a short executive education course and the MIB course, and described in three early practitioner publications (Ghoshal & Bartlett 1986; Bartlett & Ghoshal 1987a and 1987b). By 1991, four discussion cases from their work were in the HBSP system: GE (391-144), Komatsu (390-037), Matsushita (388-144), and P&G (384-139). Ghoshal’s DBA thesis (Ghoshal 1986) asked: “How can relations between the headquarters and subsidiaries of a large multinational corporation be organized so as to facilitate innovations in the company?” His study was conducted in three stages: 1) 100 interviews on 38 innovation cases at Ericsson (Sweden), Matsushita (Japan), NEC (US), Philips (Netherlands). 2) Surveys of 300 + managers at Matsushita, NEC, and Philips. 3) Survey sent to 500 N American and European multinationals, with 65 responses. The case studies revealed four innovation processes (center-for-global, local-for-local, local-for-global, global-for-global) and three MNC subsidiary types: Innovator, Contributor, Implementer. Ghoshal (1986) stated: “The organizational factors that facilitate each process are not only different but mutually contradictory. Herein lies a key challenge for MNC managers: designing an organization that can facilitate all three innovation processes simultaneously. … Based on both case and survey research, a framework is developed to suggest a basis for differentiation in allocation of subsidiary roles and structuring of headquarters-subsidiary relations….” In the Acknowledgements section of his 1986 DBA thesis, Ghoshal's praise for his supervisor points to how their discussions helped him form and reform ideas. He stated: “Professor Christopher Bartlett, the chairman of my thesis committee, must share both credit and discredit for this thesis, not only because of his personal involvement with the research ... but also because of his overall influence on the ways in which I have come to think about the task of business administration in general, and about management of multinationals, in particular. … Many of the ideas and concepts presented in this report arose in the course of many, many hours of discussions with him in the lobbies of strange hotels in three continents, in deserted class rooms of Aldrich and Cotting, and in overcrowded airplanes. …” The thesis, centered on the 38 cases in four companies (an embedded-cases study design), launched both his career and Bartlett's. Together, they published 16 case-infused journal publications by 1995. Meanwhile, Ghoshal authored seven other case-infused FT-50 publications and 14 discussion cases (alone or with other co-authors) by 1995. In the second decade of his career Ghoshal produced another five scholarly FT50 papers (none with Bartlett), 11 practitioner papers (seven with Bartlett), 20 real discussion cases (none with Bartlett, who produced 36 other cases during the same time3), and nine books with six sets of co-authors (some with Bartlett, including new editions of books). Ghoshal's view of organizations as complex social systems was strongly linked to his belief in the value of case research, both in terms of the data gathered in each case study and discussions about them with students and colleagues. In a paper published the year after he died, Ghoshal urged business schools to embrace Boyer’s recommendations for pluralistic scholarship4: “We need to … reengage with the scholarship of integration, application and pedagogy to build management theories that are broader and richer than the reductionist and partial theories we have been developing over the last 30 years.” (Ghoshal, 2005, p. 87). In this paper he cites Friedrich von Hayek's 1989 Nobel Memorial Lecture, in which Hayek asserted that a dangerous “pretense of knowledge” is often produced when positivist “scientific” methods are applied to complex social phenomena (such as institutional change). Physical systems, Hayek explained, operate independent of research; the earth is round and gravity exerts its pull regardless of what or how we think about these phenomena. Thus, Hayek asserted, positivist research on a physical phenomenon can change our understanding but does not change the phenomenon. In contrast, when positivist research is applied to a social phenomenon, people who believe the resulting theory may change their actions based on it – in turn, changing the phenomenon. Ghoshal urged management scholars to “temper the pretense of knowledge” through engaged scholarship, including through conducting and discussing rigorous field-based case studies (Ghoshal 2005, p. 87). 3.3 Christopher Bartlett Reflecting on his Research Collaboration with Sumantra Ghoshal A conference presentation or journal publication can trigger a useful discussion about a complex phenomenon, and, we suggest, another valuable way to trigger useful scholarly discussion is to prepare real discussion cases as part of one's program of case research. Our extreme-case study of Ghoshal, an extraordinarily impactful scholar in the field of international management, reveals that case discussion among scholars and with leaders, managers, and business students can yield useful insights about a complex problem, which in turn can improve theory or generate new theory. Christopher Bartlett, who has produced 74 real discussion cases in his career so far (most recently in 2017) also believes cases trigger fruitful conversations that improve theory. In 2007 Bartlett gave a talk at the inaugural Sumantra Ghoshal Conference at London Business School. A summary and commentary about this talk (Rynes 2007) restated Bartlett's view that Ghoshal exemplified several characteristics important to engaged scholarship, including “1) A commitment to field research, built on a profound respect for practitioners. 2) Engagement and ongoing dialog with practitioners … 3) Teaching and research as interdependent activities, with teaching cases being the first output of field research.” We conducted a telephone interview with Bartlett, in fall 2017. Asked to reflect on his collaboration with Ghoshal, he stated: “Case research, teaching, and publications are tightly bound and closely intertwined. I’ve always thought of a three-legged stool: academically engaging plus managerially engaging plus pedagogically engaging. All three are always an explicit part of the case research engagement.” Bartlett’s first publications with Ghoshal were “driven by both our three-industry research design and the ability to come into the MBA and executive classrooms with cases. Also, we were doing a lot of outside consulting activity. ...We were not, at first, led by a strong conceptual model in this research…” Bartlett described their collaboration: “We beat each other up in discussions; pushing ideas back and forth. Out of these cases came a belief that a very different managerial model was developing at companies.” Discussions with students and with each other led Bartlett and Ghoshal to ideas discussed in their books and many of their journal publications. Bartlett stated, “I learned a huge amount from my MBA students. Discussing a case was a way to test ideas; students would push back. Executive Education participants would push back especially hard; they were quite a reality check on the ideas. ... Sumantra and I could really push each other to go far deeper. ... It was intense ... yet we had such an easy rapport.” Their work practice was punctuated by energetic conversations: “I would take the lead on managerial articles; Sumantra would take the lead on the academic papers. Once a first draft was written, we would switch; I’d work on his, he’d work on mine. We beat each other up in the process, butting heads until we agreed we had a compelling story to tell. … It was a pretty constant ongoing collaboration.” All 20 papers that Bartlett published in the FT50 during his career were co-authored with Ghoshal (in addition, Ghoshal produced 16 other FT50 papers, alone or with others). 4 Discussion and Implications for Research on Complex IT Challenges 4.1 Main Findings from the Extreme Case of Sumantra Ghoshal Our extreme-case study of one scholar's work practices and influence illustrates how rigorous case research – and, importantly, real discussion cases produced from this research-- contribute to understanding complex challenges and to helping researchers identify plausible solutions and build better theories. Ghoshal and Bartlett produced many cases about international management, and discussing them was essential to developing their influential theories. We also observe that Ghoshal's extensive reference lists point to his respect for the “invisible college” (a form of asynchronous scholarly discourse). Cases were central to his work, but not the only tool in his kit. His respect for other research methods is evident in his other qualitative studies (questioning managers in similar jobs in many organizations) and quantitative studies (surveys). Ghoshal's engaged scholarship helped him deeply understand the messy complexity of organizational life and to avoid Hayek's “pretense of knowledge.” Both his papers and discussion cases triggered important conversations about management and (later) management education (in his later years Ghoshal criticized managerial theories and teaching approaches that, he felt, over- emphasize analytic techniques and a rational-actor perspective, while underemphasizing behavioral and social aspects of organizational life). Evidence of his impact on theories that address complex challenges of global management is incontrovertible (nearly 40,000 citations to his Top Ten FT50 papers; see Appendix 1), as is the evidence (presented here) of how case research contributed to his (and Bartlett’s) thinking. Ghoshal's body of work, combined with Bartlett's description of their work practices, point to the potential that real discussion cases offer for helping managers and scholars avoid faddish discourse –in any managerial domain, including IS management. Based on the extreme case of impactful research described here, we propose that a program of research on a complex IT phenomenon should include some field-based case studies, from which it is helpful to produce some real discussion cases. 4.2 Case Studies, Case Discussions and Discourse Case research demands cognitive flexibility. When evidence pointed in new directions, and/or when other scholars and practitioners offered alternative interpretations, Ghoshal looked in those directions. For example, an early paper (Ghoshal 1987) explained that strategic competencies and organizational structures for international operations vary, yielding three enterprise archetypes: multinational, global, international. Later, Ghoshal proposed that effective leaders of transnational enterprises master multinational, global, and international strategic competencies and structures (all of these, not just one). Importantly, this insight arose from discussing cases with his students and his main collaborator. Bartlett and Ghoshal had many vigorous and productive conversations, with each other and with MBA and executive students and other scholars. Case discussions often reveal new facets of a phenomenon. Discussants are encouraged to identify with the case protagonist and to recognize that there are usually several right answers (and also many wrong answers) to a complex situation. To some extent, discussants are free to focus on aspects that interest them and to argue why these matter. Conventional scholarly discussions (such as those triggered at conferences) can also produce helpful insights, but if these discussions are confined to small circles of interested scholars, diverse voices may not be mixed into the stew of ideas they stir. Discussing a case with students or practitioners can help a scholar see other facets. Thus, real discussion cases can play a useful role in triggering conversations that scholars have with other scholars and with business or public leaders, managers, and students. Addressing the problem of IT fads (unreasonable expectations and irrationally skewed perceptions of IT innovations), Hirschheim et al. (2012) argue that while scholarly analysis can validate reasonable claims about an innovation, the analysis can also help prevent detrimental “flighty” claims from taking off. However, lengthy peer review processes delay publications, which in turn can delay the emergence of scholarly discussions about new IT phenomena. How to get scholars involved sufficiently early to influence the trajectory of a management fad? Hirscheim et al.'s answer: don't wait for the final journal publication; improve discourse by making fine-grained scholarly analysis available at an earlier stage in a program of research. That argument assumes that ideas flow from one scholar to another as we read each other's works. However, we caution that an important earlier study based on discourse analysis reported that only about 5% of papers that cited “Power, Politics and MIS Implementation (Markus 1984) had engaged with that study's central tenets. In its Conclusion section this paper glumly asked: “Does the IS discipline support a true and active dialogue around concrete research questions about phenomena of interest? Or are we just ships passing in the night, using selected referential spotlights to see the shape and direction of other vessels?” (Hansen et al. 2006, p. 419). We argue that real discussion cases might more quickly help scholars and practitioners jointly evaluate complex IT phenomena -- such as cognitive computing initiatives, digital transformation tactics, or rapidly changing cyber-attack practices and related incident response practices. For fast-moving topics like these, scholarly theorizing can greatly benefit when scholars and practitioners talk with each other about practices in the field. To study complex IT phenomena, curious scholars with complementary expertise and interests, who are skilled in various methods and theories, may produce very helpful case studies. Subsequently, when smart people from multiple theoretical and methodological perspectives argue vigorously about the cases, good ideas should emerge. Sumantra Ghoshal greatly valued his collaboration with Chris Bartlett, and Bartlett greatly appreciated how “butting heads” together produced great ideas. 4.3 Study Limitations and Suggestions for Further Research Ghoshal’s scholarly journey leads us to believe that, for a program of research about a complex IT challenge, an optimal starting point is to conduct one or more rigorous field-based case studies and to produce a few real discussion cases from these. However, our research has limitations. We conducted a single extreme-case study, whereas a more comprehensive embedded- cases study could use content analysis to trace specific Ghoshal ideas through his discussion cases, practitioner articles and scholarly papers (each stream of Ghoshal research would be treated as an embedded case). We have not yet deeply considered rival explanations. For example, could Ghoshal's strong impact be a function of his superior intelligence or superior ability to persuade, independent of his choice to conduct case research and his enthusiasm for discussion cases? Other rival explanations may emerge from further studies of how other exemplary scholars' ideas propagate and develop - such as by considering Kathleen Eisenhardt (first recipient of the Ghoshal Award), other recipients of that award, and various influential IS case researchers. Thus far, no IS researcher has received the Ghoshal award, so to identify exemplary dobule- impact scholars to study, it will be necessary to apply similar criteria as those which are used by the Ghoshal Award committee to identify influential researchers in other domains who exemplify both rigor and relevance. By studying other influential scholars we will have an opportunity to learn to what extent conversations about real discussion cases help shape theory, compared with other forms of individual and collaborative sensemaking. One could also choose to study the contrasting case: research practices of influential management and IS scholars who never conduct case research. By examining how these scholars engage in individual and collaborative sensemaking, compared with case researchers, we might be able to home in on the unique contribution played by real discussion cases and also identify alternative mechanisms that support discourse which improves theory. We acknowledge that many influential scholars (and many good-enough scholars) in the field of IS are neither case teachers nor case researchers. Surely other forms of discourse contribute to how scholarly ideas are aired and shaped and considered in terms of their potential to extend or build management theory or IT management theory. Another limitation arises from our choice to focus on an extreme case, from which we cannot easily generalize -- Ghoshal set a nearly-unattainable standard of research productivity and excellence, whereas most scholars produce far fewer papers and books, and with modest impacts. We also recognize that case research skill is honed with practice and tempered by the researcher's deep knowledge of his/her field and related business fields. Discussion leadership skill is also honed through practice and tempered by broad knowledge. So, a scholar who seeks to emulate Ghoshal's approach to research first needs to read deeply and widely, and to develop expertise in carrying out high quality field case research, and to become skilled at writing real discussion cases, and become skilled at facilitating case discussions. It may be useful to study the scholarly journeys of good (not great) case researchers, to learn whether and how good-enough case research produces ideas that benefit theory and practice, and whether extensive discussion involving participants from multiple occupations and representing mulitple perspectives is fruitful in this context. Prior studies that employed discourse analysis in social media and other IT-enabled contexts suggest opportunities for future research that would explore whether and how case discussions trigger theoretically-important ideas. For example, a three-case Wikipedia study of editing and written discussion on three topics -- Armenian Genocide, Ethanol Fuel, Intellectual Property - analyzed the findings in light of the Theory of Rational Discourse (Habermas 1984) and concluded that it is possible to design an information system that supports “the emancipatory objectives of critical social theory ... [by circumventing] much of the influence of relations of power and domination.” (Hansen et al. 2009). Examining how “software development team members achieve a level of understanding that allows them to work in parallel yet create interdependent components or modules that work together seamlessly” another study (Hansen and Rennecker 2010) sought to understand how hermeneutic interpretation takes shape through collaboration. Further building on prior work by Weick (1995) and Weick et al. (2005), a third study proposed that “individuals make sense of new information technologies through discourse.” It concluded: “In making sense of innovations, individuals present, negotiate and argue for a range of perspectives on the value of the emergent techology ... in a sensemaking process ... ultimately influencing the adoption and evolution of a technology.” (Berente et al. 2011, p. 686). Just as the three studies described above captured peoples' arguments about new IT or IS work, an important next step in our program of research could be to conduct a study which analyzes transcripts of one or more recorded discussions (by MBA and/or executive participants) of real cases that address complex IT challenges. That study could examine the arguments advanced in discussions, in light of each argument's potential for extending or building relevant theory. Once this first step is taken (to verify that theoretically-important ideas are aired and shaped in case discussions), a next step would be to conduct further studies to investigate whether, how, and to what extent ideas aired in case discussions actually do inform theorizing, and whether, how and to what extent the theorizing actually informs influential research on complex IT challenges. Based on one extreme-case study of an extraordinary scholar in the field of international management, we have argued for how discussion of rigorously-researched real cases may help IS scholars propose and improve theories addressing complex digital transformation challenges. We hope we have identified a way for the field of IS to avoid the fate of “ships passing in the night.” Acknowledgments We gratefully acknowledge Chris Bartlett, for sharing his thoughts about case research, discussion cases, and Sumantra Ghoshal. References This list includes those Ghoshal papers and books that are cited in the text of this paper, whereas Appendix 1 lists all of his papers and books. Allen, T.J. (1966). Managing the Flow of Scientific and Technological Information. MIT PhD Dissertation. Andersen, E., Schiano, B. (2014). Teaching with Cases: A Practical Guide. Harvard Business School Press, 1495-ENG. Bartlett, C.A. (1979). Multinational Structural Evolution: The Changing Decision Environment in International Divisions. Doctoral thesis, Harvard University Graduate School of Business Administration. Bartlett, C.A., Ghoshal, S. (1987a). Managing Across Borders: New Strategic Requirements. MIT Sloan Management Review 28, no. 4 (summer): 7–17 Bartlett, C.A., Ghoshal, S..(1987b). Managing Across Borders: New Organizational Responses. MIT Sloan Management Review 29, no. 1 (Fall): 43–54. Bartlett, C.A., Ghoshal, S. (1989). Managing Across Borders. Harvard Business School Press. Berente N., Hansen S., Pike J.C., Bateman P.J. (2011). Arguing the Value of Virtual Worlds: Patterns of Discursive Sensemaking of an Innovative Technology. MIS Quarterly 35(3): 685-709, September. Boulding, K.E. (1956). General Systems Theory: The Skeleton of Science. Management Science 2(3): 197-208, April. Boyer, E.L. (1990). Scholarship Reconsidered: Priorities of the Professoriate. The Carnegie Foundation for the Advancement of Teaching. Burt, R.S. (2004). Structural Holes and Good Ideas. American Journal of Sociology 110(2): Sept. Camillus, J.C. (2008). Strategy as a Wicked Problem. Harvard Business Review 86: 98-101. Camillus, J.C. (2016). Wicked Strategies: How Companies Conquer Complexity and Confound Competitors. U of Toronto Press. Cash, J.I., Earl, M.J., Morison R. (2008). Teaming Up to Crack Innovation and Enterprise Integration Harvard Business Review November. Checkland , P. (1999). Systems Thinking, Systems Practice: A 30-Year Retrospective. Wiley. Churchman, C.W. (1967). Wicked Problems. Management Science 14(4): Dec. doi:10.1287/mnsc.14.4.B141 Conklin, J. (2005). Dialogue Mapping: Building Shared Understanding of Wicked Problems. Wiley, November, ISBN 978-0-470-01768-5. Daft, R.L., Weick, K.E. (1984). Toward a Model of Organizations as Interpretation Systems. Academy of Management Review 9. DeGrace, P., Stahl, H.L. (1990)/ Wicked Problems, Righteous Solutions: A Catalog of Modern Engineering Paradigms. Yourdon Press. Eisenhardt, K.M. (1989). Building theories from case study research. Academy of Management Review 14(4): 532-550. Ghoshal, S. (1985). Environmental Scanning: An Individual and Organizational Level of Analysis. PhD Dissertation, MIT. Ghoshal, S. (1986). The Innovative Multinational: A differentiated Network of Organizational Roles and Management Processes. DBA Thesis, Harvard Graduate School of Administration. Ghoshal, S. (1987). Global Strategy: An Organizing Framework. Strategic Management Journal 8(5): 425-440, Sep-Oct. Ghoshal, S. et al. (1999). A New Manifesto for Management. MIT Sloan Management Review 40(3): 9-20, Spring. Ghoshal, S. (2005). Bad Management Theories are Destroying Good Management Practices. AoM Learning & Education 4(1): 75-91. Ghoshal, S., Bartlett, C.A. (1986). Tap your Subsidiaries for Global Reach. Harvard Business Review, Nov-Dec. Habermas, J., Habermas, J. (1984). The Theory of Communicative Action. Beacon Press. Hansen S, Berente N, Lyytinen K (2009). Wikipedia, Critical Social Theory and the Possibility of Rational Discourse. The Information Society 25(1): 38-59. Hansen, S., Lyytinen, K., Markus, M.L. (2006). The Legacy of 'Power and Politics' in Disciplinary Discourse: A Citation Analysis. ICIS 2006 Proceedings: 407-424. Association for Information Systems. Hansen S., Rennecker J. (2010). Getting on the Same Page: Collective Hermenutics in a Systems Development Team. Information and Organization 20: 44-63. Hirschheim, R., Murungi, D. M., and Peña, S. (2012). Witty Invention or Dubious Fad? Using Argument Mapping to Examine the Contours of Management Fashion, Information and Organization 22(1): 60-84. Hohmann, H. (2000). Rhetoric and Dialectic: Some Historical and Legal Perspectives, Argumentation 14(3): 223-234. Jacobs, S. (2000). Rhetoric and Dialectic from the Standpoint of Normative Pragmatics, Argumentation 14(3): 261-286. Krabbe, E.C. (2000). Meeting in the House of Callias: Rhetoric and Dialectic, Argumentation 14(3): 205-217. Kolko, J. (2012) Wicked Problems: Problems Worth Solving, Stanford Social Innovation Review, Mar. Lawrence P.R, Lorsch J.W. (1967). Organization and Environment. Harvard University Press. Leff, M. (2002). Relation between Dialectic and Rhetoric in Classical and Modern Perspective, in Dialectic and Rhetoric. Springer, 53-63. Ma, T., McGroarty, F. (2017). Social Machines: How Recent Technological Advances Have Aided Financialisation, Journal of Information Technology 32(3): 234-250. Markus, M.L. (1983). Power, Politics and MIS Implementation. Communications of the ACM 26(6): 430-444, June. Mills A.J., Durepos G., Wiebe E. (2010) Encyclopedia of Case Research.Sage Mintzberg, H. (1978). Patterns in Strategy Formulation, Management Science 24. Mintzberg, H. (1990). The Manager's job: Folklore and fact. Harvard Business Review March-April. Naumes W., Naumes M.J. (2012). The Art and Craft of Case Writing (3rd Edition). ME Sharpe. Nickerson J., Sanders R. (2014). Tackling Wicked Government Problems. Brookings Institution Press, Oct. ISBN13: 978-0815725077 Niemimaa, E., Niemimaa, M. (2017). Information Systems Security Policy Implementation in Practice: From Best Practices to Situated Practices. European Journal of Information Systems 26(1): 1-20. Paisley, W. (1972). The Role of Invisible Colleges in Scientific Information Transfer, Educational Researcher 1(4): 5ff, April. Ritchey, T. (2011). Wicked Problems - Social Messes: Decision Support Modelling with Morphological Analysis. Springer. Rittel, H.W.J., Webber, M. (1973). Dilemmas in a General Theory of Planning, Policy Sciences. 4, 155–169. Rowland, R.C. (1987). On Defining Argument, Philosophy & Rhetoric, pp. 140-159. Rugman, A.M. (2002). The Influence of Managing Across Borders on the Field of International Management, Advances in Comparative International Management 14: 37-56, 2002. Rynes, S. (2007). Editor’s Foreword: Carrying Sumantra Ghoshal’s Torch: Creating More Positive, Relevant, and Ecologically Valid Research, Academy of Management Journal 50(4): 745-747. Sebastian, I.M., Ross, J.W., Beath, C., Mocker, M., Moloney, K.G., Fonstad, N.O. (2017). How Big Old Companies Navigate Digital Transformation. MIS Quarterly Executive 16(3), September. Toulmin, S. (2003). The Uses of Argument (Updated Edition; first publication 1958). Cambridge: Cambridge University Press. Tushman, M., O’Reilly, C. (2007). Research and Relevance: Implications of Pasteur’s Quadrant for Doctoral Programs and Faculty Development. Academy of Management Journal 50(4): 769-774. Vermeulen, F. (2007). I Shall not Remain Insignificant: Adding a Second Loop to Matter More. Academy of Management Journal 50(4): 754-761. Weick, K. (1995). Sensemaking in Organizations. Thousand Oaks, CA: Sage Publications. Weick, K., Sutcliffe, K.M., Obstfeld, D. (2005). Organizing and the Process of Sensemaking. Organization Science 16(4): 409-421. Yin, R.K. (2018). Case Study Research and Applications: Design and Methods. Sage. Appendix 1: Sumantra Ghoshal Books (later editions noted) 2006 A Bias for Action (Bruch & Ghoshal) 2005 Organization Theory and the Multinational Corporation (Westney and Ghoshal) 2004 The Future of the Multinational Company (Birkenshaw, Ghoshal, Markides, Stopford, Yip) 2002 Managing Radical Change: What Indian Companies Must Do … (Ghoshal, Piramal, Bartlett) 2002 World Class in India: A Casebook of Companies in Transformation (Ghoshal, Piramal, Budhiraja) 2000 Transnational Management: Text, Cases and Readings (co-author with CA Bartlett on first 5 editions) 1999 Managing Across Borders: The Transnational Solution (2 editions, co-authored with CA Bartlett) 1997 The Individualized Corporation (Ghoshal and Bartlettl) 1987 The Strategy Process: Concepts, Contexts, Cases (Mintzberg, Lampel, Quinn, Ghoshal). 1997 The Differentiated Network (Nohria, Ghoshal) Career Top Ten Journal Publications, per Google Scholar (last two considered a tie for 10th) 16,862 1998 AoM Review Nahapiet, Ghoshal. Social capital, intellectual capital, and organizational advantage. 6,347 1998 AoM Journal Tsai, Ghoshal. Social capital and value creation: The role of intrafirm networks 3,562 2005 AoM L&E Ghoshal. Bad management theories are destroying good management practices 2,970 1996 AoM Review Ghoshal, Moran. Bad for practice: A critique of the transaction cost theory 2,474 1990 * AoM Review Ghoshal, Bartlett. The multinational corporation as an interorganizational network 1,745 1987 SMJ Ghoshal. Global strategy: An organizing framework 1,100 1986 * HBR Ghoshal, Bartlett. Tap your subsidiaries for global reach 1,127 1989 SMJ Ghoshal, Nohria. Internal differentiation within multinational corporations 971 1988 * JIBS Ghoshal, Bartlett. Creation, adoption and diffusion of innovations by subsidiaries 838 1994 SMJ Nohria, Ghoshal. Differentiated fit and shared values: Alternatives for mananging HQ-subsidiary relations 826 1993 * SMJ Bartlett, Ghoshal. Beyond the M-form: Toward a managerial theory of the firm 38,802 citations 1 AoM Journal, 3 AoM Review, 1 JIBS, 1 HBR, 2 SMJ 4 co-authored w CA Bartlett 37 Papers in the FT 50 journals (20 papers co-authored with C Bartlett indicated with *) Academy of Management Review 1999 1996 1990 California Management Review 1998* 1997* 1992 1988 Harvard Business Review 2004 2003* 2000* 1995* 1995* 1994* 1992* 1990* 1986* Journal of International Business Studies 1998 1988* Journal of Operations Management 2006 Management Science 1994 Strategic Management Journal 2000 1994 1994* 1993* 1991 1991* 1987 MIT Sloan Management Review 2005 2003 2003 2002* 1999* 1996* 1995* 1993 1987a* 1987b* 36 Discussion Cases (in CaseCentre and HBS Publishing Collections, Feb 2017) 396-142 Lufthansa: The challenge of globalization. 902-020 Ogilvyone: Integrating the enterprise. 389-025 Scandinavian Airlines Systems (SAS) in 1988. 390-037 Electrolux: Acquisition & integration of Zanussi. 389-021 The Dunlop-Pirelli Union 393-007 Cartier: A legend of luxury 394-140 The transformation of AT&T. 396-028 Felix Constructions SA. 392-031 Canon: Competing on capabilities. 396-050 SA Chupa Chups 396-138 Philips Semiconductors: Breaking with the past. 396-154 The LG Group: Leaping to the Future. 392-055 Andersen Consulting (Europe): ….PRIZE WINNER 302-124 The transformation of Bajaj Auto. 394-019 ISS – International Service System A/S. 399-088 Housing Development Finance Corp. (HDFC) 396-153 Indian Oxygen Ltd: Transformation in India. 302-035 Nicholas Piramal LTD: Integrating diversity. 394-051 General Electric synopsis. 397-065 WIPRO Corporation: Balancing the future. 388-144 * Matsushita Electric Industrial (MEI) in 1987. 396-048 Reliance Industries Ltd 301-040 Lufthansa 2000: Maintaining the change momentum. 302-034 Sun Microsystems: Driving innovation … 392-049 Kao Corporation PRIZE WINNER 396-140 Revitalisation of the Bank of Montreal. 302-199 Hindustan Lever Limited: Levers for change. 399-054 VIP Luggage: “It takes a lot to be a VIP” 396-141 British Telecommunications PLC … 305-149 Emirates Airline: Hub of the world. 396-139 Standard Chartered Bank. 302-090 Natura: The magic behind Brazil’s …. 399-053 Siemens Nixdorf IS. 392-033 The transformation of BP Business Model Innovation and Firm Performance: The Role of Mediation and Moderation Factors MOHAMMAD-ALI LATIFI & HARRY BOUWMAN 6 CORRESPONDENCE ADDRESS: Mohammad-Ali Latifi, Ph.D. Candidate, Delft University of Technology, Technology, Policy and Mangement Faculty, Jaffalaan 5, Delft, The Netherlands, e-mail: s.m.a.latifirostami@tudelft.nl. Harry Bouwman, Professor, Delft University of Technology, Technology, Policy and Mangement Faculty, Jaffalaan 5, Delft, The Netherlands, e-mail: W.A.G.A.Bouwman@tudelft.nl DOI https://doi.org/10.18690/978-961-286-170-4.5 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract Business model (BM) innovation is vital for today’s businesses. However, BM innovations can be irreversible, and therefore, in comparison to product, service or process innovation, entail bigger risk and ambiguity. Understanding the way in which BM innovation exerts influences over firm’s performance would help business-owners to be more effective. Based on a systematic literature review, a model to examine how BMI impacts firm performance through mediating and moderating factors was developed. Based on in-depth analysis of 37 articles, we identified twenty moderating factors classified in four groups, i.e., Firm-Characteristics, Industry-Characteristics, BM Implementation, and BM Practices, and ten mediating factors, categorized in three sub-groups, i.e., Revenue Growth, Efficiency Growth, and Enhancing the Organizational Capabilities. This paper offers the grounding for empirical research as well contribute to the development of tools to assess the effectiveness of the BMI. Keywords: • Business Model Innovation • Performance • Mediation • Moderation • 1 Introduction Business is subject to rapid change in technology, regulation, and customers and competitors behaviors. To sustain continued growth, to become more profitable or simply to survive, firms have to adapt their business logic. Since the advent of the Internet, the notion of Business Model (BM) and Business Model Innovation (BMI) gained a lot of attention in industry and academia. BMs describe the logic of how a company creates, delivers and captures value (Teece, 2010). As a means of renewal and adaptation to a changing market (Hartmann, Oriani, & Bateman, 2013), Business Model innovation has the potential to create competitive advantage (Magretta, 2002), can enhance value creation (Teece, 2010) and opportunity recognition (Guo, Tang, Su, & Katz, 2017) to exploit a market niche not addressed by competitors (C. Zott & Amit, 2007). Since performance improvement is at the heart of any firms, the contribution of BMI on performance has attracted more and more attention (Hartmann et al., 2013; Lambert & Davidson, 2013; Karimi & Walter, 2016). Although these studies have enriched knowledge on the link between BMI and performance, this line of research is still at an infant stage (Zott and Amit, 2013). Unlike innovation in product, service, and process, BMI requires fundamental changes in core components of the organization’s BM (Nair, Paulose, Palacios, & Tafur, 2013). BM Innovation (BMI) is therefore associated with high risk and uncertainty also due to the possible turmoil and resistance in the organization (Yannopoulos, 2013). The best-formulated BM may fail to lead to improved performance if not handled properly (Chesbrough, 2010; Knab & Rohrbeck, 2014). For instance, Christensen, Bartman, and Van Bever (2016) identified that more than 60% of BMI efforts did not deliver the expected improved performance. BMI does not automatically trigger impressive performance gains. So far, most academic studies looked primarily at types and components of BMs and proposed a large number of distinct explanations of BM performance (Haggège, Gauthier, & Rüling, 2017). To date, we have no understanding of cause-effect relationships and mutual dependencies in the linkage between the BMI and firm performance (Methlie & Pedersen, 2008). BMI scholars, recently, have called for causal analyses of antecedences and effects of BM, for large-scale investigations and application of advanced methodologies (Zott et al., 2011; Spieth et al., 2014; Clauss, 2016; Methlie & Pedersen, 2008). With the exception of some qualitative studies, there is little empirical research examining factors which influence the success of BMI initiatives in firms (Patrick Spieth et al., 2016). In the literature, the causal relation between BMI activities and firm higher performance remains unclear (Sebastian Knab, René Rohrbeck, 2014). In this study, we are not focused on the question of whether BMI has performance implications. This study attempts to address deeper questions about (1) ‘WHY’ and ‘HOW’ BMI affects performance, i.e. researchers simply directly relate BMI to performance without understanding what is in between, and (2), ‘WHEN’ and under which conditions, the relationship between BMI and firm performance can be strengthened. The first question leads to the need to understand mediators, and the second to understand moderators factors. Therefore, our objective is: to develop a conceptual framework that explains the complex mechanisms through which BMI influences firm performance. We contribute to existing Business Model literature in two ways. First, the proposed model considers mediating and moderating effects elucidate how managers can ensure that BMI brings more benefits for the firms in terms of performance, and provides a conceptual foundation for practice and future research. Second, it helps practitioners to align their BMI efforts with firm’s strategy and daily operational activities by providing a big more concrete picture of what is going on in the firm. This paper is structured as follows: First, the literature on mediating factors between BMI and performance of the firm is shortly presented, followed by a discussion of our empirical constructs. Our research method is then described. Next, the conceptual model is presented. Finally, we draw conclusions, discuss limitations and come up with some suggestion for future research. 2 Literature Review Business Model Innovation. A business model, which serves as a tool to analyze and communicate strategic choices (Lambert & Davidson, 2013; Shafer, Smith, & Linder, 2005), is seen as a realized expression of strategy (Casadesus-Masanell & Ricart, 2010; Dahan, Doh, Oetzel, & Yaziji, 2010) and articulates how available resources can be used effectively, how costs can be reduced, and how new sources of revenues can be leveraged (Chesbrough, 2007). According to Bock, Opsahl, George, and Gann, (2012), firms striving for a long-term performance need to innovate their BM. Moreover, the potential of technologies can often only be realized by using a new BM (Chesbrough & Rosenbloom, 2002). At a fundamental level, scholars and practitioners agree that the BM is vital to growth (Teece, 2010; Terrenghi, Schwarz, Legner, & Eisert, 2017), gain competitive advantage (Mitchell, 2003; Afuah, 2000), enhance long-term performance (Bock et al., 2012), and enable further innovation (Zott, 2011). Through BMI, a firm may be able to exploit a new market which is not addressed by its competitors and open up a niche market (Hartmann et al., 2013). However, BMI is a highly complicated and risky process with an uncertain outcome (Waldner, Poetz, Grimpe, & Eurich, 2015; Marc, Sosna; Rosa Nelly, Trevinyo-Rodriguez; Ramakrishna, 2010; Chesbrough, 2010), because it necessitates experimentation (McGrath, 2010), managing conflicts, interaction with people on different level of expertise and modifying the ongoing process of BM (Latifi & Bouwman, 2017) and asks for specific leadership style (Smith, Binns, and Tushman, 2010). Hence, knowing how and when to innovate a BM is a serious challenge (Hartmann et al., 2013). Firm Performance. Performance has been at the core of management thinking (Haggège et al., 2017), as performance directly affects the continuation of the firm, it became an essential concept in management research (Rauter, Jonker, & Baumgartner, 2017). Venkatraman and Ramanujam (1986) have pointed out that firm performance is a multi-dimensional construct. They proposed three general levels of firm performance, i.e. financial performance, business performance and organisational effectiveness, each which their own indicators like for example return on assets (ROA) (Parker, 2000), growth, market share, diversification, and product development (Gray, 1997), and employees satisfaction, quality, and social responsibility (Rauter et al., 2017). Initial research model. Building on the short review and following Foss and Saebi (2016), we consider performance, innovativeness, as outcomes. Moderators can be studied on a macro, or on a micro/firm level as depicted in figure 1. This model is elaborated upon based on a systematic literature review. Moderators Macro-Level Firm-Level Micro LevelBMIOutcomes Figure 1: Initial Research Model adapted from Foss & Saebi (2016). 3 Research Method A systematic literature review is based on replicable, scientific, and transparent process (Cook et al.,1997). Using such an approach researcher's potential bias will become explicit, effects of chance decrease, and the validity of data analysis will enhance (Reim, Panda, & Ortqvista, 2015). As mentioned, we were looking for finding out under which moderating and mediating mechanism BMIs has an effect on performance. Next, we made use of academic databases such as Web of Science, ABI/INFORMS, Science Direct, and Wiley Online Library. ‘Business model,’ ‘mediating,’ ‘moderating,’ and ‘performance’ were used as keywords in search of databases. We did not set any limitations on papers’ publication date or types of documents. So we included journal papers, conference articles, working papers as well as book chapters. This search yielded 115 publications considered to be relevant based on titles, abstract and keywords. Omitting duplications, these produced 97 unique citations. In the third step, based on the abstracts, collected publications were screened for their match and correspondence with our research objective. We include papers based on the following criteria:  Are explicitly and empirically tested hypotheses about the relation between BMI and Business Performance by using a quantitative, empirical, analytical approach being proposed.  Is reference made to BMI as a way to change main components of the BM by introducing a new system of value creation, value proposition and value capture?  Refers to Business Performance, as discussed before  Refers to the firm as a unit of analysis. Based on criteria mentioned above, thirty-five papers found to be irrelevant at this stage. Only reviews reporting relevant outcomes (remaining fifty-two publications) were reviewed to find out that articles met the above-specified inclusion criteria. Through in-depth review, twenty-seven papers were identified as relevant. Furthermore, references of identified articles were used as a secondary source of literature analysis. This resulted in identifying ten additional papers. These papers were included in our sample. As a result, our systematic literature review was built on thirty-seven articles. As said before, many research has recently revealed that there is a lack of knowledge in the causal relationship between BMI and firm performance and asked to consider mediation and moderation effects between these two. Therefore, we expected to find few numbers of studies, hence did not set any limitation on publication years and paper types. Figure 2 has been shown that the topic has received attention in recent years, and approximately 76% of our 37 selected articles were published between 2012 and 2017. Moreover, 33 out of them were journaled papers (89%), three conference papers (8%), and one working paper (3%). 0 1 2 3 4 5 6 7 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Figure 2: Number of selected papers in the final analysis (Yearly) In describing, classifying and analyzing the selected papers, we employed a coding approach classifying mediator, moderator, and control variables. All key points were listed on a coding sheet (Dey, 1993) and classified into new categories (Burnard, 1991). After this open coding, the categorized list was jointed under higher level classifications. To limit the number of categories, data were grouped (Dey, 1993). 4 Results 4.1 Moderation Effects Different factors introduced in the literature as contingency and moderating variables which can effects relationship between BMI and fir performance. Some factors were not linked directly to the firm but are related to its environment such as industry sectors. Other factors were associated with firm’s attributes, for instance, its size and age. We could find a great amount of identified factors were related to how the BMI was implemented within the firm, for example, the employees’ skills and commitment can strengthen the relationship between BMI and performance. There was another group of factors were mostly related to BMI practices, such as the speed, scope, and novelty of BMI which can influence the relationship. Therefore based on abovementioned criteria, we categorized identified moderating factors into four groups; BMI-Implementation, BMI-Practices, Firm-Characteristics, and Industry-Characteristics. BMI-Implementation. In our previous study on why BMI fail to deliver expected outcomes (Latifi & Bouwman, 2017), we revealed that about 60% of identified barriers to accomplish BMI objectives and attain expected performance lies in the implementation stage of BM. Although managers mostly concentrate and spend a lot of time and energy on (re-)designing a viable BM, the major challenges can be found in the implementation of a BM, as BMI implies fundamental (Nair et al., 2013). Various studies mentioned “poor implementation” as one of the main reasons of BM failure (Osterwalder, 2012; Batocchio, Ghezzi, & Rangone, 2016; Chesbrough, 2010; Yannopoulos, 2013). Therefore the more skilfully and knowledgeably firm implement their BMI, the more performance enhancement will occur. Many studies revealed that key source of BM failure is connected to its management (Batocchio et al., 2016; Zott, Amit, & Massa, 2010; Chesbrough, 2010). Moreover, Martins, Gerasymenko, De Clercq, and Saprienza (2014) affirmed a positive relationship between BM change experience of a firm’s CEO and performance. People in a company are key to success or failure of a BM innovation program (Hittmár, Varmus, & Lendel, 2014). Depending on the degree of changes in the BM, not only some employees’ training is required to develop relevant capabilities (Batocchio et al., 2016; Hittmár, Varmus, & Lendel, 2014), but hiring new personnel with special qualification also might be needed (Knab & Rohrbeck, 2014). BM is often used as an approach to make clear what the core logic of a company is; it is also important to share and to communicate the BM within the organization. Serrano, Serrano, and Al-Debei (2010) stated that lack of communication is an important issue of successful implementation of BMI. Lack of communication also leads to distrust between employees, departments, and management (Hittmár et al., 2014), therefore can influence the effective implementation of BMI. BMI-Practices. Foss and Saebi (2017) reported that different organizational capabilities and practices are required to support BMI such as experimentation and learning through trial-and-error (Marc Sosna; Rosa Nelly, Trevinyo-Rodriguez; Ramakrishna, 2010), and tools to support practitioners in managing the BMI process were presented. Moreover, as one of the management practices in BM innovation, Bocken, Weissbrod, and Tennant, (2016) stated that business experimentation not only can lead to creating more BMI but also is viewed as a process to achieve greater levels of innovation in the BM. According to Brunswicker, Wrigley, and Bucolo (2013), the BM ‘experimentation’, helps firms to test assumptions and hypothesized outcome through empirical observations such as usage data and market share. Yli-huumo, Rissanen, Maglyas, and Smolander (2015) argued that most of their research participant acknowledge that even though using experimentations might require investing more time to create and release the features to the end-user, it is still a better approach when expanding business and introducing high-quality products. Although there are some commonly used tools and frameworks such as Canvas, STOF, CSOFT, and VISOR and tools (see businessmakeover.eu) to support the process of BMI, Terrenghi et al. (2017) expressed the necessity for developing of software and tools that support the entire process of BM management. There is however hardly empirically studies on whether BM tooling contributes to the process of BMI or not (de Reuver et al., 2018, forthcoming). Karimi and Walter (2016) argued that companies mostly use the BM concept for analysis and design, but have not yet fully embraced it as management instrument in the implementation and control phases. Moreover, according to Gerasymenko et al. (2015) and Nicholls-Nixon and Cooper (2000); the scope of innovation in BM, for instance, to change in one or two components or to change entire the BM, i.e., an architectural change, can impact the outcome of BMI. Apart from the scope, the speed of change and the path followed to reach the new BM plays a critical role (Foss & Saebi, 2016). The degree of novelty of BM also can be important. There is a different level of performance expected for new to the firm than for new to the industry or even new to the world BMs (Dahlin & Behrens, 2005; Zott & Amit, 2007). Firm-characteristics. Prior research revealed that firms do not attain equally from innovativeness because their capability to capture the value of innovativeness depends on different characteristics of firm and industry (Sorescu and Spanjol 2008; Tellis, Prabhu, and Chandy 2009). Therefore, some specific characteristics of organizations can strengthen the relationship between BMI on performance. Based on our literature review, firm-characteristics consists of firm size, firm experience, firm age, advertising intensity, expenditures on R&D, the intensity of change, and scope of change in BM. Hartmann et al. (2013), found the size of the firm and experience of the firm to be positively related to performance (Zott & Amit, 2007; Klepper & Simons, 2000) and revealed the moderating role of firm size and firm experience. The size of the firm has been associated with firm innovation in a variety of research (Gronum, Steen, & Verreynne, 2016; Heij, Volberda, & Van den Bosch, 2014; Rubera & Kirca, 2012; Damanpour, 1991). Rubera and Kirca (2012) argued that larger the firms, the more likely to benefit from innovativeness in terms of market and financial positions. They can utilize more resources and reach consumers more quickly because they have better access to distribution channels, enjoy economies of scale, benefit from brand reputation. Gronum et al. (2016) consider firm age, measured by the number of years since its founding, as moderating variables on the relationship between BMI and performance. In their research, Heij et al. (2014) also took into account firm age as a moderator variable. Rubera and Kirca (2012) and Zott and Amit (2007) considered the advertising intensity as a moderating effect in their research. Furthermore, expenditures on R&D has been seen as an important factor which influences the relationship between BMI and firm performance (Zott & Amit, 2007). Gerasymenko et al. (2015) revealed that the intensity of change in BM innovation, incremental or radical change, has a moderating effect on the relationship between BMI and firm performance. They also consider the scope of change, innovation in the core or peripheral aspects of the firm BM as a moderator. Industry-characteristics. Industry-characteristics impact has theoretical foundations in the Industrial Organisation theory discussing environmental factors in relation to the industry in which a firm acts and having a significant influence on firm’s performance (Rauter et al., 2017). We considered industry sector, industry life cycle, industry competition, environmental conditions (dynamism, complexity, and turbulence), high-technology versus low-technology industries as relevant industry-characteristics factors. Gronum et al. (2015), Heij et al. (2014), and Brettel, Strese, and Flatten (2012) considered industry sector as a moderating factor between BMI and firm performance. Moreover, industry life cycle has an important role to play in affecting BMs (Wei, Song, & Wang, 2017). Waldner et al. (2015) expressed that most BMI to occur in the emergent life cycle stage of the industry, but not in mature or in decline stages. Rules of the game in the business world are not only determined by the firm’s actions but also by competitors and environmental causes (Carayannis, Sindakis, & Walter, 2015). When a potential competitor decides to enter the industry, the firm may have to modify its plan of action based on competitor’s BM (Casadesus-Masanell & Ricart, 2010). Waldner et al. (2015) in their empirical research on a sample of 1,242 Austrian firms, argued that industry competition negatively influences the degree of BMI. In different studies, dynamic environments were considered as moderating variable. Dynamic environments can be seen as a source of opportunities (Schneider and Spieth, 2013). Thus, in more dynamic environments, developing and running a new BM can be expected to have a stronger effect on firm performance than in less dynamic environments (Heij et al., 2014). On the other hand, regulation changes occurring within and outside the industry can influence the performance of an ongoing BM innovation (Bohnsack, Pinkse, & Kolk, 2014; De Reuver et al., 2009). Zott and Amit (2007) in their research on 190 entrepreneurial firms, found a little support to moderating effect of environmental conditions on the relationship between BM design and the performance of a firm. Rubera and Kirca (2012) argued that since innovation has a different impact on high-technology and low-technology industry, the technology level of industry can be taken into account as a moderator. Innovation is crucial for competition in high-tech industries, in which firms are forced to constantly introduce new products to meet rapidly changing consumer needs. 4.2 Mediation Variables We found ten mediating effects and categorized them into four groups which are associated with Efficiency Growth, Revenue Growth, and Enhancing the Organizational Capabilities. Efficiency Growth. Zott and Amit (2007) stressed that one of the key influences of BMI on firm performance occurs when it focuses on efficiency. BMI can take the ICT ventures to complete its transactions more efficient, by reducing transaction costs within the firm and with outsiders (Ben Romdhane Ladib & Lakhal, 2015). According to Chesbrough (2007), BMI leverage performance through utilizing available resources more effectively, as well as reduction in production costs. For instance, by adopting new partnering models such as outsourcing, organizations are able to more effectively scale down operations. Efficiency-centered BM design theme, according to Gronum et al. (2015) can enhance the firm performance through reduction in inventory costs and decrease in marketing, sales, and other communication expenditures for all participants in company supply chain. Furthermore, increasing scale leads to reducing operating costs. Therefore, by focusing on lowering operating cost, benefits can be passed on to customers. Revenue Growth. BMI can provide opportunities both during periods of rapid economic growth and at times of economic downturn via exploiting ongoing industry transformation. Plenty of scholars stated that BMI through creating values (Teece, 2010) and opportunity recognition (Guo et al., 2017) leads to exploit a market niche not addressed by its competitors (Zott & Amit, 2007) and in such a way, increase firm’s revenue growth. To illustrate, by adopting new partnering models, organizations are able to create additional access to additional resources to scale up quickly or complementary resources as new opportunities for service bundling arise. BMI using new ways for economic exchanges and focusing on novelty can create value for stakeholders (Ladib & Lakhal, 2015). Introducing a new BM with new components also can provide opportunities for new complementary effects among existing components of services and products (Heij et al., 2014) an in this way it can increase revenues. Vermeer (2016) shows that new revenue model, as one of the components of BM, has the strongest effect on firm performance (Vermeer, 2016). Gronum et al. (2015) confirmed the mediatory effect of the novelty design theme between innovation breadth and firm performance. They stated the BM novelty could improve the performance through the following mechanism; firm can offers new combinations of products, services, and information to customers (bundling), links customers to products/services in novel ways (new experience) (Bouwman et al, 2008), design new transaction mechanism (Zott & Amit, 2007), and finally, innovating in one component needs to be complemented by changes in other components. Enhancing the Organizational Capabilities. Organizational capabilities are required to make BMI efforts. Firms to be enabled to renew their BM, need to possess a certain level of capabilities such as detecting new technology and market needs, open-mindedness and innovativeness. Once firm starts to explore, design, test and implement their BMI, this process of BMI, enhance its organizational capabilities. For example, being opportunity seeker is required for developing a BMI, on the other hand, conducting a BMI after a while, improve opportunity seeking capability of the firm. In other words, with BMI firms are expected to improve their organizational capabilities. The capacity to innovate is one of the most significant factors that improve business performance (Burns & Stalker, 1961; Porter, 1990). Culture, defined as expressed norms, values, and beliefs, boost behavior ultimately related to business performance (Hult et al., 2004). When specific attitudes are accommodated in the organizational culture, the consequences of behavior are expanded across circumstances, groups, and individuals within the firm. As Barney (1986) stated a culture that supports the implementation of a strategic attempt, like in our research BMI, is not easy to imitate and then can lead to a sustainable competitive advantage. Another organizational capabilities might mediate the relationship between BMI and performance is the capability of opportunity-seeking. The role of BMI in opportunity-seeking behavior has been emphasized in several studies (Chesbrough, 2010; Dewald and Bowen, 2010; Doz and Kosonen, 2010). Exploiting opportunities were proved as a mean to impact firm performance by BM (Bock et al., 2012). BMI is a result of performing innovation in the firm BM. As a consequence, BMI can increase the innovativeness of people within the firm in various ways such as sharing the business idea entire organization, developing the opportunity seeking capabilities and creating real value propositions . Moreover, organizational learning is one of the critical organizational processes through which information and knowledge can be processed, and it can change the attributes, behaviors, capabilities, and performance of an organization. Hu (2014) in his research conducted on 163 companies confirmed that BMs affect technological innovation performance through organizational learning indirectly. Mahmood and Hanafi (2013) affirmed that entrepreneurial orientation is a capability that provides a competitive advantage and impressive performance to the firm. In addition, research conducted by Ladib et al. (2015) also expressed that by learning how gathering the unique know-how and utilizing of rare resources, BMI creates a benefit of a hard to imitate innovation. Building on the literature review, we developed an exhaustive and reference model (Figure 3) to explore the relationship between BMI and firm performance. Mediating VariablesModerating VariablesOverall PerformanceOverall PerformanceRevenue GrowthEfficiency GrowthCost ReductionNovel Value PropositionBM PracticesRadical vs Incremental changeDegree of Novelty Scope of Change BM ToolingOrganizational CapabilitiesOrganizational Learning Organizational Culture Opportunity Recognition Entrepreneurial-OrientationInnovativenessNew Markets Productivity ImprovementFirm Characteristics R&D expenditureOwnershipAdvertising expenditureIndustry CharacteristicsIndustry Life CycleCompetitive intensityEnvironment DynamismIndusrty SectorFirm AgeFirm SizeHigh-Tech vs LowBM ExperimentationBusiness Model InnovationCustomer EngagementService BundlingBM ImplementationDetailed PlanReward SystemCommnicationEmployees’ SkillsEmployees’ CommitementManagement SupportReducing time to marketNew Customers Figure 3: Research outcome: BMI mechanism to boost firm performance 5 Discussion and Conclusion These days the only thing that is constant is change itself. Adapting to change is crucial for businesses. Being aware of this complex and dynamic environment, companies have to introduce a new way of earning money and improve their performance by innovating their BM. Although in the last two decades researchers conceptualized, defined and provided different frameworks and tools to support practitioners and company-owners to develop their BMI, there is still ambiguity about how BMI leads to firm performance. We found empirical research with contradictory outcomes; some revealed that there is a positive, some negative and in some cases they could not find significant relations. Understanding the mechanism under which BMI impact performance is not an easy task due to many variables at play, the long assumed causal chain of events, while in practices BMI processes are far from linear and can be considered to be a kind of tinkering process (Heikkilä et al., 2018). As stated by Foss and Saebi (2016), it is not a linear mechanism for creating value starting from suppliers moving to the firm and the customers. Value creation comprises a more complicated, interconnected set of interactions and activities among different actors. To fully understand the effect of BMI on performance, it requires clear identification of the causal structure in the relationship between BMI and performance (Fry & Smith, 1987) and in-depth know-how what actually happens in BMI projects, with paths are followed and with what outcomes, as is shown by extensive case study research (Heikkilä et al., 2018). In this paper, based on systematic literature review of empirical studies, we develop a model that presents potential moderating, and mediating effects are playing a role when researching BMI and firm performance (see figure 3). To identify under which conditions, the relationship between BMI and firm performance can be affected, we could identify twenty moderating factors and classify them in four sub-groups; Firm-Characteristics includes firm size, age, advertising and R&D expenditures, and type of ownership; Industry-Characteristics consists of industry sector and life cycle, competition intensity, environment dynamism and level of technology; BM Implementation comprise of top management support, employees’ commitment and skills, communications among different stakeholders and, having detailed plan of action, and the last one, BM Practices; includes BM tooling, BM experimentation, Scope of change, Speed of Change, and Degree of Novelty. While looking at this different moderating subgroups, the moderating factors which are related to firm-characteristics and industry-characteristics are mostly non-variant, and firms cannot manage them to improve the performance of BMI effort. For instance, although firm size, firm age, and industry sector can influence the relationship between BMI and performance, firm owners and managers are not probably able to change them. On the other hand, moderating factors which associated with BM-Implementation and BM-Practices are mostly actionable. For example, practitioners can use BM tooling, BM experimentation or increase their employees' motivation or not; they can make the decision for taking some specific action. BM-Implementation and BM-Practices related factors are more manageable. Therefore, there is a lot of room for firm owners, managers and even researchers to work in these two specific subgroups to reach a higher firm performance. To answer under which casual mechanism the BMI indirectly influence the firm performance, in next step, we found 10 mediating factors and classify them into three sub-groups; Efficiency & Cost Reduction, e.g. focus on productivity, reducing time to market, and cost reduction due to partnership; New Customers & Revenue Growth, e.g. Focus on new value propositions, customer engagement, service bundling and creating lock-in effect, and the next mediator sub-group, Enhancing the Organizational Capabilities consists of innovativeness, organizational learning, organizational culture, developing opportunity recognition and fostering entrepreneurial orientation. Although our findings revealed that the research to find the mediation and moderation factors between BMI and firm performance are limited, we found 23 studies proposing moderating effect and 19 research considered specific factors as control variables (which means that they can affect the relationship). However, only eleven studies tested the causal relationship and mediation effects between BMI and firm performance. Because of the significant role of mediation factors in understanding how BMI improve the outcomes of the company, more research in this field is required. Developing a testable casual model of how BMI impact on firm performance, we argue that we fill the gap in the literature and the model can pave the way for future researchers. Authors, at a next step, aim to examine the model by using empirical data gathered from SMEs throughout of Europe in 2017 and 2018. This paper also has some limitations. We were able to identify 37 articles in the selected academic database which hypothesized and empirically tested the mediation and moderation effects on relationships between BMI and firm performance for in-depth analysis; However, by focussing on journal papers only, we might have missed recent research, as well as research in other languages than English. Limitations related to selection bias may also play a role. Acknowledgment The work leading to these results has received funding from the European Community’s Horizon 2020 Program (2014–2020) under grant agreement 645791. The content herein reflects only the authors’ view. The European Commission is not responsible for any use that may be made of the information it contains. References Afuah, A. and Tucci, C. (2000) Internet Business Models and Strategies: Text and Cases. McGraw- Hill, New York. Baden-Fuller C, Haefliger S. (2013). Business Models and Technological Innovation. Long Range Planning. Batocchio, A., Ghezzi, A., & Rangone, A. (2016). A method for evaluating business models implementation process. Business Process Management Journal, 22(4), 712–735. https://doi.org/10.1108/BPMJ-08-2015-0117 Ben Romdhane Ladib, N., & Lakhal, L. (2015). Alignment between business model and business strategy and contribution to the performance: Empirical evidence from ICT Tunisian venture. Journal of High Technology Management Research, 26(2), 168–176. Barney, J., (1986). Strategic factor markets: expectations, luck, and business strategy. Manag. Sd. 32 (10), 1231-1241. Bock, A. J., Opsahl, T., George, G., & Gann, D. M. (2012). The Effects of Culture and Structure on Strategic Flexibility during Business Model Innovation. Journal of Management Studies, 49(2), 279–305. https://doi.org/10.1111/j.1467-6486.2011.01030.x Bohnsack, R., Pinkse, J., & Kolk, A. (2014). Business models for sustainable technologies: Exploring business model evolution in the case of electric vehicles. Research Policy, 43(2), 284–300. Brettel, M., Strese, S., & Flatten, T. C. (2012). Improving the performance of business models with relationship marketing efforts - An entrepreneurial perspective. European Management Journal, 30(2), 85–98. https://doi.org/10.1016/j.emj.2011.11.003 Brunswicker, S., Wrigley, C., & Bucolo, S. (2013). The Experimental Nature of New Venture Creation, 139–151. https://doi.org/10.1007/978-3-319-00179-1 Burnard, P. (1991). A method of analysing interview transcripts in qualitative research. Nurse Education Today, 11(6), 461–466. https://doi.org/10.1016/0260-6917(91)90009-Y Burns, T. and Stalker, G.M. (1966) The Management of Innovation. London: Tavistock. Carayannis, E. G., Sindakis, S., & Walter, C. (2015). Business Model Innovation as Lever of Organizational Sustainability. Journal of Technology Transfer, 40(1), 85–104. Casadesus-Masanell, R., & Ricart, J. (2010). From strategy to business models and onto tactics. Long Range Planning, 45. https://doi.org/10.1016/j.lrp.2010.01.004 Chesbrough, H. (2007). Business model innovation: it’s not just about technology anymore. Strategy & Leadership, 35(6), 12–17. https://doi.org/10.1108/10878570710833714 Chesbrough, H. (2010). Business model innovation: Opportunities and barriers. Long Range Planning, 43(2–3), 354–363. https://doi.org/10.1016/j.lrp.2009.07.010 Chesbrough, H., & Rosenbloom, R. S. (2002). The role of the business model in capturing value from innovation : evidence from Xerox Corporation ’ s technology spin-off companies. Industrial and Corporate Change, 11(3), 529–555. https://doi.org/10.1093/icc/11.3.529 Christensen, C. M., Bartman, T., & Van Bever, D. (2016). The hard truth about business model innovation. MIT Sloan Management Review, 58(1). Clauss, T. (2016). Measuring Business Model Innovation: Conceptualization, Scale Development and Proof of Performance.pdf. R&D Management, (1), 1–19. https://doi.org/10.1111/radm.12186 Cook, Deborah ; Randolph, Adrienne; Kernerman, Phillip; Cupido, Cynthia; King, Derek; Soukup, Clara; Brun-Buisson, Christian. ,(1997). “Central venous catheter replacement strategies: A systematic review of the literature”, Critical Care Medicine, August 1997 - Volume 25 - Issue 8 - pp 1417-1424 Dahan, N., Doh, J., Oetzel, J., Yaziji, M., (2010). Corporate NGO Collaboration: Co-Creating New Business Models for Developing Markets. Long Range Planning 43 (2-3), 326e342. Damanpour, Fariborz (1991), “Organizational Innovations: A Meta-Analysis of Effects of Determinants and Moderators,” Academy of Management Journal, 34 (3), 555–91. Dewald, J. and Bowen, F.E. (2010) ‘Storm clouds and silver linings: responding to disruptive innovations through cognitive resilience’, Entrepreneurship Theory and Practice, Vol. 34, No. 1, pp.197–218. Dey I. (1993) “Qualitative Data Analysis. A User-Friendly Guide for Social Scientists”. Routledge, London. Doz, Y.L. and Kosonen, M. (2010) ‘Embedding Strategic Agility - A Leadership Agenda for Accelerating Business Model Renewal.’ Long Range Planning, vol. 43, no. 2-3, pp. 370-382. Foss, N. J., & Saebi, T. (2016). Fifteen Years of Research on Business Model Innovation: How Far Have We Come, and Where Should We Go? Journal of Management, XX(X), 1–28. https://doi.org/10.1177/0149206316675927 Fry, L. W., & Smith, D. A. (1987). Congruence, contingency, and theory building. Academy of Management Review, 12: 117-132. George, G., Bock, A.J. (2011). The Business Model in Practice and its Implications for Entrepreneurship Research. Entrep. Theory Pract. 35, 83–11. Gerasymenko, V. De Clercq, D. Sapienza, H. J. (2015). Changing the Business Model: Effects of Venture Capital Firms and Outside CEOs on Portfolio Company Performance. Strategic Entrepreneurship Journal, 9:1, 79-98. Gronum, S., Steen, J., & Verreynne, M.-L. (2016). Business model design and innovation: Unlocking the performance benefits of innovation. Australian Journal of Management, 41(3), 585–605. Guo, H., Tang, J., Su, Z., & Katz, J. A. (2017). Opportunity recognition and SME performance: the mediating effect of business model innovation. R&D Management, 47(3), 431–442. Haggège, M., Gauthier, C., & Rüling, C.-C. (2017). Business model performance: five key drivers. Journal of Business Strategy, 38(2), 6–15. https://doi.org/10.1108/JBS-09-2016-0093 Hartmann, M., Oriani, R., & Bateman, H. (2013). The Performance Effect of Business Model Innovation: An Empirical Analysis of Pension Funds. 35th DRUID Celebration Conference 2013, 1–34. https://doi.org/10.5465/AMBPP.2013.10986abstract Heij, C. V., Volberda, H., & Van den Boasch, F. (2014). How Does Business Model Innovation Influence Firm Performance : the Moderating Effect of Environmental Dynamism. Heikkilä, M., H. Bouwman, & J. Heikkilä (2017,). From Strategic Goals to Business Model Innovation Paths: an exploratory study. Journal of Small Business and Enterprise Development. Hittmár, Š., Varmus, M., & Lendel, V. (2014). Proposal of Model for Effective Implementation of Innovation Strategy to Business. Procedia - Social and Behavioral Sciences, 109, 1194–1198. https://doi.org/10.1016/j.sbspro.2013.12.611 Hu, B. (2014). Linking business models with technological innovation performance through organizational learning. European Management Journal, 32(4), 587–595. https://doi.org/10.1016/j.emj.2013.10.009 Karimi, J., & Walter, Z. (2016). Corporate Entrepreneurship, Disruptive Business Model Innovation Adoption, and Its Performance: The Case of the Newspaper Industry. Long Range Planning, 49(3), 342–360. https://doi.org/10.1016/j.lrp.2015.09.004 Klepper, S., Simons, K.L., (2000). Dominance by birthright: entry of prior radio pro- ducers and competitive ramifications in the U.S. television receiver industry. Strategic Management Journal, 21, 997–1016. Knab, S., & Rohrbeck, R. (2014). Why intended business model innovation fails to deliver: Insights from a longitudinal study in the German smart energy market, 527–538. Retrieved from http://papers.ssrn.com/sol3/Delivery.cfm?abstractid=2447651 Lambert, S. C., & Davidson, R. A. (2013). Applications of the business model in studies of enterprise success, innovation and classification: An analysis of empirical research from 1996 to 2010. European Management Journal, 31(6), 668–681. https://doi.org/10.1016/j.emj.2012.07.007 Latifi, S. M., & Bouwman, H. (2017). Why does Business Model Innovation fail to deliver expected outcomes? Proceedings of ISPIM Conferences, (March), 1–16. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=122622746&site=ehost-live Magretta, J. (2002). Why Business Models Matter, HBR.pdf, 86–92. Marc, Sosna; Rosa Nelly, Trevinyo-Rodriguez; Ramakrishna, V. (2010). Business Model Innovation Through Try and Error Learning_Sosna_2010.pdf, 43(383), 407. Martins, L., Gerasymenko, V., De Clercq, D., & Saprienza, J. H. (2014). Changing the business model: effects of venture capital firms and outside CEOs on portfolio company performance. Strategic Entrepreneurship Journal, 1(1), 27–47. https://doi.org/10.1002/sej McGrath, R. G. (2010). Business models: A discovery driven approach. Long Range Planning, 43(2–3), 247–261. https://doi.org/10.1016/j.lrp.2009.07.005 Methlie, L. B., & Pedersen, P. E. (2008). Business model performance: Reflections from three studies of mobile data services. Nordic and Baltic Journal of Information and Communications Technologies, 2(1), 23–38. Mitchell D, Coles C. (2003). The Ultimate Competitive Advantage of Continuing Business Model Innovation. Journal of Business Strategy 24(5): 15-21. Murphy GB, Trailer JW, Hill RC (1996). Measuring performance in entrepreneurship research. J. Bus. Res., 36: 15-23. Nair, S., Paulose, H., Palacios, M., & Tafur, J. (2013). Service orientation: effectuating business model innovation. The Service Industries Journal, 33(9–10), 958–975. https://doi.org/10.1080/02642069.2013.746670 Nancy M.P. Bocken, I. W. and M. T. (2016). Business Model Experimentation for Sustainability. Smart Innovation, Systems and Technologies, 52. https://doi.org/10.1007/978-3-319-32098-4 Osterwalder, A. (2012), “Why business models fail?”, Business Model Alchemist, available at: http://businessmodelalchemist.com/blog/2012/05/why-business-models-fail.html (accessed Jan. 22, 2018). Porter, M.E., (1990). The Competitive Advantage of Nations. The Free Press, New York. Rauter, R., Jonker, J., & Baumgartner, R. J. (2017). Going one â€TM s own way : drivers in developing business models for sustainability. Journal of Cleaner Production, 140, 144–154. Reim, W., Panda, V., & Ortqvista, D. (2015). Product — Service System s ( PSS ) business m odels and tactics — a system atic literature review. Journal O F Cleaner Production, 97, 61–75. Rubera, G., & Kirca, A. H. (2012). Firm Innovativeness and Its Performance Outcomes: A Meta-Analytic Review and Theoretical Integration. Journal of Marketing, 76(3), 130–147. 4 Schneider S, Spieth P. (2013). Business Model Innovation: Towards an Integrated Future Research Agenda. International Journal of Innovation Management 17(1) Serrano, A. E., Serrano, A., & Al-Debei, M. (2010). Implementing an e-business model for a dot-com SME: Lessons learned. 16th Americas Conference on Information Systems 2010, AMCIS 2010, 6, 4566–4574. Shafer, S.M., Smith, H.J., Linder, J.C., (2005). The power o f business models. Bus. Horiz. 48 (3), 199-207. Smith WK, Binns A, Tushman ML. (2010). Complex Business Models: Managing Strategic Paradoxes Simultaneously. Long Range Planning 43(2-3): 448-461. Sorescu, Alina B., Jelena Spanjol (2008), “Innovation’s Effect on Firm Value and Risk: Insights from Consumer Packaged Goods,” Journal of Marketing, 72 (March), 114–32. Sorescu, Alina B., Rajesh Chandy, and Jaideep C. Prabhu (2003), “Sources and Financial Consequences of Radical Innovation: Insights from Pharmaceuticals,” Journal of Marketing, 67 (October), 82–102. Spieth, P. Schneckenberg, D. Ricart, J. E. (2014). Business model innovation - state of the art and future challenges for the field. R&D Management 44:3, 237-247. Teece, D. J. (2010). Business models, business strategy and innovation. Long Range Planning, 43(2–3), 172–194. https://doi.org/10.1016/j.lrp.2009.07.003 Tellis, Gerard J. Jaideep C. Prabhu, and Rajesh K. Chandy (2009), “Radical Innovation Across Nations: The Preeminence of Corporate Culture,” Journal of Marketing, 73 (January), 3–23. Terrenghi, N., Schwarz, J., Legner, C., & Eisert, U. (2017). Business Model Management: Current Practices , Required Activities and IT Support. 13th International Conference on Wirtschaftsinformatik, 972–986. Venkatraman; Ramanujam, V. (1986). Measurement of Business Performance in Strategy Research : A Comparison of Approaches Massachusetts Institute of Technology. Academy of Management Review, 11(4), 801–814. Vermeer, T. (2016). The relationship between business models and firm performance as measured through business model components. 7th IBA Barchelor Thesis Conference. Waldner, F., Poetz, M. K., Grimpe, C., & Eurich, M. (2015). Industry Structure and Business Model Innovation Antecedents and consequences of business model innovation: The role of industry structure, 33(33). Wei, Z., Song, X., & Wang, D. (2017). Manufacturing flexibility, business model design, and firm performance. International Journal of Production Economics, 193(28), 87–97. Yannopoulos, P. (2013). Business Model Innovation : Literature Review and Propositions Proceedings of International Business and Social Sciences and Research Conference. Proceedings of International Business and Social Sciences and Research Conference, Hotel Mariiott Casamagna, Cancun, Mexico, (December), 1–13. Yli-huumo, J., Rissanen, T., Maglyas, A., & Smolander, K. (2015). The Relationship Between Business Model Experimentation and Technical Debt, 210, 17–29. Zott, C., & Amit, R. (2007). Business model design and the performance of entrepreneurial firms. Organization Science, 18(2). https://doi.org/10.1287/orsc.1060.0232 Zott, C., Amit, R., & Massa, L. (2010). The business model: Theoretical roots, recent developments, and future research. IESE Research Papers, 3(September), 45. Zott, C., Amit, R., & Massa, L. (2011). The business model: recent developments and future research. Journal of Management, 37(4), 1019-1042. Towards Responsible Data Analytics: A Process Approach ROGER CLARKE & KERRY TAYLOR 7 CORRESPONDENCE ADDRESS: Roger Clarke, Xamax Consultancy Pty Ltd, Canberra, Australia, ANU Research School of Social Science, and UNSW Law, e-mail: Roger.Clarke@xamax.com.au. Kerry Taylor, Associate Professor, Research School of Social Science, Australian National University, Canberra, Australia, e-mail: Kerry.Taylor@anu.edu.au. DOI https://doi.org/10.18690/978-961-286-170-4.6 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract The big data movement has been characterised by highly enthusiastic promotion, and caution has been in short supply. New data analytic techniques are beginning to be applied to the operational activities of government agencies and corporations. If projects are conducted in much the same carefree manner as research experiments, they will inevitably have negative impacts on the organisations conducting them, and on their employees, other organisations and other individuals. The limited literature on process management for data analytics has not yet got to grips with the risks involved. This paper presents an adapted business process model that embeds quality assurance, and enables organisations to filter out irresponsible applications. Keywords: • Big Data • Data Science • Life Cycle • Business Process • Quality • 1 Introduction Big Data and Data Analytics have attracted a great deal of attention. However, the techniques are now escaping from the laboratory with only a limited degree of maturity having been achieved. Unless businesslike approaches are adopted, the intentional looseness of academic experiments may become engrained in practice. The objective of the work reported here is to specify a business process whereby organisations can ensure that applications of data analytics satisfy both strategic and policy purposes and legal and ethical constraints. A design science research approach was adopted (Brown et al. 1978, Hevner et al. 2004, Hevner 2007). Although primarily applied to the development of information technology artefacts, the approach is also relevant to socio-technical artefacts, including methods for applying technology (Gregor & Hevner 2013, p.337). As Peffers et al. (2007) acknowledges, "for design in practice, the Design Science Research Methodology (DSRM) may contain unnecessary elements for some contexts" (p.72). DSRM has accordingly been applied as a guide rather than as a specification. The research commenced with problem identification and motivation, followed by definition of the objectives. This laid the foundation for design and development of the artefact. The approach adopted in the later phases was to consider a real-world case and demonstrate that the use of the method would have been likely to identify in advance the problems that arose in the case, and hence would have avoided harm and protected investment. The paper commences with a brief review of the fields of big data and data analytics, sufficient to provide a basis for the analysis that follows. Representations of the data analytics business process are identified in text-books and the academic literature. The risks arising from data analytics activities are then considered, and techniques for identifying and addressing those risks are identified. Refinements to the conventional business process are proposed that enable data analytics to be conducted in a responsible manner. An initial evaluation of the proposed process is performed, by applying it to a case study. 2 Big Data and Data Analytics Stripped of marketing prose, the term 'big data' merely means any relatively large data collection. The key characteristics were originally postulated as volume, velocity and variety (Laney 2001). Further factors were added later, including value and veracity (Schroeck et al. 2012). Such vague formulations as 'data that's too big, too fast or too hard for existing tools to process' pervade the definitions catalogued in OT (2013), Ward & Barker (2013), and De Mauro et al. (2015). The term 'data analytics' has been used in technical disciplines for many years. It refers to the techniques whereby a data collection is used to draw inferences. Previous decades of work in statistical sciences, operations research, management science and data mining have delivered a very substantial array of analytical tools, and more are being developed. Chen et al. (2012) uses the term Business Intelligence and Analytics (BI&A), and distinguishes two phases to date. Those authors see BI&A 1.0 as being characterised by "data management and warehousing, reporting, dashboards, ad hoc query, search-based BI, [online analytical processing (OLAP)], interactive visualization, scorecards, predictive modeling, and data mining" (p. 1166). BIA 2.0, on the other hand, which has been evident since the early 2000s, is associated with web and social media analytics, including sentiment analysis, and associated-rule and graph mining, much of which is dependent on semantic web notions and text analysis tools (pp. 1167-68). The authors anticipated 'BIA 3.0', to cope with mobile and sensor-generated data. The term 'fast data' has since emerged, to refer to near-real-time analysis of data-streams (e.g. Pathirage & Plale 2015). When considering how to manage data analytics activities, it is useful to distinguish categories of purpose to which big data analytics may be applied. See Table 1. Hypothesis testing involves quite different approaches from the drawing of inferences about populations, and from the construction of profiles. Further, whereas those three categories relate to populations, several other functions to which data analytics can be applied relate to individuals. On the basis of this brief overview, the following section presents a description of the conventional business processes for data analytics. Table 1: Purposes of Big Data Analytics After Clarke (2017), Table 1 Population Focus Hypothesis Testing This approach evaluates whether a proposition is supported by the available data. The proposition may be a prediction from theory, an existing heuristic, or a hunch Population Inferencing This approach draws inferences about a population of entities, or about sub-populations. In particular, correlations may be drawn among particular attributes Construction of Profiles This approach identifies key characteristics of some category of entities. For example, attributes and behaviours of a target group, such as 'drug mules', sufferers from particular diseases, or children with particular aptitudes, may exhibit statistical consistencies Individual Focus Application of Profiles A search can be conducted for individual entities that exhibit patterns associated with a particular, previously asserted or computed profile, thereby generating a set of entities of interest. Similarly, individual entities may be classified according to previously asserted or computed profiles so that inferences about individuals may be drawn by referring to the behaviours of the class Discovery of Anomalies This approach draws inferences about individual entities within the population. For example, a person may be inferred to have provided inconsistent information to two organisations, or to exhibit behaviour in one context inconsistent with behaviour in another Discovery of Outliers Statistical outliers are often disregarded, but this approach regards them instead as valuable needles in large haystacks, because they may represent exceptional cases, or may herald a 'flex-point' or 'quantum shift' 3 The Conventional Data Analytics Process A remarkably small proportion of the multitude of papers on big data has as its focus the business process or life-cycle of data analytics. There is little evidence of a cumulative literature or a dominant authority referred back to by subsequent authors. Chen et al. (2014) suggests four phases of "the value chain of big data", but from a process perspective the authors offer little more than four phase-names and a three-step breakdown of the second phase. Their phases are: data generation, data acquisition (comprising collection, transportation and pre-processing), data storage and data analysis. Jagadish et al. (2014) uses instead a five-step process, comprising: acquisition; extraction and cleaning; integration, aggregation, and representation; modeling and analysis; and interpretation. The framework of Pääkkönen & Pakkala (2015) involves seven process phases of source acquisition (implied), extraction, loading and preprocessing, processing, analysis, loading and transformation, and interfacing and visualisation. Huang et al. (2015) has six steps, commencing with question formulation and proceeding to data collection, data storage and transferral, data analysis, report / visualisation, and evaluation. Phillips-Wren et al. (2015) uses the five phases of sources, preparation, storage, analysis, and access and usage. Elragal & Klischewski (2017) offer 'pre-stage', acquisition, preprocessing, analytics and interpretation. The Wikipedia entry for Predictive Analytics is somewhat more comprehensive, with Project Definition, Data Collection and Data Analysis, followed by Statistics, Modelling and Deployment, and culminating in a Model Monitoring phase. Text-book approaches include Provost & Fawcett (2013), which proposes the phases Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation and Deployment. In Figure 1, a process model is presented that is a composite of the elements found in the above sources. However, it embodies a key refinement. The majority of the examples that have been identified omit any preliminary phase. It would arguably be too constraining to impose a 'requirements elicitation' phase on those data analytics activities that are oriented towards opportunity discovery and creation. On the other hand, some degree of framing is needed, even for creative work. The expression 'Terms of Reference' has accordingly been used for the first phase, to encourage pre-thinking about the project's context, but also to encompass degrees of formalisation ranging from 'question formulation' to 'requirements analysis'. The subsequent phases will vary in their content depending on the nature and purpose of the particular project. The second, Data Source Discovery phase, covers such activities as search, evaluation, and negotiation of access. The third phase, Data Acquisition, deals with the collection or extraction of data from the identified source(s) and may include primary data collection from source through surveys, polls, or on-line instrumentation. In the fourth, Data Pre-Processing phase, data from various sources may be integrated and may be scrubbed, re-formatted, interpolated and/or modified. These lay the foundations for the fifth and sixth Data Analysis and Data Exploitation phases, including such activities as inferencing, visualisation, interpretation, and application. Figure 1: A Conventional Process Model for Data Analytics Projects The following section discusses risks that arise from the conduct of data analytics and conventional approaches to the discovery and management of risks. 4 Risks and Responsibilities Most of the literature on big data and data analytics is concerned with the identification of opportunities, and much of it is highly upbeat and excited about the prospects. Some articles do, however, urge caution, e.g. "well-designed and highly functional [business intelligence] systems that include inaccurate and unreliable data will have only limited value, and can even adversely impact business performance" (Park et al. 2012). Elragal & Klischewski (2017) also express concern about completeness, correctness, and consistency. The following paragraphs briefly summarise the results of prior research by the first-named author on quality factors in big data and data analytics, reported in Wigan & Clarke (2013), and Clarke (2015, 2016a, 2016b, 2017). Data quality factors comprise those characteristics that can be assessed at the time that the data is originally created. They include syntactical validity, appropriate association, appropriate signification, accuracy, precision and temporal applicability. Information quality factors, on the other hand, are those that can only be assessed at the time of use, and in the context of use. These include theoretical and practical relevance, currency, completeness, controls and auditability. Blithe claims are made about data quality not mattering when very large quantities of data are available. e.g. "the need to find a scrupulously accurate sample (the world of small data) has been overtaken by the availability of all of the data—much of it messy but in such volumes that new correlations can be found. In other words quantity trumps quality" (Turnbull 2014). Executives know not to take such claims at face value, because there are only limited and fairly specific circumstances in which they can be justified. Many of the purposes to which data analytics is put may be seriously undermined by low-quality data. Claims are also rife in the data mining literature that data can be 'cleaned' or 'cleansed' (Rahm & Do 2000, Müller & Freytag 2003). Such processes are better described as 'wrangling' (Kandel et al. 2011), and are most honestly referred to as 'data scrubbing'. By whatever name, these techniques seek to address such problems as missing values, syntactical errors in data content, syntactical differences among apparently comparable data-items, low quality at time of capture, degraded quality during storage and missing metadata. Energetic as the endeavour may be, however, it seldom achieves a state reasonably described as 'clean'. Few of the processes described in the literature and applied in practice involve comparisons against an authoritative external reference, most of the processes are merely manipulations based on statistical analyses of the data-set itself, and the changes made as a result of such activities introduce errors. Decision quality factors must also be carefully considered, because otherwise the value of data analytics work will be undermined, and harm will arise if the results are relied upon for real-world decision-making. The meaning of individual data-items is frequently opaque, and may be assumed rather than being understood in sufficient depth, with the result that misunderstandings and mistaken assumptions readily occur. The many and varied analytical techniques all make assumptions about data and about context, and some of those assumptions are implicit rather than clearly stated. The preconditions for use of each technique may or may not be taught in parallel with the mechanics of the technique, and the significance of that information may or may not be grasped, and may or may not be remembered. The rationale underlying the inferences that each data analytics technique gives rise to may be clear to the analyst, and may be successfully communicated to the decision-maker who relies on it. However, the transparency of rationale varies a great deal, and in many of the new techniques that have emerged during the last decade, is seriously lacking. Neural networks are largely empirical. They are not based on any formal model of a solution, of a problem, or even of a problem-domain (Clarke 1991, Knight 2017). Further, their behaviour may vary greatly depending on the training-set used and the detail of the implementation and parameter-setting of the method used against the training-set. Similarly, the various forms of AI and machine learning (ML) that are being applied in this field provide outputs, but seldom make available humanly-understandable explanations of how those outputs were achieved and what assumptions underlie them. Another approach that may suffer from the problem is 'predictive analytics', which is sometimes used in the very narrow sense of extrapolations of patterns derived from time-series rather than of the time-series themselves. The plethora of quality factors that can undermine data analytics efforts give rise to considerable risks. Many are borne by the organisation that exploits the outputs of data analytics techniques. These include negative impacts on the quality of organisational decisions and actions, and hence on return on investment and policy outcomes. There may also be opportunity costs, if resources are diverted to big data projects that, with hindsight, could have been better invested in alternative activities with higher return. Some risks are borne by other parties, however, because organisational decisions and actions affect other organisations and individuals. Decisions and actions that are unreasonable or even wrong can inflict harm ranging from inconvenience, via onerousness and inversion of the onus of proof, to serious economic, financial, psychological and/or societalharm. Such problems may return to haunt the organisation whose actions gave rise to them, in such forms as reputational damage, lawsuits and additional regulatory imposts. Wild enthusiasts such as Anderson (2008), McAfee & Brynjolfsson (2012) and Mayer-Schonberger & Cukier (2013) publish claims that science is obsolete and that the need is now to know, not to know why. Hard-headed directors, auditors and executives, on the other hand, seek assurance that authors of that ilk are unable to provide. Professionals and consultants support executives in managing risk, by means of a variety of tools. Quality assurance and formal risk assessment are well-known techniques and are the subject of copious documentation in both formal Standards (ISO 31000 and 31010, and the 27005 series) and commercial processes (ISACA COBIT, ITIL, PRINCE2, etc.). Risk assessment processes of these kinds are conducted from the organisation's own perspective. Where risks fall on other organisations and individuals, other techniques are needed. Relevant techniques include Technology Assessment (TA - Guston & Sarewitz 2002), Privacy Impact Assessment (PIA - Clarke 2009, Wright & De Hert 2012) and Surveillance Impact Assessment (Wright & Raab 2012). The following section considers how the conventional process model identified in section 3 can be adapted in order to manage the risks identified in this section. 5 An Adapted Business Process Previous research established a set of Guidelines for the conduct of data analytics in a manner that is responsible in terms of protecting the interests both of the organisation on whose behalf the activity is being performed and of other stakeholders (Clarke 2018). Those Guidelines provide a basis on which a business process can be established that addresses the weaknesses in existing approaches. The following sub- sections identify further process elements that need to be added to the conventional process depicted in Figure 1, present an adapted form of business process, and discuss alternative ways in which the proposed generic business model can be applied. 5.1 Additional Process Elements The purpose of the new artefact is to provide a basis whereby an organisation can ensure that problems arising during the data analytics process can be detected, and detected early. This enables appropriate measures to be put in place in order to address those problems. Two categories of process element need to be added to the conventional model described earlier: evaluation steps, and selection constructs that, under appropriate circumstances, loop the flow back to an earlier phase. Awareness of risks is a crucial pre-condition for effective risk management, and hence there is a need for evaluation of both data and decision processes against quality standards. Cai & Zhu (2015) and Hastie & O’Donnell (2017), for example, each includes an outline of a data quality assessment process. A few authors, such as Koronios et al. (2014), expressly include one or more evaluation elements, and feedback or feedforward from those elements. In the Knowledge Discovery in Databases (KDD) community, a 'knowledge discovery process' has been proposed that recognises the need for loops back to earlier phases where data or processes fall short of reasonable standards (Fayyad et al. 1996, Han et al. 2011). Elragal & Klischewski (2017) discuss epistemological challenges in data analytics, the importance of understanding model assumptions, and the need for results to be assessed. Various tests need to be performed, and appropriate locations need to be identified at which each test can be applied. Established techniques of quality assurance (QA) and risk assessment (RA) can be brought to bear, including requirements elicitation, interviews with executive, managerial, supervisory and operational staff, and review of the data. Similarly, the conventional classification of risk management (RM) strategies can be applied, distinguishing proactive strategies (avoidance, deterrence, prevention, redundancy), reactive strategies (detection, reduction/mitigation, recovery, insurance) and non-reactive strategies (self-insurance, graceful degradation, graceless degradation). Where significant impacts may arise outside the organisation, the additional techniques of impact assessment (IA) and impact management (IM) are needed. These activities develop an appreciation of the contexts, needs and values of external stakeholders. Many external stakeholders are users of, or at least participants in, the system in question. However, it is important not to overlook 'usees', by which is meant organisations, but more commonly individuals, who are affected by the system without directly participating in it (Clarke 1992, Fischer-Hübner & Lindskog 2001, Baumer 2015). Developing the necessary appreciation generally depends on consultation with representatives or and/or advocates for the various external stakeholder groups. 5.2 A Generic Business Process In this sub-section, the conventional model that was presented in Figure 1 is adapted, by building in evaluation steps and conditionally redirected flows. A general framework is provided by professional standards in the areas of QA, RA and IA, and RM and IM. More specifically, the adapted model reflects the specific risk factors identified in the Guidelines for the responsible conduct of data analytics (Clarke 2018). A copy of the Guidelines is provided as an Appendix to the present paper. The adapted business process is presented in Figure 2. The discussion in this section intentionally uses general and even vague expressions, in order to accommodate the considerable diversity of purposes to which data analytics is put, as discussed earlier, in section 2 and Table 1. During Phase 1 – Terms of Reference, it is advantageous to not only clarify the problem or opportunity that is the project's focus, but also the governance framework that applies, the expertise required within the team, and the legal compliance and public expectations that are relevant to the activity (paras. 1.1-1.3 of the Guidelines). These provide reference-points that support subsequent evaluation steps. After Phase 2 – Data Source Discovery, the Evaluation step needs to reconsider section 1 of the Guidelines (G1.1-1.3). Of particular significance is the legality of the intended acquisition and use of the data. This step also needs to examine the extent to which the team has understood the problem-domain (G2.1) and the nature of the data sources, including the data's provenance (Clarke 2015), purposes of creation, definitions, and quality (G2.2). Depending on the conclusions reached in the evaluation step, it may be necessary to return to Phase 1 and re-cast the Terms of Reference, or even to abort the project. After Phase 3 – Data Acquisition, the theoretical evaluation of data quality that was undertaken after Phase 2 needs to be complemented by practical assessment against the full suite of data quality and information quality factors discussed in section 4, including the incidence and impact of missing data and non- conformant data (G2.2). The effects of any merger, scrubbing, identity protection or data security measures undertaken prior to the data coming under the control of the project team must also be assessed (G2.3-2.8). After Phase 4 – Data Pre-Processing, the effects of all merger, scrubbing, identity protection and data security measures undertaken by or for the project team must be assessed (G2.3-2.8). In addition to data and information quality factors, a look-ahead to the Data Analysis phase is advisable, in order to anticipate any further issues with data characteristics that may arise, such as incompleteness, inconsistency, and format and measurement scale incompatibilities. It may transpire that the project flow needs to depart from the mainstream. For example, it may be that the project should be held in Phase 4 until further work relevant to data quality is performed. Alternatively, it may be necessary to loop back to the 3rd phase, e.g. by re-acquiring the data using different parameter-settings or procedures, or to the 2nd phase, to acquire data from alternative sources, or even to the 1st phase, in order to re-conceive the project. After Phase 5 – Data Analysis, the evaluation step needs to consider all of Guidelines 3.1-3.6. These relate to the adequacy of the expertise applied to the analytics, the nature of, and the intrinsic assumptions underlying, the relevant analytical techniques, the nature of the data, the compatibility of the data and the technique, the statistical confidence or error rate supporting the inferences drawn, and the transparency of the rationale for inferences drawn. Where the data or the inferences drawn involve sensitivity and/or the actions taken as a result are likely to be particularly impactful, a look-ahead may be advisable, in particular by performing some preliminary reality testing (G4.3). Figure 2: A Business Process Model for Responsible Data Analytics Projects After Phase 6 – Data Exploitation, it is important to apply all of Guidelines 4.1-4.11. These involve appreciation of the impacts (e.g. by workshopping with staff and others with familiarity with the relevant parts of the real world), internal cost/benefit and risk assessment, where relevant external impact assessment, reality testing, the design, implementation and testing of safeguards (such as metricated pilots, interviews, feedback processes, feedback evaluation), proportionality checking, contestability, testing of the understandability of the decision-rationale, and review and recourse. Depending on the nature and potential impact of the actions that are being considered, it may be advisable to commence the evaluation process at an early stage in this phase, rather than at the end of it. The adapted business process features the same phases as the conventional model. Evaluation steps have been specified after each phase, together with guidance in relation to circumstances in which looping back to prior phases is desirable and even essential. The adapted model enables the identification of problems at an early stage, and hence the implementation of measures to address them. It therefore fulfils the declared objective of the research, which was to specify a business process whereby organisations can ensure that applications of data analytics satisfy both strategic and policy purposes and legal and ethical constraints. 5.3 Instantiations The business process model in Figure 2, and the Guidelines on which it was to a considerable extent based, were expressed in somewhat abstract terms. That was necessary in order to achieve sufficient generality to enable application in a range of circumstances. One dimension of diversity among projects is the category of purpose, as discussed in section 2 and Table 1. Another way in which projects differ is their degree of embeddedness within a corporate framework. In the case of a standalone project, each phase and each evaluation step may need to be planned and performed as a new activity. At the other extremity, every phase and every step may be tightly constrained by existing corporate policies and practices, perhaps in the form of an industry standard or a proprietary process management framework imposed by or on the organisation. Another factor to consider in applying the model to a project is the extent to which the project is ground- breaking or novel. A less painstaking approach can reasonably be adopted where the project falls into a well-known category, and is being conducted by a team with both expertise and experience in relation to the problem-domain, the data-sets, the data analytic techniques, the pitfalls, the stakeholders and their interests, and the project's potential impacts. The following section makes an initial contribution to the evaluation of the proposed business process. 6 Application and Evaluation In Peffers et al. (2007), two related phases are defined towards the end of the design research approach. These distinguish 'demonstration' of the use of the artifact to solve one or more instances of the problem (by means of experimentation, simulation, case study, proof, or other appropriate activity), from 'evaluation', which involves more formal observation and measurement of the new artefact's effectiveness in addressing the stated objectives. The research reported in this paper includes a demonstration step, based on a case study. Centrelink is the Australian government agency responsible for distributing welfare payments. During the second half of 2016, Centrelink launched a new system that was intended to improve the efficiency of the agency's processes whereby control is exercised over overpayments. The new Online Compliance Intervention (OCI) system featured simple data analytics that were used to draw inferences about likely overpayments of welfare payments, combined with automated decision-making. The implementation resulted in a large proportion, and very large numbers, of unjustified and harmful actions by the agency, which gave rise to serious public concern and two external investigations, which in turn forced the agency to make multiple changes to the scheme. This represents a suitable test-case for the artefact developed in this research project. The data analytics used were trivially simple, but the big data collections that were matched were not. The project was real – and, indeed, for many of the people affected by it, all too real. It involved all steps of the life-cycle. In addition, unlike most private sector data analytics projects, it was subject to the glare of publicity, and is documented by two substantial and independent reports. A 3,000-word case study was prepared by the first-named author, based on the two reports and about 20 substantive media articles. It is provided as Supplementary Material in support of the present paper. The major problems arising in the case appear to have resulted from a small number of factors. The most critical issue was that the new system implicitly assumed that the annual income declared by welfare recipients to the taxation agency could be divided by 26 in order to establish a reliable estimate of the income that each of them earned in each fortnight of that year. Many welfare recipients, on the other hand, earn some and even all of their small incomes from short-term, casual and/or seasonal employment, and hence they work variable numbers of hours per fortnight. This has the inevitable result that their income is unevenly distributed across the year, and an assumption of even distribution is seriously problematical. Centrelink failed to appreciate how significant that issue was, and remained in denial 12 months after implementation. The generic business process proposed in Figure 2 includes within Phase 1 – Terms of Reference the establishment of a governance process, identification of the expertise required within the team, and consideration of the legal compliance and public expectations that are relevant to the activity. If the team had included people with strong familiarity with both the real world of benefit recipients and the relevant data models, together with experienced data analysts, then the unreasonableness of the assumption of income being evenly distributed over a period would have been obvious. Similarly, if the agency's legal obligations, and the vulnerability of many of its clients, had been clear in the minds of the team-members, they would have taken greater care in considering the potential impacts of the design. Even if these framing elements had been overlooked, any and all of the five evaluation steps across the recommended life-cycle would have been very likely to identify this major problem well in advance of implementation. In particular, it would have become clear that the design process needed to include consultation with the operational staff, the taxation agency, and advocates for welfare recipients' interests. A second major problem was the abandonment of checks with employers, which Centrelink had identified as an avoidable cost. The agency sought to transfer these costs to the recipients; but this proved to be unreasonable, partly because of many recipients' cognitive and performative limitations, and partly for systemic reasons. Had the proposed business process model been used, the unreality of this key assumption might also have been uncovered during one of the early-phase evaluation steps; but at the very least it would have become apparent during Phase 6 – Data Exploitation. A third problem was the automation of both debt-raising and the commencement of debt collection. The triggers were nothing more than non-response by the targeted welfare recipient, or their failure to deliver satisfactory evidence to prosecute their innocence of the accusation. The previous system had involved considerable computer-based support, but also a number of manual steps. Under the new, naive system of suspicion-generation, the case-load leapt more than 30-fold, from 20,000 p.a. to 10-20,000 per week. This overwhelmed the support services, resulting in a complete log-jam of enquiries and complaints, and escalating numbers of auto-generated debts and debt-collection activities. Several of the elements making up this compounding problem could have been intercepted early in the process, and all of them would have become apparent at latest during the Phase 6 evaluation steps, if the business process in Figure 2 had been applied. The Centrelink case demonstrates the benefits of inserting QA elements between the successive phases, and the insertion of loops where problems are found. It also provides the valuable insight that a conventional requirements analysis that incorporated interviews with operational staff, followed by internal risk assessment, may well have been sufficient to prevent much of the harm, but that these alone would have very likely missed some key factors. Because the system directly affected 'usees', external impact assessment was necessary, including consultation with advocacy organisations. 7 Implications and Conclusions Conventional business processes for data analytics lack three important features: a preliminary, planning phase; evaluation steps; and criteria for deciding whether a project needs to be looped back to an earlier phase. On the basis of established theories and prior research into risk assessment of data analytics projects, an adapted business process model was proposed, which makes good those deficiencies. A recent case was considered in the light of the adapted model. The implications for practice are clear. Data analytics embodies many risks. Organisations that conduct or commission data analytics projects are subject to legal obligations. In some contexts, such as highly- regulated industry sectors, these may be fairly specific; but in any case company directors are subject to broad responsibilities in relation to the management of risk. The concerns of consumers and citizens are increasing, the media is eager to snowball cases of large organisations treating people badly, and social media has provided means for the public to escalate issues themselves. Investors, board rooms and executives will demand that a balance is struck between data exploitation and due care. QA, RA and RM, and IA and IM, need to be applied. The adapted business process shows how. The research reported here can be strengthened in a number of ways. It is built primarily on academic work, because reliable reports of active data analytics projects are difficult to acquire. A much stronger empirical base is needed, such that actual business processes can be better understood, and the adapted model's efficacy can be evaluated. The model is, for the reasons explained above, generic in nature. It requires tailoring to the various specific contexts identified above, and further articulation. A number of implications can be drawn for research in the area. Observation of practice, and publications arising from it, are likely to encounter nervousness on the part of project teams and the organisations for whom the project is undertaken. Some concerns will relate to competitive, commercial and strategic factors, and others to ethical, legal and political considerations. Academic projects to apply data analytics tools need to incorporate controls at a much earlier stage than is currently the case, to ensure that the transitions from research to industrial R&D, and on to live use, are not undermined by the late discovery of quality issues. The work reported here accordingly provides a substantial contribution to both practice and research. References Anderson C. (2008) 'The End of Theory: The Data Deluge Makes the Scientific Method Obsolete' Wired Magazine 16:07, 23 June 2008, at ://archive.wired.com/science/discoveries/magazine/16-07/pb_theory Baumer E.P.S. (2015) 'Usees' Proc. 33rd Annual ACM Conference on Human Factors in Computing Systems (CHI'15), April 2015 Brown H., Cook R. & Gabel M. (1978) 'Environmental Design Science Primer' Advocate Press, 1978 Cai L. & Zhu Y. (2015) 'The Challenges of Data Quality and Data Quality Assessment in the Big Data Era' Data Science Journal 14 (2015) 2, at http://doi.org/10.5334/dsj-2015-002 Cao L. (2016) 'Data science: nature and pitfalls' IEEE Intelligent Systems 31, 5 (September/October 2016) 66-75, at http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=7579413 Cao L. (2017) 'Data science: a comprehensive overview' ACM Computing Surveys 50, 3 (October 2017), at http://dl.acm.org/ft_gateway.cfm?id=3076253&type=pdf Chen H., Chiang R.H.L. & Storey V.C. (2012) 'Business Intelligence and Analytics: From Big Data to Big Impact' MIS Quarterly 36, 4 (December 2012) 1165-1188, at http://ai.arizona.edu/mis510/other/MISQ%2520BI%2520Special%2520Issue%2520Introduction%2520Che n-Chiang-Storey%2520December%25202012.pdf Chen M., Mao S. & Liu Y. (2014) 'Big Data: A Survey' Mobile Network Applications 19 (2014) 171?209, at http://www.ece.ubc.ca/~minchen/min_paper/BigDataSurvey2014.pdf Clarke R. (1991) 'A Contingency Approach to the Application Software Generations' Database 22, 3 (Summer 1991) 23 - 34, PrePrint at http://www.rogerclarke.com/SOS/SwareGenns.html Clarke R. (1992) 'Extra-Organisational Systems: A Challenge to the Software Engineering Paradigm' Proc. IFIP World Congress, Madrid, September 1992, PrePrint at http://www.rogerclarke.com/SOS/PaperExtraOrgSys.html Clarke R. (2009) 'Privacy Impact Assessment: Its Origins and Development' Computer Law & Security Review 25, 2 (April 2009) 123-135, at http://www.rogerclarke.com/DV/PIAHist-08.html Clarke R. (2015) 'Quasi-Empirical Scenario Analysis and Its Application to Big Data Quality' Proc. 28th Bled eConference, Slovenia, 7-10 June 2015, PrePrint at http://www.rogerclarke.com/EC/BDSA.html Clarke R. (2016a) 'Big Data, Big Risks' Information Systems Journal 26, 1 (January 2016) 77-90, PrePrint at http://www.rogerclarke.com/EC/BDBR.html Clarke R. (2016b) 'Quality Assurance for Security Applications of Big Data' Proc. European Intelligence and Security Informatics Conference (EISIC), Uppsala, 17-19 August 2016, PrePrint at http://www.rogerclarke.com/EC/BDQAS.html Clarke R. (2017) 'Big Data Prophylactics' Chapter 1 in Lehmann A., Whitehouse D., Fischer-Hübner S., Fritsch L. & Raab C. (eds.) 'Privacy and Identity Management. Facing up to Next Steps' Springer, 2017, pp. 3-14, at http://www.rogerclarke.com/DV/BDP.html Clarke R. (2018) 'Guidelines for the Responsible Application of Data Analytics' Forthcoming, Computer Law & Security Review 34, 3 (Jul-Aug 2018), PrePrint at http://www.rogerclarke.com/EC/GDA.html De Mauro A., Greco M. & Grimaldi M. (2015) 'What is big data? A consensual definition and a review of key research topics ' Proc. AIP 1644 (2015) 97 (2015), at http://big-data-fr.com/wp- content/uploads/2015/02/aip-scitation-what-is-bigdata.pdf Elragal A. & Klischewski R. (2017) 'Theory-driven or process-driven prediction? Epistemological challenges of big data analytics' Journal of Big Data 4. 19 (2017), at https://journalofbigdata.springeropen.com/articles/10.1186/s40537-017-0079-2 Fayyad U., Piatetsky-Shapiro G. & Smyth P. (1996) 'The KDD Process for Extracting Useful Knowledge from Volumes of Data' Commun. ACM 39, 11 (November 1996) 27-34 Fischer-Hübner S. & Lindskog H. (2001) ' Teaching Privacy-Enhancing Technologies' Proc IFIP WG 11.8 2nd World Conference on Information Security Education, Perth, Australia (2001) Gregor S. & Hevner A. (2013) 'Positioning Design Science Research for Maximum Impact' MIS Quarterly 37, 2 (June 2013 ) 337-355, at https://ai.arizona.edu/sites/ai/files/MIS611D/gregor-2013-positioning- presenting-design-science-research.pdf Guston D.H. & Sarewitz D. (2001) 'Real-time technology assessment' Technology in Society 24 (2002) 93- 109, athttp://archive.cspo.org/documents/realtimeTA.pdf Han J., Kamber M. & Pei J. (2011) 'Data mining: concepts and techniques' Morgan Kaufmannn, 3rd Edition, 2011 Hastie R. & O’Donnell A. (2017) 'Training Pack: Responsible Data Management' Oxfam, August 2017, at https://protection.interaction.org/2004-2/ Hevner A.R. (2007) 'A three cycle view of design science research' Scandinavian Journal of Information Systems 19, 2 (2007) 4 Hevner A.R., March S.T., Park J. & Ram S. (2004) 'Design science in information systems research' MIS Quarterly 28, 1 (2004) 75-105 Huang T., Lan L., Fang X., Min J. & Wang F. (2015) 'Promises and Challenges of Big Data Computing in Health Sciences' Big Data Research 2, 1 (March 2015) 2-11 Jagadish H.V., Gehrke J., Labrinidis A., Papakonstantinou Y., Patel J.M., Ramakrishnan R. & Shahabi C. (2014) 'Big data and its technical challenges' Communications of the ACM 57, 7 (July 2014) 86-94 Kandel, S., Heer, J., C. Plaisant, C., Kennedy, J., van Ham, F., Henry-Riche, N., Weaver, C., Lee, B., Brodbeck, D. & Buono, P. (2011) 'Research directions for data wrangling: visualizations and transformations for usable and credible data' Information Visualization 10. 4 (October 2011) 271-288, at https://idl.cs.washington.edu/files/2011-DataWrangling-IVJ.pdf Knight W. (2017) 'The Dark Secret at the Heart of AI' 11 April 2017, MIT Technology Review https://www.technologyreview.com/s/604087/the-dark-secret-at-the-heart-of-ai/ Koronios A., Gao J. & Selle S. (2014) 'Big Data Project Success ? A Meta Analysis' Proc. PACIS 2014, p.376, at http://aisel.aisnet.org/pacis2014/376 Laney D. (2001) '3D Data Management: Controlling Data Volume, Velocity and Variety' Meta-Group, February 2001, at http://blogs.gartner.com/doug-laney/files/2012/01/ad949-3D-Data-Management- Controlling-Data-Volume-Velocity-and-Variety.pdf McAfee A. & Brynjolfsson E. (2012) 'Big Data: The Management Revolution' Harvard Business Review (October 2012) 61-68 Marchand D.A. & Peppard J. (2013) 'Why IT Fumbles Analytics' Harv. Bus. Rev. (Jan-Feb 2013) 2-9 Mayer-Schonberger V. & Cukier K. (2013) 'Big Data: A Revolution That Will Transform How We Live, Work and Think' John Murray, 2013 Müller H. & Freytag J.-C. (2003) 'Problems, Methods and Challenges in Comprehensive Data Cleansing' Technical Report HUB-IB-164, Humboldt-Universität zu Berlin, Institut für Informatik, 2003, at http://www.informatik.uni-jena.de/dbis/lehre/ss2005/sem_dwh/lit/MuFr03.pdf OT (2013) 'Definitions of Big Data' OpenTracker.com, apparently of April 2013, at http://www.opentracker.net/article/definitions-big-data Pääkkönen P. & Pakkala D. (2015) 'Reference Architecture and Classification of Technologies, Products and Services for Big Data Systems' Big Data Research 2, 4 (2015) 166-186, at https://www.sciencedirect.com/science/article/pii/S2214579615000027/pdfft?md5=5af642001f750f63eaeb 3fbe35f5ccdc&pid=1-s2.0-S2214579615000027-main.pdf Park S.-H., Huh S.-Y., Oh W. & Han A.P. (2012) 'A Social Network-Based Inference Model for Validating Customer Profile Data' MIS Quarterly 36, 4 (December 2012) 1217-1237, at http://www.is.cityu.edu.hk/staff/sangphan/mobility/papers/BI_MISQ_%5BHAN,%2520Sang%2520Pil%5 D.pdf Pathirage M. & Plale B. (2015) 'Fast Data Management with Distributed Streaming SQL' arXiv preprint arXiv:1511.03935, at http://arxiv.org/abs/1511.03935 Peffers K., Tuunanen T., Rothenberger M. & Chatterjee S. (2007) 'A design science research methodology for information systems research' Journal of Management Information Systems, 24, 3 (2007) 45-77, at http://www.sirel.fi/ttt/Downloads/Design%20Science%20Research%20Methodology%202008.pdf Phillips-Wren G., Iyer L.S., Kulkarni U. & Ariyachandra T. (2015) 'Business Analytics in the Context of Big Data: A Roadmap for Research' Commun. AIS 37, 23 Provost F. & Fawcett T. (2013) 'Data Science for Business' O'Reilly, 2013 Rahm E. & Do H.H. (2000) 'Data cleaning: Problems and current approaches' IEEE Data Eng. Bull., 2000, at http://dc-pubs.dbs.uni-leipzig.de/files/Rahm2000DataCleaningProblemsand.pdf Schroeck M., Shockley R., Smart J., Romero-Morales D. & Tufano P. (2012) `Analytics : The real world use of big data' IBM Institute for Business Value / Saïd Business School at the University of Oxford, October 2012, at http://www.ibm.com/smarterplanet/global/files/se sv_se intelligence Analytics_- _The_real-world_use_of_big_data.pdf Turnbull M. (2014) 'Benefiting from big data the government's approach' Address by the Minister for Communications (and soon afterwards, the Prime Minister) to the Australian Information Industry Association, April 2014, at https://www.malcolmturnbull.com.au/media/benefiting-from-big-data-the- governments-approach Ward J.S. & Barker A. (2013) 'Undefined By Data: A Survey of Big Data Definitions' arXiv, 2013, at https://arxiv.org/pdf/1309.5821 Wigan M.R. & Clarke R. (2013) 'Big Data's Big Unintended Consequences' IEEE Computer 46, 6 (June 2013) 46 - 53, PrePrint at http://www.rogerclarke.com/DV/BigData-1303.html Wright D. & De Hert P. (eds) (2012) 'Privacy Impact Assessments' Springer, 2012 Wright D. & Raab C.D. (2012) 'Constructing a surveillance impact assessment' Computer Law & Security Review 28, 6 (December 2012) 613-626 Appendix: Guidelines for Responsible Application of Data Analytics 1. GENERAL DO's 1.1 - Governance Ensure that a comprehensive governance framework is in place prior to, during, and for the relevant period after data acquisition, analysis and use activities, that it is commensurate with the activities' potential impacts, and that it encompasses: a) risk assessment and risk management from the perspectives of all affected parties b) express assignments of accountability, at an appropriate level of granularity 1.2 - Expertise Ensure that all individuals participating in the activities have education, training, and experience in relation to the real-world systems about which inferences are to be drawn, appropriate to the roles that they play 1.3 - Compliance Ensure that all activities are compliant with all relevant laws and established public policy positions within relevant jurisdictions, and with public standards of behaviour 2. DATA ACQUISITION DO's 2.1 - The Problem Domain Understand the real-world systems about which inferences are to be drawn and to which data analytics are to be applied 2.2 - The Data Sources Understand each source of data, including: a) the data's provenance b) the purposes for which the data was created c) the meaning of each data-item at the time of creation d) the data quality at the time of creation e) the data quality and information quality at the time of use 2.3 - Data Merger If data is to be merged from multiple sources, assess the compatibility of the various collections, records and items of data, taking into account the data's provenance, purposes, meaning and quality, and the potential impact of mis-matching and mistaken assumptions 2.4 - Data Scrubbing If data is to be scrubbed, cleaned or cleansed, assess the reliability of the processes for the intended purpose and the potential impacts of mistaken assumptions and erroneous changes 2.5 - Identity Protection If the association of data with an entity is sensitive, apply techniques to the data whose effectiveness is commensurate with the risks to those entities, in order to ensure pseudonymisation (if the purpose is to draw inferences about individual entities), or de-identification (if the purpose is other than to draw inferences about individual entities) 2.6 - Data Security Minimise the risks arising from data acquisition, storage, access, distribution and retention, and manage the unavoidable risks DON'Ts 2.7 - Identifier Compatibility Don't merge data-sets unless the identifiers in each data-set are compatible with one another at a level of reliability commensurate with the potential impact of the inferences drawn 2.8 - Content Compatibility Don't merge data-sets unless the reliability of comparisons among the data-items in the sources reaches a threshold commensurate with the potential impact of the inferences drawn 3. DATA ANALYSIS DO's 3.1 - Expertise Ensure that all staff and contractors involved in the analysis have: a) appropriate professional qualifications b) training in the specific tools and processes c) sufficient familiarity with the real-world system to which the data relates and with the manner in which the data purports to represent that real-world system d) accountability for their analyses 3.2 - The Nature of the Tools Understand the origins, nature and limitations of data analytic tools that are considered for use 3.3 - The Nature of the Data Processed by the Tools Understand the assumptions that data analytic tools make about the data that they process, and the extent to which the data to be processed is consistent with those assumptions. Important areas in which assumptions may exist include: a) the presence of values in relevant data-items b) the presence of only specific, pre-defined values in relevant data-items c) the scales against which relevant data-items have been measured d) the precision with which relevant data-items have been expressed 3.4 - The Suitability of the Tool and the Data Demonstrate the applicability of each particular data analytic tool to the particular data that it is proposed be processed using it DON'Ts 3.5 - Inappropriate Data Don't apply data analytics unless the data satisfies threshold tests commensurate with the potential impact of the inferences drawn, in relation to data quality, internal consistency, and reliable correspondence with the real-world systems about which inferences are to be drawn 3.6 - Humanly-Understandable Rationale Don't apply an analytical tool that lacks transparency, by which is meant that the rationale for inferences that it draws is expressible in humanly-understandable terms 4. USE OF THE INFERENCES DO's 4.1 - The Impacts Understand the potential negative impacts on stakeholders of reliance on the inferences drawn, taking into account the quality of the data and the data analysis process 4.2 - Evaluation Where decisions based on inferences from data analytics may have material negative impacts, evaluate the advantages and disadvantages of proceeding, by conducting cost-benefit analysis and risk assessment from an organisational perspective, and impact assessments from the perspectives of other internal and external stakeholders 4.3 - Reality Testing Test a sufficient sample of the results of the analysis against the real world, in order to gain insight into the reliability of the data as a representation of relevant real-world entities and their attributes 4.4 - Safeguards Design, implement and maintain safeguards and mitigation measures, together with controls that ensure the safeguards and mitigation measures are functioning as intended, commensurate with the potential impacts of the inferences drawn 4.5 - Proportionality Where specific decisions based on inferences from data analytics may have material negative impacts on individuals, consider the reasonableness of the decisions prior to committing to them 4.6 - Contestability Where actions are taken based on inferences drawn from data analytics, ensure that the rationale for the decisions is transparent to people affected by them, and that mechanisms exist whereby stakeholders can access information about, and if appropriate complain about and dispute interpretations, inferences, decisions and actions 4.7 - Breathing Space Provide stakeholders who perceive that they will be negatively impacted by the action with the opportunity to understand and to contest the proposed action 4.8 - Post-Implementation Review Ensure that actions and their outcomes are audited, and that adjustments are made to reflect the findings DON'Ts 4.9 - Humanly-Understandable Rationale Don't take actions based on inferences drawn from an analytical tool in any context that may have a material negative impact on any stakeholder unless the rationale for each inference is readily available to those stakeholders in humanly-understandable terms 4.10 - Precipitate Actions Don't take actions based on inferences drawn from data analytics until stakeholders who perceive that they may be materially negatively impacted by the action have had a reasonable opportunity to understand and to contest the proposed action. Denial of a reasonable opportunity is only justifiable on the basis of emergency, as distinct from urgency or mere expediency or efficiency. Where a reasonable opportunity is not provided, ensure that stringent safeguards, mitigation measures and controls are designed, implemented and maintained in relation to justification, reporting, review, and recourse in the case of unjustified or disproportionate actions 4.11 - Automated Decision-Making Don't delegate to a device any decision that has potentially harmful effects without ensuring that it is subject to specific human approval prior to implementation, by a person who is acting as an agent for the accountable organisation Examining the PoC System Implementation and Adoption: A FVM Perspective IMRAN MUHAMMAD & NILMINI WICKRAMASINGHE 8 CORRESPONDENCE ADDRESS: Imran Muhammad, Ph.D., Lecturer,Deakin University, Faculty of Health, L4, BC Building, 221 Burwood Highway, Burwood, Victoria 3125 Australia, e-mail: imran.muhammad@deakin.edu.au. Prof. Nilmini Wickramasinghe, Ph.D., Professor Health Informatics Management, Deakin University, Faculty of Health, L4, BC Building, 221 Burwood Highway, Burwood, Victoria 3125 Australia, e-mail: imran.muhammad@deakin.edu.au. DOI https://doi.org/10.18690/978-961-286-170-4.7 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract HIS implementation is complex and involves people issues as well as technological issues. The effect of sociotechnical issues such as macro level or external factors including political, social, economic, environmental infrastructure and technology, laws and regulations; meso level or organizational factors such as leadership, management style, policies, structure; and micro level or tactical factors such as information sharing, training and learning, technical staff or user behaviour, have been less widely studied. Yet, it is precisely these issues that separately or in combination derail numerous HIS implementations. To examine this dilemma, we proffer a unique application of the fit viability model (FVM) to facilitate a better understanding of key issues pertaining the implementation and adoption of a Point of Care (PoC) System at one of the not for profit private hospitals in Australia. This will help the decision makers in hospital to understand how the new system fits within the different departments and also is it a viable option to install such a new system. This study focuses on just two departments of the hospital; namely, food services and environment services. An exploratory single case qualitative study methodology is adopted. From such an analysis, it is possible to identify optimal aspects with the PoC solution and opportunities to add value. Keywords: • Point of Care Systems • Fit-Viability Model • PoC • FVM • 1 Introduction Information technology is an important part of almost every industry in the developed world. Businesses are expanding their boundaries by using integrated and collaborative IT solutions. Due to an ever-increasing need for effective and efficient healthcare services and delivery, healthcare organisations are now trying to find integrated solutions for their business needs to automate their processes to gain a strategic advantage (Ives and Jarvenpaa 1991). In addition, the complex nature of healthcare services is compelling and forcing healthcare organizations to adopt best available technologies, but the problem here organizations are facing is that they need to adopt technologies according to their requirements and best fit having considered their environment, infrastructure, government regulations, and scope of their business, availability of funds and local culture and norms (Ignatiadis and Nandhakumar 2007). Healthcare organisations are inclusive organisations, involving different stakeholders, partners, customers and suppliers from different cultures and different systems; they need more sophisticated means of communication and interaction (Harris, Moran, & Moran, 2004). Since the last decade of the last century, we have witnessed a significant growth in the adoption rate of information technology and a big shift towards the deployment of Health Information Systems (HIS) (Muhammad et al 2014). Most countries have responded positively to the changing global market place and are reaping the strategic benefits by implementing HIS (Mukesh and Betsy 2009). HIS implementation is a difficult and complex decision and system implementation is not limited to just installing software; ie, it is much more than a technology adoption. Rather, it involves people issues more than technological issues (Cresswell, Worth, & Sheikh, 2011). Research indicates that people issues are more to blame for the unsuccessful efforts of eHealth implementations (Mukesh and Betsy 2009). The effect of sociotechnical issues such as macro level or external factors including political, social, economic, environmental infrastructure and technology, laws and regulations; meso level or organizational factors such as leadership, management style, policies, structure; and micro level or tactical factors such as information sharing, training and learning, technical staff or user behavior have been less widely studied (Nguyen et al., 2015). Yet, it is precisely these issues that separately or in combination derail numerous HIS implementations (Nguyen et al., 2015). To examine this dilemma, we proffer a unique application of the fit viability model (FVM) (Liang, Huang, Yeh, & Lin, 2007) to facilitate a better understanding of key issues pertaining the implementation and adoption of a Point of Care (PoC )System at one of the not for profit private hospitals in Australia. In so doing, we answer the research question: “How can a FVM assist in unpacking the varied sociotechnical issues in the adoption and implementation of PoC system?”. This will help the decision makers in hospital to understand how the new system fits within the different departments and also is it a viable option to install such a new system. This study deals with just two departments of the hospitals namely, food services and environment services. An exploratory single case study methodology is adopted. From such an analysis, we believe it will be possible to identify optimal PoC solutions and opportunities to add value. 2 Literature Review Many healthcare information systems have been implemented around the globe with mixed results, despite the claims that HIS can play a significant role in efficiency and effectiveness of healthcare service delivery (Muhammad et al 2014) The literature provides evidence of failed clinical systems and lack of adoption by users (DesRoches et al., 2008; Protti et al., 2009). Challenges and barriers to implementation and adoption of bedside PoC systems in hospital wards have been extensively debated (Brailer, 2005; Choi et al., 2004; Yao, Schmitz, & Warren, 2005; Nguyen et al., 2015). Researchers have divided these barriers into different categories ranging from environmental, social, technical and Organisational (André et al., 2008). These factors can play a very crucial role in the decision-making process of technology adoption (Huang & Palvia, 2001). In a healthcare service context, where organisations are now required to work as a networked framework, health information technology implementation and adoption would be a more complex and challenging endeavor because of the different business processes, the available infrastructure, compatibility issues, decision centres, authorization mechanisms and hierarchies, enterprise systems and data semantics (Avgerou, 2008; Liu et al., 2011; Trudel, 2010). IT implementations can cause serious disruptions in service deliveries and in result, at productivity and healthcare services are one of the very critical areas of services that cannot afford disruptions (Kralewski et al., 2010; Scott et al., 2005). There are many organisational barriers to the implementation and the adoption of eHealth technologies, for example, poor governance, organisational culture and proper management of the change process that could harm the flow of transformation (Greenhalgh & Stones, 2010; Kennedy, 2011). These issues can aggravate the resistance to the change process and complicate the dissemination of the eHealth technology. Technological issues can also exacerbate the resistance to the adoption of health information technology (Muhammad et al 2014). The lack of infrastructure, and standards results in a fragmentation of healthcare information systems and this contributes to creating a very complex situation for coordination (Kennedy, 2011; Trudel, 2010). Pre-implementation and post-implementation vendor support is another key concern for organisations (Kennedy 2011; Liu et al. 2011). Lack of technical resources and experience with information technology implementation within healthcare settings are other problems faced by many (Trudel 2010; Liu et al. 2011; Kennedy 2011). People issues, ranging from user acceptance (Trudel, 2010), perceived ease of use (Al-Azmi et al., 2009), lack of knowledge about the system (André et al., 2008; Liu et al., 2011), lack of training, lack of stakeholder consultation (Showell, 2011), lack of willingness to assimilate the technology in to daily routines and processes (Greenhalgh & Stones, 2010) , conflict between system and user embedded values (Greenhalgh & Stones, 2010), complex and complicated user interfaces (Yusof et al. 2007), conflict between physician activities and training schedules (André et al. 2008; Yusof et al. 2007;) and complications in patient-provider communications are some of the major concerns. Further, it is paramount that the systems are user centric and have a good fit with user values as well as existing healthcare systems (Liang, Huang, Yeh, & Lin, 2007). 2.1 Fit-Viability Model As Tjan (2001) proposed fit viability dimensions for evaluating Internet initiative projects. Liang and Wei (2004), by taking these two dimensions and adding Task Technology Fit (TTF) theory, proposed a fit-viability model to study m-commerce applications. In their framework, viability measures the readiness of the organization for the technology adoption and implementation, and fit measures capabilities of the systems to optimally perform the required tasks. These two dimensions make a simple matrix with fit on horizontal and viability on vertical axis. By using the four corners of the matrix, organizations can make an informed decision for technology adoption and implementation. 2.2 Task-Technology Fit The theoretical basis of the fit construct is derived from the Task-Technology Fit model which according to Goodhue (1995; 1998) argues that a fit between task characteristics and system features need to be high for the better performance and success and this will have effect on the decision-making process of an organization. Research (Madapusi 2008; Soh et al. 2000) has indicated that if a system is more aligned with the requirements of the users there are greater chances of system success which leads to better performance. It means that if the features offered by the system fit with the task requirements the users will be more incline to use it. 2.3 Viability Viability refers to the degree of impact of environment and organizational factors on a system adoption and implementation decision. These factors at the macro level include political and social, economic, environmental as well as infrastructure/technology factors. At the organizational level literature has proposed many factors at the strategic and tactical levels (Umble et al., 2003). These factors include leadership, management style, polices, information sharing, training and learning, technical staff, and user behaviour. Taking the example of PoC, economic and technological factors are crucial factors in HIS system implementations; and ignoring these factors could lead to unsuccessful projects. Management support, physical and IT infrastructure create stronger desire of system implementation and innovation adoption, that positively impact viability of the system. 3 Research Framwork The research framework shown in figure 1 illustrates is used to identify the key constructs and factors affecting PoC system implementations. The PoC is a patient bedside solution that can be accessed across the hospital by clinicians and also has a patient portal component. It is not an EMR (electronic medical record) but has many features and capabilities similar to an EMR. It also has a patient entertainment component. The PoC system was implemented into the not-for private tertiary hospital system (the chosen case study) gradually starting in late 2015. Primary objectives for its implementation included to enhance the patient experience and provide value-based patient-centred care. The system has many modules of which this paper focuses on catering/food services and environmental services. Fit is measured by matching the requirements of the organization with the functionalities offered by the PoC system e.g. data format, operating procedures, and output format while viability is measured by assessing the impact of national and organizational factors on the adoption decision of the organization and individual user adoption. Considering that the PoC has many similar factors to other e-health solutions such as political and sociotechnical factor identified by (Muhammad et al 2016) in evaluation of MyHealth record in Australia (the Australian national e-health solution) and smart card solution in Germany (the German national e-health solution), thus it is logical to use these factors as the basis of the model. This conceptual model serves to capture the important aspects of the barriers and facilitators for the prediction of the successful adoption and implementation of the PoC. The proposed model identifies a network of different actors interconnected to each other. It further illustrates that a central issue with the evaluation of IT based healthcare is influenced by the complexity of the evaluation objects and includes both social and technical considerations (Greenhalgh & Stones, 2010). For instance, the nature of the integration of healthcare information systems with the culture and business processes of healthcare organisations puts more emphasis on the evaluation methods and goes beyond the technology aspects of hardware and software, furthermore, external and internal environmental factors as well as an understanding of the diverse nature of system effects in the healthcare settings is required (Greenhalgh & Stones, 2010). This emphasis is on creating a better fit between human, contextual and technological factors for the successful implementation and adoption of health information systems (Yusof et al., 2008). Figure 1: FVM 4 Methodology Based on the criteria given by Yin (2010); the appropriate choice of methodology to test the use and usability of the proposed framework (figure 1) is a qualitative case study research because this is an exploratory study of a new phenomenon of bedside PoC information system’s implementation. Further, we wish to explore how the PoC solution at different sites of the hospital can be implemented successfully and what are the factors that impact on the implementation and adoption of this HIS based intervention. Qualitative research is holistic, humanistic, and interactive, it can provide more support to focus on the study of a complex phenomenon of human and system interaction and relationship; as in our case multisite bedside PoC system implementations. Qualitative research can provide deeper understanding of the phenomenon as compared to quantitative study because of the exploratory nature of the study and focus which would not be on quantitative measures (Yin, 2010). For this study, several archival records and documents relating to the health information and communication technologies implementation and adoption in healthcare service delivery settings along with hospital and OneView reports and evaluations were critically analysed. These documents were of great value in developing an understanding of the need for a PoC system and factors important for the implementation and adoption of this system. This analysis assisted in developing the theoretical re-search framework and in planning the primary data collection strategies for the larger study. A priori themes were developed through a pilot study and then literature was analysed using thematic analysis and hermeneutic analysis (Boyatziz, 1998; Kvale, 1996), then we performed a gap analysis. The analysis led us towards the development of FVM for this study. 5 Analysis and Results Data for this piolet study was collected using unstructured interview and analysis of archival material. As far as possible, the multiple stakeholder views (i.e.; Service providers, regulator, payer, Food and environment services, and hospital and patient perspectives) in healthcare were captured. Data analysis included standard qualitative techniques such as thematic analysis where a priori themes were derived from the components in the conceptual model. The case studies all exemplify various aspects of the proffered conceptual model in this way serving to validate the model and demonstrate its usefulness in unpacking critical aspects with HIS implementation. 5.1 The context before and after OneView PoC system This section presents a process map for two departments of the hospital before and after the implementation of the OneView PoC system. It must be noted here that prior to the PoC system, there was another system known as Infotainment. This system had many problems concerning mainly to poor technical support, frequent hardware and software failures, smaller screens, and usability issues especially with the user interface and the biometrics registries. That systems really did not fit with in the hospital quality services environment thus the change was impending. This section will explain the process before and after OneView PoC system within two departments; namely Food Services and Environmental Services. These departments are chosen as they are fully transitioned to the new PoC system now. 5.1.1 The Context of Food Services Prior to the current PoC system, the processes in the Food Services were facilitated using an application called Delegate. This application had been in use since the opening of Epworth Eastern and has been replaced by OneView PoC system. Hence, the old PoC system was not used by the staff from Food Services. 5.1.2 The Context of Food Services Prior to OneView PoC system Managing patients’ meal orders was handled by the Department of Food Services. The contact point was the Menu Monitors, who take the meal orders 24 hours prior to the actual delivery of meal. A computerized system called Delegate Prior to OneView PoC system was used for the order and delivery services. Delegate system was installed on a number of computers on wheels (mainly laptops). Taking orders and delivery meals to the patient includes discussing the options and personal preferences for the breakfast, lunch, and dinner for the next day. This process would usually consume nearly 70% of the Menu Monitor’s work time with a rate of 14 patients per hour. That is 3-4 minute per patient. Patients can choose three full courses for the next day from a menu that changes every three years hospital wide. The gathered information about the meal preferences was then entered by the Menu Monitors into Delegate, and then spread sheets were printed off this system to circulate to kitchens. These sheets were then assigned to different chefs as ‘Production Lists’. Throughout the day, three production lists were prepared, one in the morning (7:00 am), the second is after breakfast (10:00 am), and the third is at 3:30 PM for the evening as Figure 2 below illustrates. Two main issues were found in this way of handling patients’ meals. The first relates to the information collection, and the second issue about Delegate performance and functionality. Although preparing the production lists was time consuming and required loads of time for the interaction between patients and menu monitors, still four types of information were not possible to capture using Delegate and the process around it. Those are: I. Late orders: Meal orders for patients admitted after 8 pm during week days and after 7 pm during weekends were not attended. This group of patients did not had a choice of food for the first day of their admission because they could not see the menu monitors. Rather, they would have the default meal for the day. II. Food allergy for visited patients: Delegate did not had any capability to record any food allergy patients may have. This piece of information normally comes from iPM. In many cases this information was then not passed on to the chefs, so they make meals without taking that into consideration. III. Discharged people: Depending on the time patients are discharged, many cases reportedly happened where meal orders were made, but patients had been discharged. Again, this piece is coming from iPM, and not passed on to the Food Services staff at the right time. IV. People with changed diet codes: As treatment plans progress, patients may change their diet codes, such as changing from ‘not eating’ to ‘eating’ and from ‘soft’ to ‘hard’ food. These changes were also managed by nurses using iPM, however, the co-operation between nurses and Food Services was not maintained at all times, which resulted in many cases meals were not made according to these changes. Figure 2: A Flow chart for ordering meals before PoC These issues had direct and indirect impacts on the cost and quality of provided services as Table 1 below summarizes. Table 1: Issues resulting from the conventional way to handle meal orders prior to OneView PoC system Issue Impact on Cost Impact on quality Late orders None, as patients receive the default meal for the next day Negative impact on patient satisfaction Food allergy for visited patients Wasted food, unplanned care for the resulting allergy which may result in implications on patients’ insurance cover Negative impact on patient satisfaction and safety as well as trust in the hospital Discharged patient Wasted food More pressure on chefs Patients with changed diet codes Wasted food Negative impact on patient satisfaction Delegate has a number of issues as the interviewees from the Food Services agreed. These issues include: I. It was perceived to be a slow system. II. System was not really fit for the task. III. Long term viability of the system was really doubtful even though infrastructure and management support were available. IV. The system had many technical problems and fixing these problems took long times as described by this interviewee: “If there is some technical problems with Delegate this would take long time to fix. I mean when the system goes down it really goes down and takes a while to fix”. V. As mentioned before, Delegate did not have the capability to capture food allergy data, which forced the staff at the Food Services to look up these data from iPM. This was not possible all the time, which resulted in a number of cases where patients had food that they were allergic to, consequence this had negative impact on patient satisfaction and healthcare services and delivery of the hospital. While some issues with Delegate had negative impacts on the users, thus they became dissatisfied about the system; but few problems were really serious problem with real consequences on patient’s health and safety. Not only did it negatively affect patient satisfaction and experience, but it also represented a real risk factor and caused safety issues to the patients and the hospital at the same time. 5.1.3 The Context of Food Services After OneView PoC system Using OneView PoC system, patients can place their orders of meals through their user interface. The arrival of this function to the PoC system has partially solved the issues faced by the conventional system; namely late patients (after 8 pm weekdays and after 7:00 pm weekends), allergy data, and patients with changed diet coeds. Late patients can order their favorite meals for the next day if they want to, they can state their allergy status through the admission form, and nurses can change patients’ diet codes right from OneView PoC system. Currently, no more than 10% of the patients are using the PoC system to order their meals. On asking on the reasons behind that, two main reasons were identified. The first is some issues with the user interface, especially with elderly patients, as patients need to scroll down to the bottom of the screen to reach the meal ordering function. During the scrolling down, a number of pop-ups will appear and may disrupt their endeavor: “We go up and introduce the system to them. When you go into the system at the minute, the way they implemented it, it is hard to use. You’ve got to scroll down to find the ordering”. Figure 3: A Flow chart for ordering meals after PoC The other reason relates to the conceptual resistance by some patients to rely on a ‘machine’ to order their meals, preferring human-human interaction more than human-machine interaction. This was agreed upon by both interviewees from the Food Services. As a result of being in a hybrid environment, i.e. a minority of patients are using the PoC system to order their meals, and the majority are still using conventional way to order their meals, Delegate now has colored dots to indicate the patients who used the PoC system to order their meals. Hence, the menu monitors do not need to visit them to organize their meals. With the expected increase of the uptake of this function in the PoC system, more patients will adopt this function, and more time required to see patients by their menu monitors will be freed-up. This is expected to have positive impacts on the hospital and its patients. In addition, it is expected that there will be a significant saving regarding food wastage. 5.2 The Context of Environmental Services The Environmental Services is a dedicated team whose two main responsibilities are 1) providing all types of cleaning (steam cleaning, buffer cleaning, advanced cleaning, curtains cleaning, etc.), and 2) patient transportation. In terms on human resources, the Environmental Services team comprises about 60 staff. Of this figure, about 40 works in cleaning, and the reminder work in patient transportation. Unlike the Food Services, which had Delegate as a computerized system to facilitate food-related processes prior to the OneView PoC system, the operations of the Environmental Services at the hospital were mainly based on phone to phone and face to face communications. Introducing the OneView PoC system and integrating the Room Ready Module into it has made considerable change in the processes of this vital department. The following is a summary of the process map of the Environmental Services at the hospital before and after OneView PoC system. 5.2.1 The Context of Environmental Services before OneView PoC system The process of performing jobs by the Environmental Services before the PoC had three main steps. 1) Initiating the job by nurses and specific cleaners; 2) receiving job orders by the supervisors within the Environmental Services; and 3) assigning tasks to cleaners as Figure 4 depicts. Figure 4: A map for the Environmental Services before OneView PoC Nurses normally initiate job orders when needed. This includes preparing rooms before admitting new patients to these rooms, cleaning rooms after patients have been discharged, and as needed if a patient had an incident such as bleeding or vomiting. These job orders go to the supervisors from the Environmental Services using face to face or phone to phone communications tools. The supervisors in turn convey these orders to the cleaners across the hospitals using same communication means, i.e. phone and face to face. Apart from nurses, buffer cleaners and permanent cleaners can initiate job orders if need be. This normally happens when one of these cleaners realizes, while doing their jobs, there exist some curtains or carpet need to be cleaned, and they do not have the required equipment to do so. Once the need of a cleaner has been established, the path of this order in dependent upon its urgency. If the job was of a higher urgency, then these cleaners inform their supervisors either by phone calls or by hand-written notes. Then the supervisors assign the tasks to different cleaners on floor accordingly. If the job was not considered urgent by these cleaners, then they will wait till the next shift of cleaners has come to do the job based on hand written notes, which caused extended times to do specific jobs. The level of urgency was left to the cleaners to decide. This system had caused many problems, which can be summarized as follows: 1. Over relaying on human factors: As can be seen, the whole processes of the Environmental Services team are centered on human communications and judgment. This resulted in many cases where jobs were not done or took longer time to be addressed. This is particularly acute in the cases where patients had to wait outside their rooms while cleaning the room, which resulted in unsatisfied patients and nurses. 2. Lack of co-ordination in the multi levels of communications: Given that nurses, supervisors, and permanent and buffer cleaners could initiate job orders, cleaners in many cases were confused about their tasks and what tasks had higher priorities. This lack of coordination was due to adopting phone calls, hand-written notes, and face to face means. 3. Inability to address language barriers: Given that a considerable portion of the cleaners had language barriers, their understanding of their assigned tasks over the phone or hand-written notes was reportedly limited in many cases, which caused many jobs not done properly. 4. More effort by cleaners: As the job orders did not have enough information about their jobs (locations, level of urgency, and required equipment), the cleaners had to go to the site to manually collect all of this information and come back to their workplace to collect the right equipment for that specific job. This caused them to walk back and forth many times, which reduced their productivity and the quality of their jobs, and increased their fatigue. 5. Lack of accountability: As the majority of the needed jobs were verbally conveyed from one stakeholder to another, the possibility of creating accountability and tracking the performance of different units and individuals were almost impossible. This is especially acute in the case of cleaners with limited literacy, as well as the communications between nurses (Clinical Services) and the cleaners and their supervisors as Figure 4 depicts. These issues had direct and indirect impacts on the cost and quality of provided services as Table 2 below summarizes Table 1: Issues resulting from the conventional way to handle environmental services prior to PoC system Problems Impact on Cost Impact on Quality Over relaying on Human Factors Double work resulted in many cases, which implied extra cost and less quality Lack of Co-ordination: Multi levels of communications Language barriers As many jobs needed to be repeated, double works resulted in many cases, which implied extra cost and less quality More effort by cleaners Given the missing information on the nature of their tasks, cleaners had to survey the location of their jobs in person, which resulted in them walking for extended times/ distances. Due to the unneeded increased workload, the quality of cleaning services was negatively affected Lack of accountability Tracking the performance of individuals and different unit was almost impossible 5.2.2 The Context of Environmental Services after OneView PoC system Introducing the OneView PoC system, particularly the Room Ready Module, has notably streamlined the cleaning related processes of the Environmental Services, while the other vital role of the Environmental Services; namely patient transportation, is still conducted using the conventional way, with a vision to integrate this function into the OneView PoC in later enhancements. The Room Ready Module enables nurses, permanent cleaners, and buffer cleaners to log into the system and place cleaning orders with enough details about the job, its location, requirements and level of urgency. This information is then conveyed to the cleaners on floor as short text messages on their PoC phones. Based on the nature of the jobs, cleaners can choose the jobs of higher urgency, closer to their geographic location, and/ or achievable using their current equipment. This has resulted in saving cleaners’ times and efforts, which in turn has shown faster responses to the cleaning needs initiated by different wards, units, and individuals. Not only has the Room Ready module enhanced quality and productively of cleaners, but it has also resulted in a simpler map of cleaning processes performed by the Environmental Services as Figure 5 depicts. Figure 5: A map for the Environmental Services after OneView PoC The initial findings from the interviews show that the use of OneView PoC system to support the cleaning processes at Epworth Eastern has addressed most of the problems faced by the conventional way to manage the cleaning needs for the hospital. Table 2 summarized these findings: Table 2: The impact of Room Ready Module from the OneView PoC system Problems How has OneView PoC addressed the problems Impact of OneView PoC on Quality Cost Over relaying on Human Factors This problem has been partially solved, as human still need to log in and place job orders. The existence of Room Ready though has increased the ability to place jobs and track them. Positive M Positive M Lack of Co-ordination: The introduction on Room Ready Module has eliminated one layer of communications in the process map of cleaning services. That is the supervisors, which has freed up their time, and has rebuilt their roles around coordinating different tasks and following up with different stakeholders. Positive H Positive M Multi levels of communications Language barriers As job orders come to cleaners in a form of short text messages on their PoC phones, this problem has been partially solved. Positive L Positive L More effort by cleaners As the jobs orders come with a relatively comprehensive set of information, cleaners don’t need to go and assess the job before actually doing the job, which resulted less Positive H Positive M effort from them, and more tasks performed every day than before Lack of accountability All jobs’ orders are now documented and stored in the system. Hence, tracking different jobs and their progress and the responses from different stakeholders is always possible. Positive H Positive H Legends: L: Low, M: Medium, H: High The initial results suggested that the PoC system for food and environment services is fit for task and very much viable to implement. Almost all sites of hospital were ready for implementation. IT Infrastructure was fit for the purpose while physical infrastructure needs changes for one site. Environment was favorable, and project have full support from top management. Appropriate budget was allocated for the project. Our initial analysis suggest that the use of system will be depended on the fitness of the system for the tasks. Users were very positive in its very early stages of deployment. We contend that a large-scale study is prudent to further understand the capacities of implementation and use of this system for clinical services. 6 Conclusion The purpose of this research paper is to conceptualize a framework to investigate the implementation and use viability of a bedside PoC health information and entertainment system for food and environment services for patients in private hospital wards, since we recognize that with e-innovations not only the technology solution is necessary but it is also critical to look at organizational and societal aspects concurrently especially in today’s global business environment. ehealth system implementations comprise of technology as well as human involvement. In view of this, it is important to investigate the impact of political and social, economic, environmental and infrastructure/technology factors on the organizational decision making. We underscore the importance of studying the system viability and fit before making any decision about system adoption and deployment. Technology needs to have a good fit to perform the required tasks, only then can system viability and fit have positive effects on the organization’s performance and that can be measured by system use and user satisfaction and thus ensure the full potential of a solution is realised. The theoretical contribution of the paper is the use of Fit-viability model for the evaluation of health information technology (HIT) implementation in private hospital settings. This model has never been used of HIT implementations, thus we believe the model will present more informed lens for decision makers to understand HIT implementation. The practical implication of the Fit-Viability model presented in this paper can be realised by using the developed model to measure the fitness and viability of implementing health information technology in hospital settings. This should be able to envisage the possibility of success. There are also limitations of this study First, the theoretical generalizability of the FVM needs critical testing in future studies. Second, the findings of this paper are based on just two segments of the hospital namely food services and environmental services, thus more extensive studies including more segments of hospital may be necessary. References Al-Azmi, S. F., Al-Enezi, N., & Chowdhury, R. I. (2009). Users’ attitudes to an electronic medical record system and its correlates: A multivariate analysis. HIM J, 38(2), 33–40. André, B., Ringdal, G. I., Loge, J. H., Rannestad, T., Laerum, H., & Kaasa, S. (2008). Experiences with the implementation of computerized tools in health care units: a review article. Intl. Journal of Human–Computer Interaction, 24(8), 753–775. Avgerou, C. (2008). Information systems in developing countries: a critical research review. Journal of Information Technology, 23, 133–146. doi:10.1057/palgrave.jit.2000136 Boyatzis, R. E. (1998). Transforming Qualitative Information: Thematic Analysis and Code Development (1st ed.). Sage Publications, Inc. Brailer, D. J. (2005). Interoperability: the key to the future health care system. Health Affairs, 24, W5. Cresswell, K., Worth, A., & Sheikh, A. (2011). Implementing and adopting electronic health record systems: How actor-network theory can support evaluation. Clinical Governance: An International Journal, 16(4), 320–336. https://doi.org/10.1108/14777271111175369 Choi, J., Chun, J., Lee, K., Lee, S., Shin, D., Hyun, S., … Kim, D. (2004). MobileNurse: hand-held information system for point of nursing care. Computer Methods and Programs in Biomedicine, 74(3), 245–254. DesRoches, C. M., Campbell, E. G., Rao, S. R., Donelan, K., Ferris, T. G., Jha, A., … Blumenthal, D. (2008). Electronic Health Records in Ambulatory Care -- A National Survey of Physicians. New England Journal of Medicine, 359, 50 – 60. Greenhalgh, T., & Stones, R. (2010). Theorising big IT programmes in healthcare: strong structuration theory meets actor-network theory. Social Science & Medicine (1982), 70(9), 1285–1294. doi:10.1016/j.socscimed.2009.12.034 Huang, Z., & Palvia, P. (2001). ERP implementation issues in advanced and developing countries. Business Process Management Journal, 7(3), 276–284. doi:10.1108/14637150110392773 Ignatiadis, I., and Nandhakumar, j. (2007). “The impact of enterprise systems on organizational resilience”, Journal of Information Technology, 22: 36-43. Ives, B. and Jarvenpaa, S.L. (1991). “Applications of Global Information Technology: Key Issues for Management”, MIS Quarterly 15 vol 1, pp. 33-49 Kvale, S. (1996). InterViews: An Introduction to Qualitative Research Interviewing (1st ed.). Sage Publications, Inc. Kralewski, J., Dowd, B. E., Zink, T., & Gans, D. N. (2010). Preparing your practice for the adoption and implementation of electronic health records. Physician Executive, 36(2), 30–33. Kennedy, B. L. (2011). Exploring the Sustainment of Health Information Technology: Successful Practices for Addressing Human Factors. NORTHCENTRAL UNIVERSITY. pp. 7-17. Liang, T.-P., Huang, C.-W., Yeh, Y.-H., & Lin, B. (2007). Adoption of mobile technology in business: a fit‐viability model. Industrial Management & Data Systems, 107(8), 1154–1169. https://doi.org/10.1108/02635570710822796 Liang, T.P. and Wei, C.P. (2004), “Introduction to the special issue: a framework for mobile commerce applications”, International Journal of Electronic Commerce, Vol. 8 No. 3, Liu, L. S., Shih, P. C., & Hayes, G. R. (2011). Barriers to the adoption and use of personal health record systems. In Proceedings of the 2011 iConference (pp. 363–370). Muhammad, I., & Wickramasinghe, N. (2014). How an Actor Network Theory (ANT) Analysis Can Help Us to Understand the Personally Controlled Electronic Health Record (PCEHR) in Australia. Technological Advancements and the Impact of Actor-Network Theory, 15-34. Mukesh S., Betsy J. G. (2009) “Chinese Cultural Implications for ERP Implementation” Journal of Technology Management & Innovation Volume 4, Issue 1 Nguyen, L., Bakewell, L., Wickramasinghe, N., Haddad, P., Muhammad, I., Moghimi, H., … Botti, M. (2015). Transition from Paper to Electronic Nursing Documentation in Residential Aged Care: an Actor Network Theory Analysis. Electronic Journal of Health Informatics, 9(1), 4. Protti, D., Johansen, I., & Perez-Torres, F. (2009). Comparing the application of Health Information Technology in primary care in Denmark and Andalucía, Spain. International Journal of Medical Informatics, 78(4), 270–283. doi:10.1016/j.ijmedinf.2008.08.002 Scott, J. T., Rundall, T. G., Vogt, T. M., & Hsu, J. (2005). Kaiser Permanente’s experience of implementing an electronic medical record: a qualitative study. BMJ (Clinical Research Ed.), 331(7528), 1313–6. Showell, C. M. (2011). Citizens, patients and policy: a challenge for Australia’s national electronic health record. Health Information Management Journal, 40(2), 39–43. Soh, C., Kien, S. S., and Tay-Yap, J. (2000) “Enterprise Resource Planning: Cultural Fits and Misfits: Is ERP a Universal Solution?” Communications of the ACM, 43 Vol. 4, pp. 47-51. Tjan, A.K. (2001), “Finally, a way to put your internet portfolio in order”, Harvard Business Review, Vol. 79 No. 2, pp. 76-85. Trudel, M. C. (2010). Challenges to personal information sharing in interorganizational settings: Learning from the Quebec Health Smart Card project. THE UNIVERSITY OF WESTERN ONTARIO. Umble, E., Haft, R., and Umble, M. (2003). “Enterprise Resource Planning: Implementation Procedures and Critical Success Factors,” European Journal of Operational Research, 146, Vol. 2 pp. 241-257 Yao, J., Schmitz, R., & Warren, S. (2005). A wearable point-of-care system for home use that incorporates plug-and-play and wireless standards. IEEE Transactions on Information Technology in Biomedicine, 9(3), 363–371. Yusof, M. M., Papazafeiropoulou, A., Paul, R. J., & Stergioulas, L. K. (2008). Investigating evaluation frameworks for health information systems. International Journal of Medical Informatics, 77(6), 377–385. doi:10.1016/j.ijmedinf.2007.08.004 Influence of Digital Coaching on Physical Activity: Motivation and Behaviour of Physically Inactive Individuals TUOMAS KARI & PETRIINA RINNE 9 CORRESPONDENCE ADDRESS: Tuomas Kari, D.Sc., Postdoctoral researcher, University of Jyvaskyla, Faculty of Information Technology, P.O. Box 35 40014 Jyvaskylan yliopisto, Finland, e-mail: tuomas.t.kari@jyu.fi. Petriina Rinne, MSc, University of Jyvaskyla, Faculty of Information Technology, P.O. Box 35 40014 Jyvaskylan yliopisto, Finland, e-mail: petriina.r.rinne@student.jyu.fi DOI https://doi.org/10.18690/978-961-286-170-4.8 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract Digital wellness solutions and their use have become increasingly common. Respectively, the need to understand their users and usage has increased. A general problem with wellness technologies is that they typically provide feedback through numbers and graphs instead of providing actual guidance. One potential and novel solution for this is digital coaching. This study is one of the first to examine digital coaching features and their influence. We define digital coaching in the context of sports and wellness technology and investigate its influence on physical activity motivation and behaviour of physically inactive individuals. The study is based on thematic analysis of 20 semi-structured interviews conducted for 10 participants who used a novel digital coaching solution for one month. The digital coaching solution was found to be beneficial for physical activity and exercise motivation and behaviour. It assisted the users in many ways through different elements and mechanisms. Further, the benefits also spurred to other aspect of physical wellness. The results are discussed and implications provided. Keywords: • Digital Coaching • Digital Coach • eWellbeing • Digital Wellness • Physical Activity • Wellness Technology • Sports Technology • Motivation • Behaviour • IS use • 1 Introduction Technological development has promoted the emergence of various new technologies that allow their user to track, measure, and evaluate a multitude of personal activities and biosignals. During the past years, both individual consumers as well as the healthcare sector have become more and more interested to measure and promote health, wellness, and physical performance by using such technologies. Indeed, the number of such technologies in the market has exploded, as their popularity and significance in both individual level and in healthcare has gradually risen. Wellness technologies often include features for some form of self-tracking. The use of different self-tracking technologies has become a part of daily life for an increasing number of people (McFedries, 2013), and they are commonly used, for example, as tools to promote personal health and wellness (Swan, 2013). There once was a time when these kinds of technologies were designed purely for athletes, but nowadays there is a huge amount of different digital wellness technologies, such as devices, services, and applications aimed for regular users with all levels of physical activity. Indeed, the users of these technologies are a very varying group of users (Kettunen et al., 2017). Wellness technologies are still typically designed and developed for working age or younger population, but their increasing need and potential also among older users has been presented (Carlsson & Walden, 2017). Whilst technology in general has developed, physical inactivity has become a major global health issue. According to WHO (2017), in a global scale, around 25% of adults and more than 80% of the world's adolescent population are not sufficiently physically active. Insufficient physical activity is the fourth leading risk factor for non-communicable diseases and death worldwide, while physical activity has significant health benefits across all age groups and contributes to the prevention of diseases (WHO, 2017). Researchers in various fields have begun a serious research effort to find solutions to combat the problems of sedentary lifestyle, which are becoming increasingly widespread in our society. One stream of research concerning this has been the investigation of could different wellness technologies be used to promote the physical activity of their users. Studies on wellness technology use have found that feedback from wellness technologies can heighten the user’s awareness of personal physical activity and motivate towards it (e.g., Chan, Ryan & Tudor-Locke, 2004; Faghri et al., 2008; Kang et al., 2009; Kari et al., 2016b; Kari et al., 2017; Wang et al., 2016). However, while tracking wellness related data may result in better awareness of daily activity, it may not be sufficient to maintain the use of wellness technologies (Miyamoto et al., 2016), which can also affect maintaining wellness routines (Warraich, 2016). Thus, providing guidance and goals for using these technologies would probably increase the adherence to using them and subsequently, adherence to wellness related routines, such as physical activity. A general problem with sports and wellness technologies has been that they are more focused on providing feedback through numbers and graphs instead of providing actual manageable solutions. Research has shown that the users of wellness technologies want clear, relevant, and easy-to-understand feedback from their activities as well as information on how they should go about with their physical activity and exercise. This would likely lead to the use being more goal-oriented (e.g., Kari et al. 2016a; Kari et al. 2017b), which is generally seen beneficial (Locke & Latham, 2002; Shilts, Horowitz, & Townsend, 2004). Thus, providing the users with actual manageable steps, for example, in the form of a personalized workout plans, would increase the possibility of making both the use of the wellness technologies and the behaviour of the users more goal-oriented. One potential solution for this is digital coaching. Schmidt et al. (2015) present a demand for a goal-driven support of fitness goal achievement to be addressed by digital coaching. We define digital coaching in the context of sports and wellness technology: Digital coaching refers to information technology mediated sports and wellness related coaching activities that are generated by a software without human intervention. Coaching activities can be any kind of activities that guide, direct, train, or instruct the person being coached. The interest towards different solutions to fall under the digital coaching definition has increased among users, industry, and healthcare as well as in academia. However, as these digital coaching solutions are quite novel, the number of studies concerning their effects or the experiences that people have with them is still rather limited. Therefore, it is important to kick-start the research on the topic. This kind of research can offer numerous valuable implications for the development and use of technologies with digital coaching features, and further, aid in promoting physical activity. This study investigates the influence of digital coaching features on the physical activity motivation and behaviour of physically inactive individuals. This will provide new, relevant, and valuable information on digital coaching and technologies with digital coaching features. The main research questions of the study are: 1) How is the influence of digital coaching to physical activity motivation and behaviour of physically inactive individuals? 2) What are the central aspects of digital coaching that respond to users’ needs? The focus of the study is on the users’ subjective experiences. The study is explorative in nature and follows a qualitative approach. The study is based on thematic analysis of 20 semi-structured interviews conducted to ten individual participants. The participants were using a wellness technology with novel digital coaching features for a period of one month. Our study is one of the first ones to examine digital coaching solutions and their influence. The study presents some highly valuable first insights to digital coaching research and increases the understanding of digital coaching solutions and their use and influence. Our findings will provide the industry and parties working with physical activity promotion new science-based knowledge and implications, which will assist in designing and developing digital coaching features for wellness, health, and sports, and in planning new kinds of novel interventions to promote physical activity. In addition, our findings can be valuable to the healthcare sector when deploying technologies with digital coaching features to the patients. 2 Background 2.1 The Digital Coaching Feature Used in this Study The digital coaching feature used in our study was the Next Workout Recommendation feature by Firstbeat Technologies (Firstbeat Technologies Oy, 2018a). It is designed to work as a digital coach. A knowledge base with scientifically validated facts behind the feature can be found from Firstbeat’s White Papers and Publications website (Firstbeat Technologies Oy, 2018d). The Next Workout Recommendation feature can be varied to a certain extent depending on the requirements of the (corporate) customer. In our study, the functionality of the feature was as described below. In the beginning of the use, the user conducts a test with the feature and it provides an estimation of the person’s VO2max fitness level using “a proprietary method shown to be 95% accurate compared to laboratory measurements”. The test can be conducted during any freely performed walk, run, or ride (Firstbeat Technologies Oy, 2017; 2018b). The VO2max fitness level estimation (presented as the fitness level in the feature) gets more accurate the more the user uses the feature. After the base-level test, the user determines whether s/he is aiming to maintain the current fitness level or to improve it. If the user sets a goal to improve the fitness level, s/he will also get to choose the degree between improve and improve fast. The feature will then create an adaptive target goal for the user. The chosen approach will influence the degree of physiological impact targeted for the given workouts. Based on the background and the goal of the user, the chosen approach, and the estimated base-level of fitness, the feature creates a personalised and adaptive continuous seven-day workout plan that automatically adapts after each workout based on the activities of the user. The feature provides a specific workout or a rest day for each day. If the user misses a session or does the training longer or harder than suggested, the feature takes this into account and adapts the training plan accordingly. Moreover, when the fitness level changes, the feature will adapt the program accordingly: if the fitness level increases, the planned workouts will require more effort to be completed and if it decreases, the planned workouts will become easier (Firstbeat Technologies Oy, 2018c). Thus, the feature is suitable for people with different fitness levels and different physical activity levels, including physically inactive people. The feature is designed and best suited for aerobic training but places no restrictions regarding the specific sport conducted. If the user chooses so, the feature can also provide real-time guidance during the workout. The real-time guidance communicates to the user to maintain, increase, or decrease the current exertion level, and shows how much of the current workout has been conducted. It also shows when the goal of the workout has been reached, but the training continues until the user decides to end it. The feature also provides different views that display, for example, the user’s progress and various other information. Some of the views contain elements aiming to wake the experience of gamification (Kari et al., 2016b) in the user to further increase the motivation. Such elements include, for example, progress bars, comparison, and levels. An example view is presented in the appendix. VO2max – the used metric to present the fitness level in the feature – is a metric that defines one’s personal aerobic/cardiorespiratory fitness level. It describes the body’s capability to bring oxygen into the body through the lungs, transport it through the circulatory system to the muscles, and use it as part of body’s energy production method. Higher VO2max scores are strongly related to not just aerobic performance but also health and longevity in general (Firstbeat Technologies Oy, 2018b). For more detailed description of VO2max, see (Firstbeat Technologies Oy, 2017). 2.2 Goal Setting Goal setting can be an important facilitator for behaviour change (Locke & Latham, 2002; Shilts et al., 2004) and it provides a potential strategy for transforming physical activity information into practical actions (Shilts et al., 2004). To be effective, goal setting requires some form of self-tracking to better support the realisation of goals and the increase in self-efficacy (Fukuoka et al., 2010; Locke & Latham, 2002; Shilts et al., 2004). Goal setting can be viewed through three central dimensions: 1) self-set goals – assigned goals – participatory goals; 2) personalised goals – non-personalised goals; 3) fixed goals – adaptive goals (Zhou et al., 2018). Despite self-set goals typically being of highest personal importance, a literature review (Shilts et al., 2004) on goal setting shows that assigned goals are more effective in comparison to self-set goals (Zhou et al., 2018). Studies have also shown that interventions utilising wellness technology with fixed and non-personalised physical activity goals are often significantly less effective than interventions with adaptive goals (Adams et al., 2017; Adams et al. 2013; Jakicic et al., 2016; Zhou et al., 2018) or with personalised goals (Chan et al., 2004; Jakicic et al., 2016; Tudor-Locke, 2002; Zhou et al., 2018) in increasing physical activity levels. The weakness in providing fixed and non-personalized goals to all users is that in many cases it irresistibly leads to either goals that are too high or even unrealistic or too low and unchallenging. This obviously decreases the effectiveness of the goal setting (Tudor-Locke, 2002; Zhou et al., 2018) and leads to non-optimal behaviour. Studies have also shown that in the case of sports and wellness technology, the utilitarian goals are one of the important drivers behind use intention (Makkonen et al., 2012a; Makkonen et al., 2012b). Thus, the adaptive and personalized goals and training plans provided by the Next Workout Recommendation feature pose much potential regarding physical activity and fitness. 3 Methodology The basic setting of our study included participants who implemented the Next Workout Recommendation digital coaching feature into use for a period of one month. Due to its novelty, at the time of the use period (September-October 2017), the Next Workout Recommendation feature was not yet implemented on any commercial physical devices, though the first products implemented with it were released not long after (early 2018). Therefore, the participants used a mobile application version (provided by Firstbeat) that was installed to their smart phones before the use period started. For clarity, we refer to the used Next Workout Recommendation feature as application in the remaining of the paper. All the participants were also given a heart rate belt and personally instructed on how to use the application and how to interpret the received information. As physically inactive individuals seldom want to maintain their current fitness level but rather to increase it, all the participants were set with a goal of improving fitness level in the application (improve approach). The application then created and provided each one with a personalised target goal to be aimed for in one month and started providing the users with adaptive continuous seven-day workout plan to reach that target goal. As the aim of the study was not in increasing the physical fitness levels of the participants per se but to investigate the application’s influence on motivation and behaviour, we did not require the participants to conduct every workout the application suggested, but rather wanted the use to be as natural as possible. Thus, it was a voluntary choice to do or not to do the suggested workouts. This kind of autonomy is also central in the self-determination theory (Ryan & Deci, 2000). To conduct our study, we chose a qualitative approach. Qualitative research aims to understand people, their sayings and behaviour, as well as the cultural and social context they are living in. The goal is to find new knowledge and understand real life phenomena. One of the central benefits of qualitative research is enabling the researcher to see and understand the underlying contexts in which decisions are made and actions take place (Myers, 2013). To collect the data, we chose interviews as the data collection method. According to Myers & Newman (2007, p. 3) qualitative interview is the “most common and one of the most important data gathering tools in qualitative research”. More precisely, we chose a semi-structured interview, which is the most used type of interview in IS research. In a semi-structured interview, there is an incomplete script, but it is typical that some pre-formed structure is prepared that the interviewer follows (Myers & Newman, 2007). This was also the case in our interviews. In planning and conducting the interviews, we followed relevant guidelines (e.g., Guest, Bunce & Johnson, 2006; Myers, 2013; Myers & Newman, 2007) regarding semi-structured interviews. The goal was to gain maximal benefit from using the method and to avoid the potential problems and pitfalls. We conducted two interviews for each interviewee, one before and one after the use period. The first one mainly focused on the expectations towards the application and its influence, whereas the second one focused mainly on the experiences and perceived influence of using the application on physical activity and exercise motivation and behaviour. Therefore, two distinct interview scripts were created. One for the interviews before (Script 1) and one for the interviews after (Script 2) the use period. It is to be noted that the interviews also included themes for a purpose of another study. However, this cannot be seen to affect the results of this study. However, it affected the development of the scripts so that they also included themes and questions deriving from theories less connected to this particular paper. The themes for Script 1 were mainly developed based on the research questions, previous literature, and the theoretical background, while some themes also derived from the UTAUT2 (Venkatesh, Thong & Xu, 2012) and innovation-decision process (Rogers, 2003). Script 2 included the same aspects but focusing more on the experiences from the use period plus questions deriving from self-determination theory (Ryan & Deci, 2000). The themes of both interviews are presented in the appendices. But as mentioned, not all themes were used in this particular paper. Following Myers and Newman (2007), the interviews included the opening, the introduction, key questions related to certain themes, and the closing. To recruit the participants, we used a snowball sampling approach (Patton, 2002). First, we searched persons that met the criterion and then suggested them with the possibility to participate in the study. Then, the selected participants were asked for and provided information on additional potential participants and these again were contacted. Following McCracken (1988), in selecting the interviewees we aimed for a fairly homogenous sample that share critical similarities concerning the research question, this way aiming to maximise the richness and depth of the data. In addition, certain criteria were used: 1) the person was physically inactive at the moment, 2) the person had the possibility and interest to implement the application into use, and 3) the person was motivated to take part in the research. In the end, the study was conducted with ten participants. We conducted 12 interviews before the use period, but had two dropouts during the use period due to sickness affecting not being able to exercise. Thus, we conducted 10 interviews after the use period. We did not include the interviews from the two dropouts in our analysis. The first round of interviews was held early September 2017 before the use period began, and the second interviews during November 2017 by one of the authors. The interviews were held face-to-face with the interviewee, with the exception of one interview that was conducted via mobile phone. On average, the interviews lasted 25 and 49 minutes. The interviews were recorded and transcribed (in the interviewees native language). Based on the transcriptions and notes made during the interviews, the analysis began. For the purposes of this particular paper, our analysis mainly focused on the data collected with the second-round interviews. To analyse the data, we used thematic analysis, which is the most widely used method of analysis in qualitative research (Guest, MacQueen & Namey, 2012). Thematic analysis is a method for “identifying, analyzing and reporting patterns (themes) within data” (Braun & Clarke, 2006, p. 79). Using thematic analysis, a researcher can organise and describe the data set in rich detail, and typically go even deeper by interpreting various aspects of the research subject (Braun & Clarke, 2006). In addition to identifying, analysing, and reporting the themes in our data, we also aimed to interpret various specific related aspects and exceptions. In conducting the analysis, we applied guidelines from Braun and Clarke (2006) and Patton (2002). As they suggest, these guidelines were applied flexibly to fit the research questions and the collected data. Further, as suggested, the analysis process was not a linear phase-to-phase process but a recursive one, in which we moved back and forth between the different phases of the analysis. 4 Results The ten participants consisted of five females and five males with the mean age of 31,6 years. The description of each participant can be found in Table 1. In general, the participants had rather little previous experience on using sports or wellness technology. Only two owned some such device at the moment, but neither of them was using it actively. The physical activity background of the participants varied a lot. Some had a more active history through some hobby and some had even competed earlier in their life, whereas others had been more or less physically inactive their whole life and regarded themselves as beginners in terms of exercise. However, it is to be noted that even though some participants had been more physically active in the past, at the moment when the use period started, they all considered themselves to be physically inactive, and at most, taking part in exercise activities was occasional. Everyone also perceived their current aerobic fitness level as low or fair at best. Table 1: Description of the Sample Participant Gender Age Physical Activity Background Previous Experience on Sports Technology 1 Female 34 Competition history Heart rate monitor 2 Female 28 Beginner – 3 Male 28 Competition history – 4 Female 21 Beginner – 5 Male 33 Competition history Mobile applications 6 Male 41 Competition history Heart rate monitor 7 Male 30 Hobby history Heart rate monitor, app 8 Female 25 Competition history Mobile applications 9 Female 51 Competition history – 10 Male 25 Hobby history Pedometer (in a phone) 4.1 Influence of Digital Coaching to Physical Activity Motivation and Behaviour The digital coaching application was found to beneficial for physical activity and exercise motivation. Each of the participants reported that their motivation towards physical activity had increased during the use period. The element that most affected the motivation was reported to be that some outsider told what to do and how to do it. Also, being able to follow own progress was perceived highly motivational. Another aspect that influenced the motivation was that the digital coach had a role in changing the person’s whole mentality regarding physical activity. Many reported that they had previously had different kinds of challenges that inhibited being physically active, for example, related to false assumptions regarding how they should exercise. The application was also perceived to be a better motivator than most of the previous things that the users had tried, as the application was more capable in solving those challenges that the users had previously experienced. The application was also able to make the physical activity more goal-oriented. The users wanted to accomplish the goals of individual workouts, and achieving those was perceived as steps in achieving the overall goal of improving fitness. Workout specific goals were reported to be sufficiently short-term, and thus, easier to achieve and good in maintaining motivation. Further, the users reported that if they only had a much larger goal to achieve at once, it would likely feel unachievable and subsequently lower the motivation. Achieving these smaller workout specific goals was perceived to increase personal capability towards exercise and additionally make one proud of him or herself, further increasing the motivation. Seeing the upcoming workouts from the application was perceived as very interesting, and many reported to be waiting for the moments when they received updates to their adaptive workout plan. The length or the exertion level of the upcoming workout did not affect whether the users planned to go perform it or not. In addition, being able to see the workout plan beforehand helped in organising the day’s schedule and the daily life in general, which for some was a big but positive change. This aided in fitting physical activity into the daily routines and transforming the increased motivation to actual physical activity behaviours. The digital coaching application also influenced the users’ actual behaviours in many ways. First of all, it increased their level of physical activity. Whereas many had felt exercising challenging for one reasons or another before implementing the application, the application lowered the level of starting a workout by providing a ready-made plan and guidance to individual workouts. Further, it helped to realise that being physically active and doing workouts is not actually that demanding as they had thought. Everyone reported that the application had developed their habits and ways of being physically active. The application also changed the way they conducted their individual workouts. Many reported that when they had previously tried exercising for a while, they had done so with a much higher exertion level, because they had thought that for a workout to be beneficial, one has to go all out. Thus, they usually had exercised too hard and subsequently been exhausted after each workout, which had made them feel negative towards exercising and stop it altogether. Instead, now they felt that the exercise intensity during the workouts had been much lower and more pleasing, which subsequently had made them experience better feeling during the workout and also after it. The real-time guidance had a major role in keeping the right heart rate and exertion level during the workouts, and also in teaching the users how certain kinds of workouts should feel physically. It also motivated to keep going until the goal of the workout had been reached, especially when they didn’t otherwise feel so keen on exercising. Indeed, even though the application had made the feeling during and after the workouts more pleasing, it did not magically make everyone enjoy every aspect of exercising all at once, and the sofa was still seen as an intriguing option, yet one that was chosen much less frequently than before. Reaching the workout specific goals was perceived important and trust towards the application was strong. Therefore, the real-time guidance was generally followed quite accurately. However, this also had its turnside. If one wanted to follow the real-time guidance accurately, it made social exercising more difficult. The need to follow a certain exertion level for a certain time was sometimes seen as difficult to do together with someone else. Though some also reported that the application and especially its real-time guidance made them feel like they had a companion when exercising alone. Receiving a plan also affected other aspects of life than just physical activity. The users reported that they had started to pay more attention to nutrition and proper breakfast or skipped drinking alcohol if they had a workout planned for the next day. In other words, the application did not just influence those moments of exercise but also the time between them. Almost everyone reported that their appreciation towards physical wellness had changed during the use period. As they had been physically more active, they had noticed it influencing their sleep positively and made them pay more attention to their nutrition and eating rhythm. Everyone planned to continue being more physically active also in the future. 4.2. Central Aspects of Digital Coaching Regarding Users’ Needs and Challenges Digital coaching can answer to various kinds of needs and bring solutions to various challenges. However, we also found certain aspects that should be paid more attention to in the design process. The most notable need that the application answered to was providing the users personalized workouts suited for their personal fitness level, doing of which had previously been challenging to many. This led the user to exercise with right exertion level and frequency. The ready-made workout plans also decreased the perceived stress regarding physical activity, as the planning and tracking was done by the application. The users also reported that despite not necessarily having a particularly good overall knowledge on physiology, it didn’t matter, as the application in a way took care of that part. The perceived advantage over optional wellness technologies without digital coaching features was evident, and everyone would have been willing to pay for such digital coaching features alone. The three most central features regarding matching own previous needs and challenges were reported to be personalised workout plans, real-time guidance, and the ability to follow own progress. Personalisation was especially highlighted and appreciated. It was also highly regarded that the application was able to adapt the workout plan based on the user’s activity, for example, if one had overachieved or in contrast skipped a workout. An additional need that the application answered to was the ability to start a guided workout straight from one’s own door, without the need to separately go for a group workout session in a gym, which could be challenging due to requiring more time and having a higher starting step. One interesting finding was that the application was perceived to be as any coach or personal trainer in a sense that the users would have wanted to communicate with it more. For example, if a user knew beforehand that s/he was unable to conduct the upcoming workout due to work, travel, not feeling well, or another obstacle, s/he would have wanted to tell the digital coach this, so that the coach could have moved the workout for the day before or otherwise adapt the plan. Some also had specific weekdays when they were never able to conduct workouts due to other hobbies. It also seemed to us that among some users, the application had roused a sense of obligation towards the digital coach, and they would have wanted to explain why they were unable to conduct certain workouts. Thus, it can be said that for some, the interaction with the digital coach application had some human characteristics in it. Even though the users mostly perceived following the development of their fitness in a general level as clear, for some, presenting the actual values and fitness level through VO2max was difficult to comprehend. They had difficulties in understanding the scale and what kind of changes in VO2max would be realistic in the first place. It was also wished that the application would have notified about increased fitness level with more clarity and accompanied rewarding. Some elements of the real-time guidance were also reported to have waken the experience of gamification, for example regarding reaching own goals. Those elements were liked and some expressed they would have liked to see more of such elements in the application. On the other hand, even though it was generally wished that the application would have provided further (digital) rewards, the users would not have wanted it to exaggerate accomplishing goals of individual workouts, as the most important feeling of success came from within. The application did not provide separate reminder notifications for the users about individual workouts. Whether it should have divided opinions. Whereas some would have valued daily notifications of the next workout, some reported that it might have felt pressuring. The users reported that they had learned much from using the digital coach, for example, about heart rate levels and how to exercise according to them, how a certain exertion level should feel, and scheduling the workouts into the daily life. Finding the solutions to these needs and challenges would influence the ways of exercising in the future. 5 Conclusions The purpose of this study was to increase the understanding of novel digital coaching solutions and their use. More precisely, we investigated the influence of digital coaching features on the physical activity motivation and behaviour of physically inactive individuals. The focus was also on central aspects of digital coaching that respond to users’ needs. In our qualitative study based on interviews and thematic analysis, the participants used the Next Workout Recommendation digital coaching feature by Firstbeat Technologies for one month. We also provide a definition of digital coaching in the context of sports and wellness technology: Digital coaching refers to information technology mediated sports and wellness related coaching activities that are generated by a software without human intervention. Coaching activities can be any kind of activities that guide, direct, train, or instruct the person being coached. The application with the digital coaching feature was found to beneficial for physical activity and exercise motivation. The element that most affected the motivation was some outsider, in this case the digital coach, instructing the what, when, and how regarding exercise. Further, knowing that these instructions were targeted specifically to them as individuals and based on their personal fitness level and actions was much appreciated and perceived important. Also, being able to follow personal progress was perceived highly motivational. The application was also able to make the whole physical activity regime more goal-oriented, which had a further motivating effect. The application also aided in fitting physical activity into the daily routines and transforming the increased motivation to actual physical activity behaviours. The digital coaching application influenced the users’ actual behaviours in many ways. First and foremost, it was able to increase the level of physical activity for these previously physically inactive individuals. It also guided and taught them to conduct their workouts with appropriate frequency and exertion level, which subsequently led them to enjoy exercising more and get more positive feelings after the workouts. Thus, the application was able to shift their whole mentality regarding physical activity and fitness to more positive direction. The digital coaching feature influenced also other aspects of life than just physical activity. The users had started paying more attention to, for example, nutrition, diet, sleep, and alcohol consumption. In other words, the application did not just influence those moments of exercise but also the behaviours in between them. Subsequently, it increased the general appreciation towards physical wellness in a positive way. The application was able to meet several kinds of needs and give solutions to various previously felt challenges. Thus, it seems that a digital coaching feature like the one used in our study can really be beneficial. At least among physically inactive individuals and on a short-term. As the use period lasted only one month, it is difficult to draw longer-term conclusions. Also, the needs and challenges of, for example athletes, can be very different. Thus, the particular digital coach used in our study might not be so influential and successful among other user groups, and more research is needed in that regard. The digital coaches should indeed be designed with the target group in mind, first investigating the target group. One interesting finding was that the digital coach was perceived as any coach or personal trainer in a sense that the users would have wanted to communicate with it more, for example, regarding upcoming or missed workouts. This is something for the designers to acknowledge. They should provide communication and interaction possibilities between the user and the digital coach. The users clearly want to explain if they are not or have not been able to follow the plan and want the digital coach to adapt the workout plan accordingly. Also, as the users clearly perceived the interaction with digital coach to have human characteristics, the digital coaches could have some human like avatars or similar. As for further practical implications, it would probably be valued if the digital coach could better be able to adapt individual workouts so that it would better support social exercising, should the user want to exercise with a friend for example. In the beginning of the use, the metrics used to display fitness level should be thoroughly explained. The users could also be provided with different options regarding the metrics, so they could choose between different metrics based on their own will and background knowledge. The designers should also undertake the process of gamification (Kari et al., 2016b) and implement digital coaches with gamified features. But needing to be careful not to overemphasise individual elements or overgamify the use. Special attention should also be paid to conducting the base-level tests as accurately as possible. If the digital coach would estimate the user’s base-level incorrectly and with relatively high error, it would probably lead to wrong kinds of plans and individual workouts and to presenting the progress in physical fitness falsely. Nevertheless, even with the need for great accuracy, conducting the base-level test should still be easy and effortless enough also for physically inactive users, who do not necessarily possess much knowledge on physical activity. This is one of the main challenges for the designers to overcome. To conclude, this study is one of the first ones to examine novel digital coaching solutions and their influence. Increasing the understanding of digital coaching solutions and their use and influence, the study provides highly valuable first insights both for the researchers and for the wellness and sports technology providers. The industry as well as the actors working with physical activity promotion can utilise our findings and implications in designing and developing digital coaching features for wellness, health, and sports, and in planning new kinds of novel interventions to promote physical activity. The study also lays foundation for future research on digital coaching. 6 Limitations and Future Research There are few notable limitations in the study. First, although the interviews provided a very valuable set of information, the number of participants in the study could have been higher. The study had to consist of physically inactive people willing to implement a sports technology application with digital coaching features into use, which is not the most typical of target groups, though a highly important one. However, all interviewees had interest towards the solution, and thus, the implementation was based on own interest. We also followed previous guidelines (McCracken, 1988) in selecting the interviewees. We also believe that an adequate number of interviews were conducted, as we recognised that their marginal benefit was notably reduced. Second limitation concern the relatively short usage period of one month, which reduces the ability to make long-term interpretations. However, from a physical point of view, a month is considered long enough to be able to see changes in fitness level, especially among people with below average fitness level (Jürimäe et al., 1985). Third limitation relates to the general limitations of qualitative interview. To overcome these, we followed well-established guidelines for semi-structured interviews in planning and conducting the interviews (e.g., Braun & Clarke, 2006; Guest et al., 2006; Myers, 2013; Myers & Newman, 2007). This, we believe, enabled us to avoid the possible problems and pitfalls and helped us to gain maximal benefit from using a semi-structured interview. As Guest et al. (2012) point out, the identified themes in the thematic analysis are always based on the interpretations of the researchers. Therefore, we also applied relevant guidelines in doing the analysis. As is typical to qualitative research, it is difficult to make generalizations from the sample to a larger population (Myers, 2013) and one should be cautious if doing so. Despite these limitations, the research offers some highly valuable first insights and a good overview on the subject. The findings of the study also provide potential paths for future research. Being one of the first studies to examine these kinds of digital coaching solutions and their influence, it can serve as a starting point and a reference to future studies. Future research would benefit if to be conducted with larger samples and longer use periods. The target group could naturally also be other than physically inactive people, as digital coaching has the potential to be beneficial for people with all levels of physical activity and fitness as well as for people of different ages. Future studies could also specifically focus to certain types of users who share similar goals, for example, regarding a certain sport, certain competition, physical ability, weight loss, or other specific area. We would also encourage researchers to conduct more research in regards to digital coaching and the different aspects of it, not just from the IS perspective but from other perspectives as well, for example, health and exercise science and psychology. Thus, we make a call for cross-disciplinary research on digital coaching and their effects on human behaviour and wellbeing. References Adams, M. A., Hurley, J. C., Todd, M., Bhuiyan, N., Jarrett, C. L., Tucker, W. J., ... & Angadi, S. S. (2017). Adaptive goal setting and financial incentives: a 2× 2 factorial randomized controlled trial to increase adults’ physical activity. BMC Public Health, 17(1), 286. Adams, M. A., Sallis, J. F., Norman, G. J., Hovell, M. F., Hekler, E. B., & Perata, E. (2013). An adaptive physical activity intervention for overweight adults: a randomized controlled trial. PloS One, 8(12), e82901. Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101. Carlsson, C., & Walden, P. (2017). Digital Coaching to Build Sustainable Wellness Routines for Young Elderly. In The 30th Bled eConference “Digital Transformation – From Connecting Things to Transforming Our Lives” Research Volume, 18.-21.6.2017 (pp. 57-70). Bled, Slovenia: University of Maribor. Chan, C. B., Ryan, D. A., & Tudor-Locke, C. (2004). Health benefits of a pedometer-based physical activity intervention in sedentary workers. Preventive Medicine, 39(6), 1215-1222. Faghri, P. D., Omokaro, C., Parker, C., Nichols, E., Gustavesen, S., & Blozie, E. (2008). E-technology and pedometer walking program to increase physical activity at work. The Journal of Primary Prevention, 29(1), 73-91. Firstbeat Technologies Oy. (2017). Automated Fitness Level (VO2max) Estimation with Heart Rate and Speed Data. White Paper. Firstbeat Technologies Oy. (2018a). Retrieved 6.2.2018 from https://www.firstbeat.com/ Firstbeat Technologies Oy. (2018b). VO2max Fitness Level. Retrieved 6.2.2018 from https://www.firstbeat.com/en/consumer-feature/vo2max-fitness-level/ Firstbeat Technologies Oy. (2018c). Personalized Training Plans: Cardio. Retrieved 6.2.2018 from https://www.firstbeat.com/en/consumer-feature/personalized-training-plans-cardio/ Firstbeat Technologies Oy. (2018d). White Papers and Publications. Retrieved 19.4.2018 from https://www.firstbeat.com/en/science-and-physiology/white-papers-and-publications/ Fukuoka, Y., Vittinghoff, E., Jong, S. S., & Haskell, W. (2010). Innovation to motivation – Pilot study of a mobile phone intervention to increase physical activity among sedentary women. Preventive Medicine, 51(3-4), 287-289. Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? an experiment with data saturation and variability. Field Methods, 18(1), 59-82. Guest, G., MacQueen, K. M., & Namey, E. E. (2012). Applied thematic analysis. Los Angeles, CA: SAGE. Jakicic, J. M., Davis, K. K., Rogers, R. J., King, W. C., Marcus, M. D., Helsel, D., ... & Belle, S. H. (2016). Effect of wearable technology combined with a lifestyle intervention on long-term weight loss: the IDEA randomized clinical trial. JAMA, 316(11), 1161-1171. Jürimäe, T., Viru, A., Pedaste, J., & Toode, K. (1985). Changes in physical capacity and serum lipids and lipo-proteins during running training program in female university students. Biology of Sport, 2(4), 243-253. Kang, M., Marshall, S. J., Barreira, T. V., & Lee, J. O. (2009). Effect of pedometer-based physical activity interventions: a meta-analysis. Research Quarterly for Exercise and sport, 80(3), 648-655. Kari, T., Kettunen, E., Moilanen, P., & Frank, L. (2017a). Wellness Technology Use in Everyday Life: A Diary Study. In The 30th Bled eConference “Digital Transformation – From Connecting Things to Transforming Our Lives” Research Volume, 18.-21.6.2017 (pp. 279-294). Bled, Slovenia: University of Maribor. Kari, T., Koivunen, S., Frank, L., Makkonen, M., & Moilanen, P. (2016a). Critical Experiences During the Implementation of a Self-Tracking Technology. In The 20th Pacific Asia Conference on Information Systems (PACIS), 27.6.-1.7.2016 (16 pages). Chiayi, Taiwan: AIS. Kari, T., Koivunen, S., Frank, L., Makkonen, M., & Moilanen, P. (2017b). The expected and perceived well-being effects of short-term self-tracking technology use. International Journal of Networking and Virtual Organisations, 17(4), 354-370. Kari, T., Piippo, J., Frank, L., Makkonen, M., & Moilanen, P. (2016b). To gamify or not to gamify?: gamification in exercise applications and its role in impacting exercise motivation. In The 29th Bled eConference "Digital economy" Research Volume, 19.-22.6.2016 (pp. 393–405). Bled, Slovenia: University of Maribor. Kettunen, E., Kari, T., Moilanen, P., Vehmas, H. M., & Frank, L. (2017). Ideal Types of Sport and Wellness Technology Users. In The 11th Mediterranean Conference on Information Systems (MCIS), 4.-5.9.2017 (12 pages). Genoa, Italy: AIS. Locke, E. A., & Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist, 57(9), 705-717. Makkonen, M., Frank, L., Kari, T., & Moilanen, P. (2012a). Explaining the usage intentions of exercise monitoring devices: The usage of heart rate monitors in Finland. In the 18th Americas Conference on Information Systems (AMCIS) 2012, 9.-11.8.2012 (Paper 13). Seattle, USA: AIS. Makkonen, M., Frank, L., Kari, T., & Moilanen, P. (2012b). Examining the usage intentions of exercise monitoring devices: The usage of pedometers and route trackers in Finland. In The 25th Bled eConference “eDependability: Reliable and Trustworthy eStructures, eProcesses, eOperations and eServices for the Future” Research Volume, 17.-20.6.2012 (pp. 439–453). Bled, Slovenia: University of Maribor. McCracken, G. (1988). The long interview. Newbury Park, CA: SAGE. Mcfedries, P. (2013). Tracking the quantified self. IEEE Spectrum, 8(50). 24. Miyamoto, S. W., Henderson, S., Young, H. M., Pande, A., & Han, J. J. (2016). Tracking health data is not enough: a qualitative exploration of the role of healthcare partnerships and mhealth technology to promote physical activity and to sustain behavior change. JMIR mHealth and uHealth, 4(1), e5. Myers, M. D. (2013). Qualitative research in business and management (2nd edition). Los Angeles, CA: SAGE. Myers, M. D., & Newman, M. (2007). The qualitative interview in IS research: examining the craft. Information and Organization, 17(1), 2-26. Patton, M. Q. (2002). Qualitative research & evaluation methods (3rd edition). Thousand Oaks, CA: SAGE. Rogers, E. M. (2003). Diffusion of innovations (5th edition). New York, NY: Free Press. Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68-78. Shilts, M. K., Horowitz, M., & Townsend, M. S. (2004). Goal setting as a strategy for dietary and physical activity behavior change: a review of the literature. American Journal of Health Promotion, 19(2), 81-93. Schmidt, B., Benchea, S., Eichin, R., & Meurisch, C. (2015). Fitness tracker or digital personal coach: how to personalize training. In The 2015 ACM International Joint Conference on Pervasive and Ubiquitous Computing and ACM International Symposium on Wearable Computers, 7.-11.9.2017 (pp. 1063-1067). Osaka, Japan: ACM. Swan, M. (2013). The quantified self: Fundamental disruption in big data science and biological discovery. Big Data, 1(2), 85-99. Tudor-Locke, C. (2002). Taking steps toward increased physical activity: Using pedometers to measure and motivate. President's Council on Physical Fitness and Sports Research Digest, 3(17), 3-10. Venkatesh, V., Thong, J. Y., & Xu, X. (2012). Consumer acceptance and use of information technology: extending the unified theory of acceptance and use of technology. MIS Quarterly, 36(1), 157-178. Wang, J. B., Cataldo, J. K., Ayala, G. X., Natarajan, L., Cadmus-Bertram, L. A., White, M. M., ... & Pierce, J. P. (2016). Mobile and wearable device features that matter in promoting physical activity. Journal of Mobile Technology in Medicine, 5(2), 2-11. Warraich, M. U. (2016). Wellness Routines with Wearable Activity Trackers: A Systematic Review. In The 10th Mediterranean Conference on Information Systems (MCIS), 4.-6.9.2016 (13 pages). Paphos, Cyprus: AIS. World Health Organization (WHO). (2017). Physical activity. Retrieved 1.2.2018 from http://www.who.int/mediacentre/factsheets/fs385/en/ Zhou, M., Fukuoka, Y., Mintz, Y., Goldberg, K., Kaminsky, P., Flowers, E., & Aswani, A. (2018). Evaluating Machine Learning–Based Automated Personalized Daily Step Goals Delivered Through a Mobile Phone App: Randomized Controlled Trial. JMIR mHealth and uHealth, 6(1), e28. Appendices Appendix A: Themes of the 1st round interviews 1. Background 1.1 Demographics 1.2 Socioeconomic characteristics 2. Physical activity background and meaning of wellbeing 3. Previous experience 4. Innovativeness, personality variables 5. Challenges and needs with physical activity 6. Performance expectancy and personal goal setting 7. Effort expectancy and facilitating conditions 8. Hedonic motivation 9. Habit 10. Physical activity and exercise motivation 11. Social aspects More detailed descriptions with key questions are available from the authors by request. Appendix B: Themes of the 2nd round interviews 1. Meaning of wellbeing 2. Use experiences 3. Social aspects 4. Challenges and needs with physical activity 5. Performance experience and personal goal setting 6. Effort experience and facilitating conditions 7. Hedonic motivation 8. Elements of the solution and physical activity motivation 9. Habit 10. Relative advantage 11. Adaptability 12. Price value 13. Physical activity and exercise motivation 14. Other aspects More detailed descriptions with key questions are available from the authors by request. Appendix C: An example view of the Next Workout Recommendation Feature Perceived Intrusiveness of Personalized Marketing KATRI KEREM & MONIKA ULLA 10 CORRESPONDENCE ADDRESS: Katri Kerem, Ph.D., Professor, Estonian Business School, Department of Marketing and Communication, Lauteri 3, Tallinn 10114, Estonia, e-mail: katri.kerem@ebs.ee. Monika Ulla, MA, Estonian Business School, Department of Marketing and Communication, Lauteri 3, Tallinn 10114, Estonia, e-mail: monika.ulla@hotmail.com DOI https://doi.org/10.18690/978-961-286-170-4.9 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract Businesses collect vast amounts of customer data via forms that consumers fill in themselves and with automatic data collection tools. Advances of data collecting methods outpace consumers’ understanding of how their personal data is used. This paper investigates how the consumers perceive the pros and cons of marketers using their personal data for customized marketing messages and product offers. A qualitative study was conducted among Estonian consumers. In depth interviews with 18 consumers revealed surprisingly positive attitudes but also highlighted how consumers do the privacy calculus. Data based personalization is welcomed if the consumers perceive to receive benefits for disclosing their data. The results demonstrate that women are more willing to share their personal data for business purposes. Keywords: • Personalized marketing • Privacy concerns • Personalization-privacy paradox • 1 Introduction Personalization is by no means a new idea or concept in the world of marketing. Consumers are increasingly seeking customized solutions and businesses are trying to meet these needs without compromising the economies of scale. Advances in data collection methods allow companies to gather vast amount of non-self-disclosed information about consumers, to automate the monitoring of consumers’ purchasing behaviour and to use the accumulated data for targeted marketing communication and customizing their offers. Personalized marketing enabled by data management platforms is integral to consumers receiving relevant, timely, engaging, and personalized messaging and advertisements that resonate with their unique needs and preferences. Although personalization, customization, and exact targeting is appreciated by the consumers while saving time and money, such personalized services demanding detailed customer information may cause privacy related problems. To avoid negative attitudes and loss of goodwill, businesses must be sensitive to consumer expectations associated with data collection and usage. The aim of this research paper is to map the attitudes of Estonian consumers and study the link between personalized marketing and perceived intrusiveness among Estonian consumers. The authors will analyse the consumers’ attitude towards personalization, their willingness to disclose personal data and the perceived usefulness of personalized marketing communication. 2 Literature Review 2.1 Personalization in Marketing The aim of personalization is to speed up customers’ decision and purchase process via tailoring marketing messages and product offers. By definition, personalization is the ability and skill to proactively tailor products and customer experience to the preferences of individual consumers based on their personal information (Chellappa & Sin, 2005). As Vesanen (2007, pp. 409-410) points out, “the concept of personalization is currently used to cover a very fragmented set of ideas, and a clear theoretical framework is lacking”. Both consumer acceptance of personalization and the tools and techniques of personalization are in constant flux as the technologies change at a rapid pace. Tactics that worked well in their early phases may become counterproductive as customers get fed up by certain type of communication. Personalization seems to promise savings in time and money both for the seller and the buyer but nevertheless there is a thin line between appropriate personalization and perceived intrusiveness (Truong & Simmons, 2010). Personalization can both strengthen or weaken the relationship customer has with a brand (Aguirre et al., 2016) and the reactions to personalized messages can be either positive or negative (White et al., 2008). Thus the question of personalization poses significant strategic challenges to marketers. The benefits of personalization include increasing the efficiency of the client’s purchase process and specifically narrowing down alternative solutions/actions to the most relevant ones. Time savings are an important bonus (Aguirre et al., 2016). At the same time it may feel unpleasant to realize that someone is following and recording your activities and data and this can lead to lower adoption rates, increased sense of vulnerability (Aguirre et al., 2016) or avoiding doing business with specific companies at all (McDonald & Cranor, 2010). Successful personalization can only be based on having relevant and sufficient information about the consumers, including, but not limited to, the information that consumers have shared about themselves on informed consent and information that has been collected with the help of digital technologies. The quality of personalization and ability to meet individual needs is vital for successful personalization (Jahng, Jain & Ramamurthy, 2002) and it is possible to improve the usefulness of personalization for both the company and the consumer only via a very thorough customer data collection and analysis. Collecting, storing and utilizing customer data will inevitably lead customers to question the value of allowing businesses to work with the data and ponder about the ethical limits of tracking customers. Therefore it is important to understand consumers’ perception of privacy and their willingness to trade privacy for personalization. 2.2 Privacy Conceptualizations and Concerns Various authors have defined privacy and there is no single definition that would encompass all the aspects of the concept. Westin (1997) has defined privacy as a right of an individual to decide how, when and to what extent to disclose personal information to others. Clarke (1999) has added that privacy also includes a person’s will to sustain a personal space that would be free from interference by others (including also businesses in the context of marketing). From consumers’ point of view privacy is highly subjective and depends on one’s ability to understand privacy related issues and take decisions regarding that. Past research has identified a privacy concern as a two-component construct. If a customer has good knowledge and (perceived) control over collecting, storing and using data about her/him, the privacy concern tends to be low, and vice versa (Milne 2000, Sheehan & Hoy 2000). Alt, Hartwig and Reinhold (2016) have included the dimension of privacy awareness into the discussion, claiming that without awareness consumers are unable to make decisions and give educated permission to use their personal data for business purposes. Awareness, in turn, does not necessarily lead to rational decisions. Knowledge-behaviour gap is well documented in various fields, and it applies for privacy decisions as well (Bakos, Marotta-Wurgler & Trotten, 2009; Murumaa-Mengel, et.al., 2014). Most of the privacy laws and regulations postulate that people should be given complete and thorough information about data collection, storage and usage but in reality people hardly pay attention. The key reasons for not paying attention to privacy conditions and agreements are information overload and lengthy material (Helberger, 2013), easiness of just ticking the box and agreement as a default option (Murumaa- Mengel, et.al., 2014), or unclear wording of the text (Lurger 2013; Murumaa-Mengel, et.al., 2014). Dinev and Hart (2006) have described e-commerce customers as rational actors who conduct cost-benefit analysis comparing gains and losses of giving away data. But frequently data disclosure is not under the control of the consumer, or at least consumer cannot avoid giving delivery address, credit card data, etc, for a vendor for fulfilling the order (Smith & Shao, 2007) and this creates a personalization-privacy paradox, a situation where consumers are willing to share data for some purposes and reluctant to do so for some other reasons (Aquirre et al., 2016). The questions and decisions regarding privacy and data disclosure are not easy ones for consumers. The concept itself is multifaceted and difficult to comprehend for the consumers, as are contemporary data collection methods. It is difficult for consumers to take control over their privacy because most of the data based marketing activities happen in a “black box” for them and they can only guess what will happen if their data is analysed. 3 Methods The aim of the current study was to map the attitudes and perception towards privacy and study the link between personalization and intrusiveness. The study adopted qualitative approach as privacy related issues are perceived and addressed very differently by consumers. A quantified research instrument would not have been adequate to capture the meaning of the phenomenon in the minds of consumers. In-depth interviews are considered appropriate when the goal is to collect information on emerging issues (Marshall & Rossman, 2006). The semi structured interview plan was composed to include the main themes arising from past research and structured around the key research questions. The questions in the interview plan were constructed by the authors of the current study. The first section of questions was designed to understand how consumers understand personalization and to seek insights about their experiences and beliefs. Previous studies pointed out both positive and negative aspects regarding personalization and the authors aimed to clarify the balance among Estonian consumers. The second section of interview plan included guiding questions to see how consumers understand privacy and how they rationalize giving access to their data. The authors further wanted to explore if the consumers undertake some sort of privacy calculus expecting value in return to access to their data. The third section of interview questions addressed personalized marketing communication and aimed to identify the balance between losing privacy and enjoying benefits from precisely targeted communications. A non-probability convenience sampling method was used to recruit voluntary participants for this study. The call for participation was posted to 5 different social media groups in Estonia (heritage society, gardening group, marketplace, neighbourhood group and baby products group). These groups were selected to achieve variety in the sample and all of them were known to the researchers to have an active membership. Further, the researchers recruited an equal number of male and female participants and observed that the interviewees would be of different backgrounds. The initial plan was to conduct a total of 20 interviews but data saturation (Miles & Huberman, 1994) was achieved with 18 interviews. Such size of the final sample is considered suitable (Baker & Edwards, 2012). The interviews lasted between 45 minutes and 2.5 hours and were conducted in locations preferred by the interviewees. The age range of the interviewees was between 28 and 74. All the interviews were recorded with the permission of the participants, and transcribed in verbatim. Coding was conducted manually. The initial codes were developed deductively from the previous studies and additional inductive codes emerged from the material. Content analysis was used to interpret the textual data through classification that involves coding and identifying common themes (Hsieh & Shannon, 2005). Structural coding and categorization were conducted following the principles and guidelines presented by Saldaña (2009). Specific information on coding does not fit the length limits of the current paper and is available from the authors. 4 Results and Discussion 4.1 Consumer experiences and attitudes regarding personalization The initial analysis of the research material made it clear that marketers and researchers have a tendency to overestimate the knowledge of consumers regarding various marketing techniques and concepts related to digital business. Only one participant among the 18 associated personalization with mapping customer needs on individual level and making offers based on that. The typical reasoning was as follows: Personalization.... is that marketing according to someones’ needs? You pick a target group and make offers to them? (P4, M2811) 11 P denotes the number of participant, M/F denotes gender, and the number shows age of the study participant. Because of the widespread practice of phone marketing this topic was also frequently linked to the perception of personalization. When the researchers explained personalization with guiding questions the respondents started to recall their own experiences with individualized offers. The prevalent opinion about personalized marketing was positive, participants pointed out time saving and narrowing down choice options. The study confirmed that the consumers actually appreciate offers that are based on their purchase history and current needs. A few consumers were critical towards the idea of personalization claiming that there is a lot of very shallow personalization. One of the more critical participants commented: Personalization seems to come to an end with using my name, everything that follows is mostly generic. They do not pay enough attention. The first stage shows me that it is personalized but contentwise it is empty (P4, M28) Additionally, the interviewees highlighted that personalization helps them to feel special, even if they recognize someone’s business interests behind that. Still, such approach gives customers a feeling that someone cares about them and addresses their problems. This in turn is vital for successful long term customer relationships I feel myself special, someone has taken trouble to deal with my problem. Makes me feel important.... (P3, F38) Nearly all of the interviewees extended to the topic of privacy quite early into the interview. The main concerns were that “being followed” generates an uneasy feeling and narrowing down selection was perceived as limiting ones’ freedom of choice, that can even lead to resistance or avoiding certain brands or businesses. In general, the reasoning was in line with what has been suggested previously, e.g. by Aguierre et al. (2016), but also a few new ideas were added. One of the participants used the word “intensity” referring to the situation where the personalization becomes disturbingly overwhelming: I have the loyalty card of the Rimi supermarket and if they put together a shopping list for me there is a very thin line when it becomes annoying. This becomes too intense, they overdo it… (P10, M40). The consumers argued also that acceptability of personalization and data use has category differences. The word “discreet” was used by one of the male interviewees. Frequently marketers fail to understand that there are very private and personal products and services that consumers want to keep secret even from their closest ones. Several participants commented on the fact that personalization tends to overly rely on past data and behaviour. While in many categories past behaviour is a good proxy for future needs, it may not be so in all the categories, especially if to look beyond first necessity. Speaking of AirBnB, it should propose me new destinations, not only these where I have already been…. If they would really understand that I like to travel, and they would look where I have been and they could propose me new destinations based on my preferences, that would be cool…. (P5, F32) So to conclude, in general the Estonian customers’ understanding and attitude towards personalization is in line what has been pointed out in past studies. The positive gains included time saving, making choice process easier and communicating relevant content. The negatives were related to the feeling of being followed and concern about privacy. Also, narrowing down choice or limiting options can be perceived as negative. The more unique aspects highlighted in the study emphasized the need to be discreet, think in terms of product/service category specifics and not to fall in the trap of recommending solutions that the customer has happily purchased before. 4.2 Consumer attitudes regarding disclosing personal data The main aim of the following section was to find out how the consumers think about disclosing personal data and what motivates them to share data with businesses. The secondary aim was to map the customers’ awareness of various digital data collection methods. The results reveal that customers are willing to disclose data in case the related product or service is important to them and there are perceived benefits in data disclosure. Consumers do a quick privacy calculus in their minds. One of the contributions of the current study is the tendency that women are a lot more likely to share their data with businesses. The privacy calculus also included considering whether the required data is really relevant and necessary for the company. There seem to be clear “layers” of data: I am ready to share all the obvious data: my age, gender, city-level location, number of kids, and their ages. Going beyond the basic data sharing is dependent on the client history: If I have been a client for a while and developed a trustworthy relationship, I can share my address. And high-risk information is shared only for specific purposes when there is a very clear benefit for that: Uber and Taxify have got my credit card details so that I can pay automatically (all three quotes P3, F38) Analysing the opinions of the male participants it was more than obvious that they’d like to share as little data as possible. They do not welcome marketing offers and if something is communicated to them personally it needs to be very precisely targeted to their individual profile. Otherwise the companies have no reason to believe that men will pay attention to their messages and offers. Some of the male interviewees claimed to regularly insert fake data to registration forms to avoid any future communication. Men seem to be clearly disturbed by the marketing messages: If I register for a loyalty card and I need to give my e-mail address I will do that. Although I know that it really disturbs me when I get something like this /refers to marketing offers/. I want to take my own decisions (P11, M37) They are careful to make sure that they do not get additional messages and may decide not to join a loyalty program if they fear to receive communication in the future I really pay attention to disclosing data. I can share my ID code if it is really justified, but I will definitely not tick the box to say I am interested in further offers. Frequently I just do not join loyalty programs because there is so much clutter (P9, M46) Although all of the participants expressed their concern about the privacy intrusion and giving away (too much) data they also saw this as unavoidable reality that one has to agree with general terms and conditions. As pointed out earlier the secondary aim of this subsection of the study was also to find out how much knowledge consumers have about contemporary data collection methods. Several participants indicated that they have a feeling that the recent developments in the digital world are difficult to comprehend: I have a feeling that I do not get it and I am not able to handle it. It seems that sometimes I have shared my data without even noticing it (P6, F40). Although the interviewees had noticed information about cookies and also clicked to accept cookies, only a few really understood what cookies do and why they are needed. Consumers give permission to track their activities on the Internet without actually understanding what they have agreed to. Furthermore they consciously decide not to delve into the complicated lingo of the internet small print. I tell you honestly, I do not understand it. I see this ‘cookie-talk’ popping up but I never pay attention. If I am interested in the website I just click “I agree” (P1, F40) The interviewees commented that if the company clearly asks for data and also informs that data will be used for marketing purposes, it actually increases trust and improves consumers’ attitude towards the brand. Informing about the clear purpose why data is used is vital for fruitful marketing. If consumers have given their data for a specific purpose they dislike when it is cross-used for something different. Consumers have had previous negative experience with receiving offers and communication in areas that they have not indicated as a preference themselves and that is not related to their past activities. If I disclose data I need to choose the types of communication I am ready to receive. E.g. if I have indicated that I am into cooking they should no way send me e-mails about garden tools or clothing. It is vital that I indicate and edit my preferences, not someone else (P17, M40). To sum up, the attitudes towards data sharing were varied. It emerged as a clear tendency that women are more willing to disclose data and see more benefits to it. Men, on the contrary, are mostly reluctant to share data and try to find ways to avoid it as much as possible. Different strategies are employed for that – from disclosing fake information to just not doing business with a brand or a website. 4.3 Consumer attitudes regarding personalized marketing communication In general the interviewees had noticed that in depth personalization is still at its infancy. Dominant reasoning was that personalization pays off only with a wealthy clientele and because of the size and economic power of the local market personalized marketing communication is not too widespread in Estonia. Those study participants who had made purchases with global retailers and service providers had a more substantial experience and they were also aware that the ads they see are based on their previous behaviour and purchases. Some of the respondents though had not been thinking why specific ads are served to them even if these match their preferences quite closely. The key expectations to personalization were precisely selected messages/offers and a clear understanding that the messages really are personalized beyond adding person’s name to otherwise generic newsletter. One of the negative aspects participants had experienced was the overly familiar tone of the messages that was not perceived appropriate for a business relationship. Traditional norms of Estonian language prescribe using you (plural) when addressing someone who is not your personal friend and you (singular) is reserved to closer relationships. This distinction is also indicating degree of social distance (Keevallik, 1999) and although younger generations do not always stick to this rule, consumers find it intruding if businesses fail to use the formal you. Positive tone of voice and good humour are warmly welcomed by customers, especially if the business manages to be consistent in its messages. In case consumers realize that something has been personalized based on their profile but they have not clearly given a permission for data use, it makes them really annoyed: I do understand that there is a lot of such data collection that I am not aware, and I will not get to know about, and let it be like that. But if it becomes visible that someone has really been stalking me and ads appear in totally ’wrong’ places for me, it is deplorable (P6, F40) The analysis of the data shows that consumers dislike most cross-usage of data, i.e. when consumers have given their data for one purpose but the company uses it for other purposes. Some consumers are just surprised when they see that activities they have conducted in one channel or website reflect to what they see in other locations, but a lot of consumers feel really annoyed and disturbed. Participants criticized retargeting quite unanimously as nearly all of them had no idea how to stop the ads ‘following’ them long after they had lost interest or need for the product or just had already made the purchase. Retargeting seems to be a favourite tool for marketers and has indeed shown some good business results, but it becomes annoying really quickly. Consumers also realized that retargeting is based on data that is collected without their explicit permission and this realization made them feel uncomfortable and disturbed. In comparison, e-mail marketing and marketing in social media were better received due to the “unsubscribe” options or possibility to “un-like” a page that shares info that is no longer relevant. Although there were polarized opinions on personalized e-mail messages it was generally regarded as the most acceptable channel for sending marketing related information. Minority of participants, all of them male, were strictly opposed to e-mail communication from brands and were annoyed if they had given e-mail address for specific purposes, e.g. informing about the order status, or sharing info regarding specific flight, and later the same businesses had added them to the address base of the newsletter recipients. The majority of participants commented on the low intrusiveness of e-mail marketing: it is easy to delete or unsubscribe and it gives the recipient freedom to decide if they want to interact with the communication or not. Banner ads were regarded as something that you have to withstand. Some of the personalized web advertising was regarded as valuable, especially on social media where targeting is based on user’s hobbies and lifestyle. Interviewees who had experienced precise high quality personalization had also more positive attitudes towards it. They were of opinion that good personalization compensates for giving access to data and losing some privacy. Only one of the participants reported an experience with personalized advertising that violated privacy to extreme: After the Berlin Marathon I opened the website of Postimees /major daily newspaper/ and I was shown a banner with my photo and name advertising the option of ordering the photos. This was extremely unpleasant and scary. How did the photo get there? Will everyone see it? (P3, F38) Some of the interviewees had also experienced personalized messages that were not framed to their personal preferences and induced them to decide not to buy. I had been looking at Jura coffee machines and I have thought to buy it for a really long time. And then they offered that if I make the purchase I get cappuccino cups as a gift. But I do not drink cappuccino and I don’t have anything to do with the cups. I did not buy the machine either (P13, F41) So if companies want to go into personalization, it must be done tactfully, with certain moderation and real knowledge about customers’ needs and preferences. 5 Conclusion Personalization of marketing messages and product/service offerings poses specific strategic challenges. This research provides a qualitative analysis of consumers’ attitudes towards privacy and personalized marketing communication. While part of the results were in line with previous studies, there were quite a few outcomes that add value both theoretically and empirically. Stereotypically (and also scientifically) we may know the differences between males and females as customers, but the current study provided proof that men have a lot more negative attitudes towards personalized communication, do their privacy calculus more rigidly and go to great lengths to avoid giving away their data. Using fake contact details and e-mail addresses are extreme examples of this. The evidence from current research strongly suggests that consumers’ awareness and understanding of the options of data gathering, usage and personalization is low and marketers should not assume that when a client agrees to give access to data, accepts cookies or switches on location tracking, he or she really is able to evaluate the costs and benefits of these activities. Even when consumers have given permission to use data for marketing purposes the outcome may be dissatisfaction and distrust. The assumption that retargeting is an efficient means of advertising as it is based on consumers’ previous activities is strongly challenged by the current research as most of the consumers reported retargeting to be “stalking” and intrusive. It is very difficult to draw a “red line” of privacy violations as consumer tolerance and acceptance of personalization varies to a great extent. It is however obvious, that technological skilfulness of the marketers needs to be well balanced with understanding the consumer sentiment – occasions of very precise and personal communication in unexpected situations was perceived extremely negatively. The current research supports the view that well-crafted personalized marketing messages and advertisements are generally well received. However it would be necessary to pursue further research to understand better the exact situations when a well personalized message will start to appear as intrusive. References Alt, R., Hartwig, M., & Reinhold, O. (2016). Privacy Awareness in Mobile Business: How Mobile OS and Apps Support Transparency in the Use of Personal Data. In Bled eConference, June 19-22, 2016 (pp. 28-43). Maribor: University of Maribor Press. Aguirre, E, Roggeveen, A. L; Grewal, D; & Wetzels, M. (2016). The Personalization-Privacy Paradox: Implications for New Media. The Journal of Consumer Marketing, 33(2), 98-110. doi: 10.1108/JCM-06-2015-1458. Baker, S. E., & Edwards, R. (2012). How many qualitative interviews is enough? Expert voices and early career reflections on sampling and cases in qualitative research. Southampton: ESRC National Centre for Research Methods, University of Southampton. Bakos, Y., Marotta-Wurgler, F., & Trotten, R. R. (2009). Does anyone read the fine print? Testing a Law and Economics Approach to Standard Form Contracts. In CELS 2009 4th Annual Conference on Empirical Legal Studies Paper, 6.10.2009, NYU Law and Economic Research Paper No. 09-40. Chellappa, R. K., & Sin, R. G. (2005). Personalization versus Privacy: An Empirical Examination of the Online Consumer’s Dilemma. Information Technology Management, 6, 181-202. doi:10.1007/s10799-005-5879-y. Clarke R. (1999). Internet Privacy Concerns Confirm the Case for Intervention, Communications of the ACM 42(2), 60-68. Dinev, T., & Hart, P. (2006). An extended privacy calculus model for e-commerce transactions. Information Systems Research, 17(1), 61-80. doi:10.1287/isre.1060.0080. Helberger, N. (2013). Forms matter: Informing consumers effectively. Study commissioned by BEUC. Amsterdam, 2013. Hsieh, H.-F. & Shannon, S. E. (2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15, 1277-1288. doi:10.1177/1049732305276687. Jahng, J. J., Jain, H., & Ramamurthy, K. (2002). Personality traits and effectiveness of presentation of product information in e-business systems. European Journal of Information Systems, 11(3), 181-195 Keevallik, L. (1999). The use and abuse of singular and plural address forms in Estonian. International Journal of the Sociology of Language, 139(1), 125-144. doi:10.1515/ijsl.1999.139.125. Lurger, E., Moran, S., & Rodden, T. (2013). Consent for All: Revealing the Hidden Complexity of Terms and Conditions. Proceedings of ACM CHI 2013: Human Factors in Computing Systems. New York, 2687-2696. McDonald, A. M., & Cranor, L. F. (2010). Beliefs and behaviors: Internet users’ understanding of behavioral advertising. Tech. Rep. Carnegie Mellon University, Pittsburgh. Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks, CA: Sage. Milne, G. R. (2000). Privacy and ethical issues in database/interactive marketing and public policy: A research framework and overview of the special issue. Journal of Public Policy and Marketing, 19(1), 1-6. Murumaa-Mengel, M., Pruulmann-Vengerfeld, P., & Laas-Mikko, K. (2014). Privaatsus inimõigusena ja igapäevatehnoloogiad. Tallinn: Inimõiguste Instituut. Saldaña, J. (2009). The Coding Manual for Qualitative Researchers. Thousand Oaks, CA: Sage. Sheehan, K. B. & Hoy, M. G. (2000). Dimensions of privacy concern among online consumers. Journal of Public Policy & Marketing, 19(1), 62-73. Smith, R. & Shao, J. (2007). Privacy and e-commerce: A consumer-centric perspective. Electronic Commerce Research, 7 (2), 89-116. doi:10.1007/s10660-007-9002-9. Truong, Y. and Simmons, G. (2010). Perceived intrusiveness in digital advertising: strategic marketing implications. Journal of Strategic Marketing, 18 (3), 239-256. doi: 10.1080/09652540903511308. Westin, A. F. (1997). Legal safeguards to insure privacy in a computer society. Communications of the ACM, 10, 533-537. White, T. B., Zahay, D. L., Thorbjornsen, H. & Shavitt, S. (2008). Getting too personal: reactance to highly personalized email solicitations. Marketing Letters, 19(1), 39-50. doi:10.1007/s11002-007-9027-9. Digital Wellness Services’ Servicescape for Young Elderly HANS ALLMÉR 12 CORRESPONDENCE ADDRESS: Hans Allmér, Åbo Akademi, 20500 Åbo, Finland / Linnaeus University, School of Business and Economics, 39182 Kalmar, Sweden, hans.allmer@lnu.se DOI https://doi.org/10.18690/978-961-286-170-4.10 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract Digital devices reshape the servicescape of wellness, helping people, e.g. young elderly, to benefit from digital wellness services (henceforth DWSs) provided in a digital servicescape. This paper builds on five prior studies and explores: (i) what requirements young elderly have, in order to benefit from DWSs, and (ii) what service providers should consider, and pay regard to, when they develop and offer DWSs to young elderly. Hence, this paper argues for DWSs to be subsets of a digital servicescape. This research developed a model for how different parties, e.g. young elderly, their friends and relatives, developers, providers, but also the society per se, could gain from the DWSs. The DWSs create opportunities for a win-win-win situation. Important for the young elderly are motivation, trust, and credibility. Important for developers and providers are to motivate the users, to create trustworthiness, to provide user guidance, and to develop through co-creation. Keywords: • Personalized marketing • Privacy concerns • Personalization-privacy paradox • 1 Introduction Wellness is a conception that combines many aspects on several levels and dimensions. “Wellness is a multidimensional state of being describing the existence of positive health in an individual as exemplified by quality of life and a sense of well-being” (Corbin & Pangrazi, 2001, p. 3). As people grow older, they often want to use new technology even if there could be problems that originate from the fact that young people often design for young people. Still, there are quite a lot of technical devices and applications available for young elderly (60 – 75 years), and seniors (76+), that they would gain wellness from using. Campbell (2015) highlights that even though people age in their own manner, with different health issues to cope with; there are some features about aging applicable to most people. Vision and hearing are reduced, motor skills are declining, and cognition and memory might vary a lot. Further, they have often grown deep relationships over time that they want to keep and maintain, but they are more geographically restricted than younger people generally are. Hence, virtual contacts would be a great opportunity if they would find the technical solutions useful. It is important to keep in mind that elderly people could be used to technology, but could still need some guidance and support. Campbell (2015) describes an older man, who wanted to read further from various links but did not understand how to open them, because he was not used to the interface. He proposes some reminders for those who develop technical solutions for elderly e.g. “avoid small-screen devices”, provide sizing opportunities, “reduce the distance between interface elements that are likely to be used in sequence”, “enable connection with a smaller, more important group of people”, “during longer tasks, give clear feedback on progress and reminders of goals” (ibid.). Wellness is an ongoing process in which different dimensions interact with each other and shape a holistic picture of individuals. Along the years, many have dealt with wellness e.g.:  Dunn (1959, p. 447) proposing that “High-level of wellness for the individual is defined as an integrated method of functioning which is oriented toward maximizing the potential of which the individual is capable, within the environment where he is functioning”.  León et al. (2016, p. 1) stating that wellness is about “dynamic balance of physical, emotional, social, spiritual, and intellectual health”.  Smith Maguire (2007) proposing wellness to aim at offering and achieving a better life in balance.  Hattie, Myers, and Sweeney (2004) adding environmental as a seventh dimension, to the six dimensional model of wellness by Hettler (1976), which originally contained social, intellectual, spiritual, physical, emotional, and occupational aspects. For young elderly, digital wellness services (henceforth DWSs), i.e. apps and devices, could be an appropriate instrument for achieving wellness adapted to individual needs and requirements. Services exist together, and alongside, with physical products and constitute the core of the offer to customers “customers consume and use goods, services or any resource as service to achieve something” (Grönroos, 2015, p. vii). “Service is a simple, yet powerful and multifaceted construct and that it is the correct designation, not only to characterize emerging and converging marketing thought, but also to accurately inform and motivate the associated research, practice, and public policy” (Vargo & Lusch, 2008a, p. 36). Service as a phenomenon is complicated to define as the term has many meanings (Grönroos, 2015). The foundation of a service lies in that it is intangible and therefore the customer cannot see, touch or smell services (Solomon, Marshall, Stuart, Mitchell, & Barnes, 2013). The latter becomes particularly interesting seen in the light of digitalization, and how these potential shortcomings could be dealt with online. Studying wellness and services in connection to social, cultural, psychological, cognitive, and biological aspects of aging would be a start for example by putting forward the importance of assistive technology, where devices and applications can decrease elderly or ill people’s need for help and assistance (Bagwell, 2016). Further, DWSs for young elderly can make a positive impact on their health that will reduce the probability of getting serious illness, and there is a need for developing DWSs (Carlsson & Walden, 2015). Marklund (2016) argues for the importance of a holistic view on wellness, since it is not only one single factor affecting peoples’ wellness, rather a combination of different factors. Exercising both body and brain is crucial and must be done properly; otherwise, there is a serious and increasing risk for illness and premature death. Hence, the research questions are: What demands and needs are required from young elderly as customers of DWSs? What should providers consider when offering DWSs to young elderly? 2 Related work Harris and Goode (2010, p. 231) define e-servicescape as “online environment factors that exist during service delivery”. Hopkins, Grove, Raymond, and LaForge (2009) state that the e- servicescape is a site on the Internet where customers and service providers interact. The digital services can be provided by using various devices e.g. computers, tablets, smartphones, and activity bracelets. A digital servicescape, for enabling physical wellness apps, can be developed and designed for training, sleeping, and measuring various body functions. Apps provide information to the user and can encourage physical training by sending a reminder to the user, or by showing how the users’ fitness develops over time. Intellectual wellness can be strengthened by providing the apps that encourage training of the brain. Emotional wellness are apps providing e.g. possibilities for the user to show emotions like happiness, sadness, loneliness, and participation in other people’s feelings. Chiu and Hu (2015) argue that well designed app technologies can be beneficial to what they describe as older adults as it can strengthen physical and psychological well-being in a cost-effective way. Platt, Outlay, Sarkar, and Karnes (2015) propose the wellness app technology to be an important and promising tool to help people maintain good health, to decrease human suffering and in parallel decrease costs for healthcare. Harris and Goode (2010, p. 231) define e-servicescape “as the online environment factors that exist during service delivery”. Hopkins et al. (2009) argue that the e- servicescape is an Internet site where customers interact with a company. The interaction includes the companies’ design and features. Harris and Goode (2010, p. 239) discuss different terms regarding offers provided on internet, e.g. e-servicescape, online servicescape, and online environment, and state “Nevertheless, we are favourably biased toward the term ‘e- servicescape’, since the label emphasizes the context of online exchange and highlights that purchasing online involves element of self-service, even when products are purchased”. In an e-servicescape, the physical-to-virtual transfer of activity requires providers to rethink the traditional rules for building trust and loyalty in order to meet the needs and demands of prospects and customers (Rivard, 2012). It is important to strive towards message credibility that is comparable with the sender’s credibility, since trust is necessary in an e-servicescape. Otherwise, the customers will probably reject the offered services. Papadopoulou, Andreou, Kanellis, and Martakos (2001) highlight trust, and state that trust are closely connected to both relationship marketing and servicescape. Doney and Cannon (1997) identified five trust- building processes: Intentionality, capability, prediction, transference, calculative; Papadopoulou et al. (2001) added credibility to the list. The trust building processes are: (i) to make a promise, (ii) to enable a promise, and (iii) to keep a promise (Papadopoulou et al., 2001). To create trustfully electronic servicescapes, Papadopoulou et al. (2001) propose that Websites should be transformed to customer-centric e-servicescapes that offer digital experience as a way to develop indelible relationships between business and customers. Trust can be transformed through customer networks by recommendations from others, and lead to customers who are willing to try digital services. Trust will develop over time with increased customer loyalty when customers are repeatedly satisfied with their servicescape interactions. “Each time a promise is made, enabled and kept, it is evaluated with the intentionality, the capability and the credibility process confirming the customer´s trusting beliefs in the business benevolence, competence and credibility. The level of trust is also related to the experience that the customer gains within the e-servicescape” (Papadopoulou et al., 2001, p. 327). The design of user interfaces plays a central role on digital devices (Al-Showarah, Al-Jawad, & Sellahewa, 2014). The interfaces must be designed so that elderly people can benefit from their offerings. The demands and requirements on the digital devices, e.g. smartphones could vary depending on user preferences such as age according to Tsai, Tseng, and Chang (2017). Elderly compared with children and adults, often respond more slowly when using smartphones, which providers need to take into account when developing interfaces. Smith and Chaparro (2015) claim that there are differences between elderly and other groups that must be remembered during development of digital devices for elderly, i.e. elderly use smartphones less than others due to touchscreens’ shortcomings. The size, spacing and location of the target influences the performance and the performance is greater with audiotactile feedback in comparison with tactile and auditory (Hwangbo, Yoon, Jin, Han, & Ji, 2012). Proactive feedback can be seen as guidance if common problems that users experience are taken into consideration, so feedback could be given to provide help in advance. Electronic feedback could increase or decrease the gain of the amplifiers (Black, 1934). “When there are only two parts joined so that each affects the other, the properties of the feedback give important and useful information about the properties of the whole” (Ashby & Pierce, 1957, p. 54) the proactive feedback is built in with the user interface. “The user interface is constantly making progress and ubiquitous nowadays, however its design currently is still lack of concern towards usability and accessibility of the elderly” (Hsiao, Lee, Yang, & Chen, 2017, p. 158). Motivation occurs when an individual directs behaviour towards a goal (Nationalencyklopedin, 2015). Further, it is about forces that initiate, give energy, guide, and maintain behaviour (Vroom, 1964); goal setting is important for task performance whether an individual gives up or completes a task (Locke, 1968). Two behavioural elements deal with motivation; the individual’s choice of path among alternatives and amplitude / vigour of the action, for its tendency to persist over time in the chosen path (Atkinson, 1957). There are three types of motivation: Internal motivation, external motivation, and amotivation. The psychological needs mediate between the first and the third element, and are of three types: Autonomy, competence, and relatedness. The outcomes are of three types: Affective, cognitive, and behavioural. Motivation can be divided into three main groups: Internal motivation, instrumental motivation, and motivation as an interaction between the individual and the environment (Alvesson, 2013). Internal motivation emphasizes the individual's internal needs and forces; instrumental motivation stresses a perspective of individuals trying to achieve rewards and avoiding punishments, and an interactive motivation approach involving norms, reciprocity, and identity. In the human mentality three levels occur: Human nature, culture, and personality. Human nature is universal and inherited, culture is for a specific group / category and is learned, and lastly personality is specific to the individual and is inherited and learned (Hofstede, 1994). 3 Method This research carried out five studies, in which three methods were applied; the first study, one survey and a focus group; the second study, two surveys and two focus groups, the third study, a survey, the fourth study, a literature review, and the fifth study, a focus group. The first three respondent groups consisted of different network members. There were 32 respondents in the first group, 15 in the second, and 16 in the third group, in total 63. The first group was mostly retired people, who bowled together weekly. The second and third groups contained two different groups of retired people, who played boule together. Three surveys and focus groups conducted in 2012 and in 2013, collected data about the respondents and about their IT use and preferences. A literature review was conducted with the purpose of finding patterns about what is put forward regarding servicescape, e-servicescape, and information system. A complementary survey, and yet another focus group (D), based on the above studies, were carried out. 4 Findings The four surveys concerned Internet services, i.e. whether or not the respondents were interested in making use of others’ knowledge, and of sharing their own knowledge with others. Generally, the respondents were not particularly interested in sharing their knowledge online, except as regarded knowledge connected to their professional skills (49 %) and travelling (56 %) (Allmér & Råberg, 2013). The results from the interview with focus group A, showed that the group mainly used the Internet for searching for information, and that very few shared information, knowledge and/or opinions. They commented that in some cases they were forced into using services on the Internet (e.g. bank services, healthcare services, and communication with organizations and authorities). All of the respondents mentioned bank services, since many banks are closing their physical offices, decreasing opening hours and personal services, increasing the costs for services in bank offices and more, all of which are forcing customers to use Internet services whether they want to or not (Allmér & Råberg, 2013). Group B focused on why they do not utilize services available through the Internet today. The results showed that they were extremely concerned about having a personal connection with the people taking an active part in services on the Internet (e.g. when discussing health issues they want to see, and have direct communication with the expert/doctor, and preferably, a history with that person). They needed to feel confident in whom they interact with, what kind of knowledge that person has, what kind of previous record of accomplishment that person has, and what will happen with the information they are willing to share and discuss? They also felt worried about entering new services, as a whole, even if it was just for trial reasons (Råberg & Allmér, 2013). Group C felt anxiety towards trying digital services and sharing information, but also worried about knowledge and opinions on digital service applications and social media. Primarily three things concerned the young elderly in relation to previous results. Firstly, they needed to feel secure about their data, identity, username, password, credit card number, comments and knowledge in that it would not be abused and/or hijacked in any way. They also needed to be secure about that they are communicating with the right person, and about that person’s good intentions. Secondly, they needed a personal touch and guidance about digital services. Personal touch means recommendations from acquaintances, who they could contact before using the service, or could ask for help to find the right kind of service. Thirdly, they thought some service applications were so general and offered too many options, which only complicated things. Hence, they wanted to have the ability to decrease the number of options to solely those alternatives that they really wanted. They wanted to customize the service options through interaction with the service provider (Råberg & Allmér, 2013). For group D the top-three reasons for using were help and support with technology, professional knowledge, planning, scheduling, and being reminded. The top-three reasons for sharing were professional knowledge, memory training, and health advice. The three of least interest for using were (not / no answer) health advice, technology help and support, and planning, scheduling, and being reminded. The three of least interest for sharing were intermediation of work, technology help and support, and training coaches. Some conclusions could be made (Allmér & Marcusson, 2018):  They have skills, such as knowledge and experience, as well as time to help other people. They do so in clubs and societies or with friends and acquaintances. When the parties know each other, the offers can be clear, transparent, and distinct. When offers are addressed to people outside these zones different problems can occur. The young elderly are keen to feel secure in a support situation and with a beneficiary.  It was shown that clarity in how the offer is disseminated is of importance for the young elderly.  There are two types of support that may be offered; time and skills. Time can for example be used for a walk in the park, a wheelchair walk, reading aloud from a magazine or a book, and as company to a museum, theatre, or concert. Skills can be knowledge and experience that are professional (e.g. tax returns) or that originate from hobbies and interests (e.g. fruit tree pruning and fixing the car).  The young elderly need to feel secure in a support situation. They also argue that the person who receives support must bring about the type of security they need, which depends on several factors, e.g. are the persons known to each other, are there only the two persons in the situation, is it in one party's home, or is the beneficiary disabled? To strengthen this feeling of security an intermediary can play a role. The literature review resulted in the supposition that a digital servicescape is possible. However, in some cases a detour is necessary since it is not possible to apply the senses smell, touch, and taste digitally (Allmér, 2014). 5 Discussion This section discusses young elderly’s views on DWSs (5.1) and presents a model of digital wellness services’ servicescape (henceforth DiWeSS) (5.2). 5.1 Young elderly’s views on DWSs The findings have clearly shown that the young elderly commonly use IT, have Internet access, and are worried about the risks and threats that lie in wait for them online. They showed scarce interest to utilize others’ knowledge and to share their own knowledge with others, with few exceptions. Hence, from this starting point it is obvious that the actual use of DWS solutions comes from some kind of motivation (Alvesson, 2013; Atkinson, 1957; Hofstede, 1994; Locke, 1968; Vroom, 1964) and it is just as obvious that usage depends on trust (Doney & Cannon, 1997; Papadopoulou et al., 2001; Rivard, 2012) for the provider, who are expected to create a safe and secure environment. It is utterly important that the young elderly feel that it is both secure and safe to use DWSs. Young elderly use digital equipment in their daily life and have access to Internet, which means that the technology is no obstacle per se; it is rather their feelings towards the consequences that might come from the usage. The worries connected to sharing online, and to using for example apps, must be overcome in order to make use of the possibilities that actually are available, and that can come in handy for a lot of purposes and reasons. Obstacles like anxiety about trusting others, including the technical facilities and apps online, must be handled from the providers’ side for the young elderly to overcome their resistance and hesitation. What a proposition conveys must be obvious to the customer, i.e. to the young elderly, which is crucial for the provider to understand and to act on. To offer reliable and thoughtful guidance (Ashby & Pierce, 1957; Black, 1934; Hsiao et al., 2017), as well as a good usability (Al-Showarah et al., 2014; Hwangbo et al., 2012; Smith & Chaparro, 2015; Tsai et al., 2017) would make this possible, according to the research findings. One way of achieving this is to co-create trust and motivation with the young elderly, to invite them to participate. Group D stresses the necessity of guidance / support with technological issues as something that really has to be carefully handled. A provider should meet these demands, and by that also create trust (Doney & Cannon, 1997; Papadopoulou et al., 2001; Rivard, 2012). Digital servicescape designed to provide good usability (Al-Showarah et al., 2014; Hwangbo et al., 2012; Smith & Chaparro, 2015; Tsai et al., 2017) and guidance (Ashby & Pierce, 1957; Black, 1934; Hsiao et al., 2017) are likely to create security, trust (Doney & Cannon, 1997; Papadopoulou et al., 2001; Rivard, 2012), and motivation (Alvesson, 2013; Atkinson, 1957; Hofstede, 1994; Locke, 1968; Vroom, 1964) to use. It should be very clear that the responsibility to fulfil these demands is in the hands of the provider, who should meet these demands to the fullest. 5.2 DiWeSS’ context model The purpose was to put together findings from five previous studies (four papers / articles) regarding young elderly’s DWSs into a model. Four research papers’ key findings end up in a focusing on digital servicescape that is further developed into a focus on providing DWSs (apps and devices). In a digital servicescape, the provider and the customer / user meet in a moment of truth and a context of trust, and this is crucially important. It is in the digital servicescape the clash occurs, and therefore where the services offered will be accepted or rejected. The interface of the digital servicescape should emulate a place between the provider and the customer / user that works almost like a Velcro in order to build a sustainable relation between the different parts involved. On this basis, it can be argued for a need for a DiWeSS that emerges and appears as a subset of the e-servicescape. A digital servicescape has in comparison with a physical servicescape some limitations when it comes to what is possible to present (e.g. the sense of smell, touch and taste) on for instance a computer, tablet, or a smartphone. However, it could be argued that it is plausible to provide the feeling of a smell, touch, and taste even if it is not possible to the full extent. This in turn, affects the possibilities to provide DWSs, as it is hard to show people what the service actually can bring them. The DiWeSS’ context model (figure 1) puts forward three perspectives: on the one hand the developer and provider, and on the other hand, the young elderly. These three parts interact where the provider offers the DWSs with DiWeSS to the young elderly. Figure 1: DiWeSS’ context model DWS with DiWeSS points towards a win-win-win-situation. On one side are the developers and providers, and on the other side are the young elderly with their needs in everyday life. Further, the part about how these needs are met with DWSs should be included in the model. To some extent, e.g. family, friends, and society among others, can also be considered as stakeholders. Companies and organizations that provide digital services, which meet the demands of customers, have the opportunity to make a profit, which means a win situation for them. Well working DWSs with DiWeSS will help people to have a more eventful, healthier and longer life, which certainly can be seen as a win for them. People that are healthier and satisfied with their life can have a positive influence on other people and therefore demand less service from society, which will be a win for the society per se. 6 Conclusions, limitations and further research The first research question is about needs and requirements of the target group and the second research question is about demands on providers, both in the context of DWSs for young elderly. A short answer to these questions is presented in a program (figure 2). With the purpose to underline the keys to DiWeSS for young elderly, the most important issues are brought together in a program. Figure 2: DiWeSS’ program If DiWeSS works for young elderly, it probably will work for other age groups as well. There are some implications for practise, and for further research, to highlight and discuss.  Firstly, there is a need for deeper understanding of the benefits that a well-designed DiWeSS can bring to people in general, and to the group of young elderly in particular, as most people eventually will be among the group of young elderly.  Secondly, there is a need to put research effort into the group of seniors, as it can be beneficiary to them too.  Thirdly, there is an implication for putting effort into research on if it could lead to a win – win – win situation if young elderly are provided with DWSs through DiWeSS. If so, how should it be constructed? For everybody it is nicer to get old if you are in good shape; to get a few more good years. A well-designed servicescape would be most helpful in order to support the use of DWSs. Therefore, use DWSs with DiWeSS! Acknowledgement I am grateful to the reviewers for their comments. References Al-Showarah, S., Al-Jawad, N., & Sellahewa, H. (2014). Effects of user age on smartphone and tablet use, measured with an eye-tracker via fixation duration, scan-path duration, and saccades proportion8514, 3-14. Retrieved from doi:10.1007/978-3-319-07440- 5_1 Allmér, H. (2014). E-servicescape is plausible. IRIS Selected Papers(5). Allmér, H., & Marcusson, L. (2018). The five-key element model for altruistic young elderly. (In review). Allmér, H., & Råberg, M. (2013). Young-elderly and digital use. Paper presented at the IRIS 2013 Conference, Oslo, Norway. Alvesson, M. (2013). Organisation och ledning: ett något skeptiskt perspektiv. Lund: Studentlitteratur. Ashby, W. R., & Pierce, J. R. (1957). An Introduction to Cybernetics. Physics Today, 10(7), 34-36. doi:10.1063/1.3060436 Atkinson, J. W. (1957). Motivational determinants of risk-taking behavior. Psychological Review, 64(6p1), 359. Bagwell, D. K. B. S. (2016). Geriatrics and Gerontology. In. Ipswich, USA: Salem Press. Black, H. S. (1934). Stabilized feed-back amplifiers. Electrical Engineering, 53(4), 590-590. doi:10.1109/EE.1934.6539889 Campbell, O. (2015). Designing For The Elderly: Ways Older People Use Digital Technology Differently. Smashing Magazine. Retrieved from (2018-02-20) https://www.smashingmagazine.com/2015/02/designing-digital-technology-for-the- elderly/ Carlsson, C., & Walden, P. (2015). Digital Wellness Services For “Young Elderly” A Missed Opportunity For Mobile Services. AIS Electronic Library (AISeL). Chiu, C., & Hu, Y. (2015). Promoting Older Adults’ Well-Being Through Apps-based Technology Training Services: A Mixed Method Approach. The Gerontologist, 55(Suppl2), 758-758. doi:10.1093/geront/gnv395.04 Corbin, C. B., & Pangrazi, R. P. (2001). Toward a Uniform Definition of Wellness: A Commentary. Washington D. C.: President's Council on Physical Fitness and Sports Research Digest. Doney, P. M., & Cannon, J. P. (1997). An Examination of the Nature of Trust in Buyer-Seller Relationships. Journal of Marketing, 61(2), 35-51. doi:10.2307/1251829 Dunn, H. L. (1959). What high-level wellness means. Canadian journal of public health = Revue canadienne de santé publique, 50, 447. Grönroos, C. (2015). Service management and marketing: managing the service profit logic (4 ed.). Chichester: John Wiley & Sons. Harris, L. C., & Goode, M. M. H. (2010). Online servicescapes, trust, and purchase intentions. Journal of Services Marketing, 24(3), 230-243. doi:DOI: 10.1108/08876041011040631. Hattie, J., Myers, J., & Sweeney, T. (2004). A Factor Structure of Wellness: Theory, Assessment, Analysis, and Practice. Journal of Counseling and Development : JCD, 82(3), 354-364. doi:10.1002/j.1556-6678.2004.tb00321.x Hettler, B. (1976). The six dimensions of wellness. Retrieved from (2017-03-05) http://www.hettler.com/Origins.htm Hofstede, G. (1994). Cultures and organizations: software of the mind: intercultural cooperation and its importance for survival. London: Harper/Collins. Hopkins, C. D., Grove, S. J., Raymond, M. A., & LaForge, M. C. (2009). Designing the e- Servicescape: Implications for Online Retailers. Journal of Internet Commerce, 8(1/2), 23-43. doi:10.1080/15332860903182487 Hsiao, S.-W., Lee, C.-H., Yang, M.-H., & Chen, R.-Q. (2017). User interface based on natural interaction design for seniors. Computers in Human Behavior, 75, 147-159. doi:10.1016/j.chb.2017.05.011 Hwangbo, H., Yoon, S. H., Jin, B. S., Han, Y. S., & Ji, Y. G. (2012). A Study of Pointing Performance of Elderly Users on Smartphones. International Journal of Human- Computer Interaction. doi:10.1080/10447318.2012.729996 León, M., Nieto-Hipólito, J., Garibaldi-Beltrán, J., Amaya-Parra, G., Luque-Morales, P., Magaña-Espinoza, P., & Aguilar-Velazco, J. (2016). Designing a Model of a Digital Ecosystem for Healthcare and Wellness Using the Business Model Canvas. J Med Syst, 40(6), 1-9. doi:10.1007/s10916-016-0488-3 Locke, E. A. (1968). Toward a theory of task motivation and incentives. Organizational behavior and human performance, 3(2), 157-189. Marklund, B. (2016). [Conversation over health-related issues 2016-10-27]. Nationalencyklopedin. (2015). Different Dictionary Words In. Papadopoulou, P., Andreou, A., Kanellis, P., & Martakos, D. (2001). Trust and Relationship Building in Electronic Commerce. Internet Research, 11(4), 322-332. doi:10.1108/10662240110402777 Platt, A., Outlay, C., Sarkar, P., & Karnes, S. (2015). Evaluating User Needs in Wellness Apps. International Journal of Human-Computer Interaction. doi:10.1080/10447318.2015.1099803 Rivard, S. (2012). Information technology implementers' responses to user resistance nature and effects. Management information systems : mis quarterly, 36(3), 897-920. Råberg, M., & Allmér, H. (2013). Young-elderly and digital services. Paper presented at the IRIS 2013 conference, Oslo, Norway. Smith, A. L., & Chaparro, B. S. (2015). Smartphone Text Input Method Performance, Usability, and Preference With Younger and Older Adults. Human Factors: The Journal of Human Factors and Ergonomics Society, 57(6), 1015-1028. doi:10.1177/0018720815575644 Smith Maguire, J. (2007). Fit for consumption: sociology and the business of fitness. Milton Park, Abingdon, Oxon; New York: Routledge. Solomon, M. R., Marshall, G. W., Stuart, E. W., Mitchell, V., & Barnes, B. (2013). Marketing: real people, real decisions. Harlow: Pearson. Tsai, T.-H., Tseng, K. C., & Chang, Y.-S. (2017). Testing the usability of smartphone surface gestures on different sizes of smartphones by different age groups of users. Computers in Human Behavior, 75, 103-116. doi:10.1016/j.chb.2017.05.013 Vargo, S. L., & Lusch, R. F. (2008a). Why "service"? Journal of the Academy of Marketing Science, 36(1), 25-38. Vroom, V. H. (1964). Work and motivation. NY: John Wiley & Sons. Developing an E-learning Course on Serious Games: A Multidisciplinary Approach TANJA KORHONEN, TEIJA RAVELIN & RAIJA HALONEN 13 CORRESPONDENCE ADDRESS: Korhonen, Ph.D. student, Senior Lecturer, Kajaani University of Applied Sciences, School of Information Systems, Ketunpolku 1 87101 Kajaani, Finland, e-mail: tanja.korhonen@kamk.fi , Ravelin, Ph.D., Senior Lecturer, Kajaani University of Applied Sciences, School of Health, Ketunpolku 1 87101 Kajaani, Finland, e-mail: Teija.ravelin@kamk.fi, Halonen, Ph.D., Adjunct professor, University of Oulu, Faculty of Information Technology and Electrical Engineering, 90014 Oulu, Finland, raija.halonen@oulu.fi DOI https://doi.org/10.18690/978-961-286-170-4.11 ISBN 978-961-286-170-4 © 2018 University of Maribor Press Available at: http://press.um.si. Abstract The use of serious games (SGs) for different purposes is increasing. The development of successful and effective SGs requires a team with professional knowledge from different areas and a joint understanding of the SG development process. The research goal was to find a way to educate multidisciplinary students via an e-learning course about SGs. An action research study was initiated, and social constructivism theory-based e-learning courses on SGs were developed. The Carpe Diem Learning design method was used in the planning phase, followed by three iterations on course development and implementation. The feedback and observations were analysed in each iteration, and changes to the courses were made according to them. This iterative development and incremental improvement of an online course according to feedback was found to be a good approach. Keywords: • Serious Games • E-learning • Education • 1 Introduction The primary objective of a serious game (SG) is more than simply entertainment (Djaouti, Alvarez, Jessel & Rampnoux, 2011; Susi, Johannesson & Backlund, 2007; Zyda, 2005). These games, also called applied games, can be used to either change a player’s behaviour or motivate them in a particular way (Baranowski et al., 2013; Ryan, Rigby & Przybylski, 2006). Due to the many different potential applications of SGs (see, e.g. Zyda, 2005; Kemppainen, Korhonen & Ravelin, 2014; Ricciardi & Paolis, 2014), their development requires the involvement of experts from a variety of disciplines. To develop effective SGs, meaning they effectively achieve their serious purpose, more education is needed (De Troyer, 2017). In this study, our research goal was to find a solution to educate a multidisciplinary audience about SGs. The Internet and rapidly developed information and communication technology (ICT) have built new methods of education and learning (Gomez, Wu & Passerini, 2010). Especially, the role of e-learning has grown as part of the acquisition of new skills and knowledge. Despite its technical requirements, the pedagogical aspects should be noted, as well. To evaluate the effectiveness of e-learning, Yunus and Salim (2008) proposed a framework consisting of five dimensions (individual, knowledge, learning, content, and relationship between the instruction and the learner) to enable e-learning. The main research question in this study was: How to develop an e-learning course on SGs for multidisciplinary students? To answer this question, the following sub-questions were asked:  How to design an e-learning course with social constructivism theory as an approach?  What kind of content is needed regarding SGs when designing a course for multidisciplinary students?  How to utilise gamification to motivate students? This research followed action research techniques. An action research study aims to solve current practical problems while expanding on scientific knowledge, typically using an iterative research process where researchers are part of the study (Baskerville & Myers, 2004; Baskerville, 1999; Coghlan & Brannic, 2014). This paper describes an action research study of the development and implementation of e-learning courses on SGs, and it is organised into five major sections. The next section reviews related works. Section three presents the research methodology, and the results are presented after that. The paper ends with a discussion and conclusions. 2 Related Work 2.1 Pedagogical Background ICT-based communication increases learning by activating students, and team-based methods in online environments improve the effectiveness of learning (Gomez et al., 2010). Third-generation distance education, meaning that student support is integrated with teaching, giving students possibility to create learning content instead of adopting ready-made content, is a radical change for pedagogy. Also educational institutions are challenged to adopt technology that develops at a rapid pace. (Tait, 2014.) The concept of learning related to this development work is based on social constructivism theory. In social constructivism theory, social interaction between people is seen as the primary source of cognitive development in a learning context. (Durairaj & Umar, 2015.) Knowledge is constructed by exchanging dialogue in a social setting and co-constructed in the environment with others (Shaikh et al., 2017). Students acquire knowledge by interacting and collaborating with a knowledgeable individual. Involvement and interaction in group discussions provide students with the opportunity to generalise and transfer their knowledge. Social constructivism theory highlights the importance of feedback to help students construct their knowledge. (Durairaj & Umar, 2015.) Although learning happens through collaboration, it is an internal mechanism within the individual. Individual learning is a product of knowledge creation, and it happens when collaboration takes place and when knowledge itself is co-created in the environment. (Shaikh et al., 2017.) The social constructivism approach has been applied in online courses (Shaikh et al., 2017). Online learning and self-study are seen possible in the social constructivism context because the learner has inherited evolved cultural tools deliberately devised for intentional learning (Marsh & Ketterer, 2005). The tension between the synchronous and asynchronous delivery of online courses is recognised. Shared deadlines result in improved motivation, and in contrast, asynchronous delivery gives students more freedom. (Shaikh et al., 2017.) Considering the dropout rate, Onah, Sinclair and Boyatt (2014) note that in most MOOC (Massive Open Online Course) courses, less than 13% of students finished the course. In online courses, it is challenging to provide individual help to students, and to guarantee that the course outcomes have been achieved. Crowdsourcing has been seen to have the potential to answer these questions. Crowd formation in an online setting related to specific knowledge domains can connect students and make interacting and collaborating with knowledgeable individuals possible. (Shaikh et al., 2017.) 2.2 Design Method Kali, Levin-Peled and Dori (2009) reported about the challenges to design courses that promote collaborative learning in higher education. According to them, teachers willing to adopt socio-constructivist pedagogies often face challenges related to fulfilling their objectives without the thorough application of carefully planned design methods. They also encouraged teachers to formulate and refine guidelines to fit with the objectives and to find sensible methods in the design. Carpe Diem Learning design method describes a team-based, collaborative online learning design process, that typically comprises two or three academics (Salmon et al., 2014; Salmon & Wright, 2014). This method is suitable for designing new courses. Carpe Diem focuses on learning design needs for specific units of study. The basic process is delivered in a two-day workshop, and the basic idea is that people are designing something that could be put into immediate use with participants. By the end of the workshop, there will be a unit or module partly built in the online environment and an action plan to support it. (Salmon & Wright, 2014.) 2.3 Serious Games as a Topic The concept of a SG involves a digital game whose main purpose is something other than pure entertainment, and it is designed to be used in training, education, and healthcare (Loh, Sheng & Ifenthaler, 2015). Figure 1: Phases and actors in SGs development. Developing a SG requires not only game development skills and an understanding of good game design, but also the ability “to solve an organizational need or to be of utility in some other aspect beyond entertainment” (Backlund et al., 2017, p. 15). SG development is seen as an interdisciplinary scene (see Fig. 1.), where input is needed from many domains (Backlund et al., 2017). Client procurement skills are seen as a challenge: they have significantly high or low expectations, and they lack knowledge of possible solutions regarding gamification and SGs (Backlund et al., 2017). The education and consultancy of customers could be considered a solution for this challenge. De Troyer (2017) also presents the need for multidisciplinary tools to assist during the development of SGs, guidelines for developing SGs, and knowledge of SGs. Several disciplines should be involved when designing and implementing SGs. Professionals participating in the development process should acquire knowledge and an understanding of the SG development process. (Korhonen et al., 2017.) The term multidisciplinary expertise refers to the gathering of professionals from multiple academic disciplines to solve collaboratively a defined problem (Nancarrow et al., 2013). Furthermore, game development requires a team of experts with different backgrounds, and the development of SGs requires the participation of professionals in the target content area (Kemppainen et al., 2014; Korhonen & Halonen, 2017a). Figure 1 illustrates the aspects of SG design: SGs are designed using basic game design techniques, but for the defined purpose (Loh et al., 2015) and users. 3 Research Methodology Action research aims to solve current practical problems while expanding scientific knowledge, typically using an iterative research process where researchers are part of the study (Baskerville & Myers, 2004; Baskerville, 1999). As a concept, action research was introduced by Lewin (1946), who studied workers’ intergroup relations and actions, as well as the barriers the workers met at work. Specifically, action research is an effective research method in cases that include active participation in organisations that are in the process of changing and that raise research interest among researchers. In practice, the change can be carried out by professional researchers who want to develop their work processes and practices in their environment. Action research appears in a cyclical process, and it is collaborative in nature. (Coghlan & Brannic, 2014.) Figure 2: Action research cycle in developing an e-learning course on serious games. In the current study (see Fig. 2), two of the authors were involved in developing the course as part of the organisation and its practitioners: one from information systems with technical background in gaming and one from healthcare expertise area. They observed the overall process of developing the course, and they participated in practical work as developers (see Coghlan & Brannic, 2014). 4 Development of the Course on Serious Games This section presents the phases in the development of the course on SGs that apply the cycle of action research (see Fig. 2). Planning started with research on the-state-of-the art in SG courses, and it continued by applying the Carpe Diem Learning design methodology. This was followed by content selection and creation in a learning management system (LMS), including the use of gamification techniques. Observation took place during three implementation phases, and all feedback was analysed and used in the iterative development of the course. 4.1 The State-of-the-Art in Serious Game Education At the time of this study, year 2016, there were few online SG courses and education available in English, and the quality and content of the courses varied greatly. There were no courses in Finnish. After searching using Google, we listed courses that were found in English and offered as online courses. Then the search was narrowed down to courses that suited also to experts outside game development (see Table 1). Table 1: Available SG Online Courses Course name Table of available online SG courses - top3 Link Syllabus Remarks Serious Gaming https://www.coursera.org/learn/serious-gaming “You will learn the ins and outs of games that are designed with exactly those purposes in mind: serious games.” Most suitable for a multidiscipli-nary audience. Serious Game Design and Development https://iversity.org/en/courses/serious-game-design “Students will go through all the steps of the development of a Serious Educational Game.” Ended 2015. Design and Development of Games for Learning https://www.edx.org/course/design-development-games-learning-mitx-11-127x-0 “We will hear from experts working on many aspects of learning games from design to development to implementation.” No verified certificate available. Table 1 describes the three courses that were offered as online courses on SGs and considered to suite for non-game development audiences. 4.2 Planning the Course In planning the online course in higher education (HE) level on SGs, a Carpe Diem Learning design method (see Salmon et al., 2014) was applied. This method was chosen because it is effective, creative, student-centred, and multidisciplinary teamwork was needed in planning the brand-new course. The process started with the guidance of the facilitator. There were six steps in the process (see Salmon et al., 2014): 1. Write the blueprint – envision the future. 2. Make a storyboard – become a designer. 3. Build your prototype online. 4. Check reality. 5. Review and adjust. 6. Plan your next step. First, HE healthcare and sports bachelor's students were selected as the pilot course's primary target group. While writing the blueprint, the mission for the online course from the students’ viewpoint was defined. The mission was set (see Table 2), and the look and feel of the course was described as creative, playful, professional, and engaging. After that, it was time to define what was supposed to be achieved after the course: “By the end of the course, I will be able to generate ideas concerning SGs, and I will know how to go further with them.” Then, it was explored how to assess these outcomes and how to fulfil the promises made. For example, to be engaging, course methods included peer-reviewing and rapid feedback. For assessing knowledge, there were online tests and a game evaluation assignment. The learning objectives of the course are described in Table 2. Table 2: The SGs course plan. Online course on SGs Target group Primary: Higher Education healthcare and sports students (Bachelor’s level). Secondary: All interested students and professionals The mission “I have courage to generate SG ideas and go further with my ideas.” Learning objectives Student will.. 1. get familiar with the concept and the development process of serious games and gamification. 2. find out about the owner's role in a serious games development project and learn a variety of innovation methods for developing your game ideas. 3. learn to understand the technological perspective behind gamified solutions. The methods Gamification, peer discussions, peer-reviewing, online tests. During the storyboard phase, the progression of the course was planned. The ways how gamification (defined as “using game design elements in non-game contexts” (see Deterding et al., 2011)) can be used to motivate students. One gamified element was proceeding in LMS from level to level (see Fig. 3). Each level included a certain assignment to complete before proceeding to the next level. Social interaction was emphasised during the course via peer discussions included in the assignments. The delivery of the course was both synchronous and asynchronous. At first, the timetable for proceeding from level to level was set, but it was considered to limit the students. Thus it was possible to proceed at one’s own pace but via the level structure of the course, the social interaction was ensured. 4.3 Content of the Course The course content was selected so that students with a limited background in and knowledge of games could participate. The learning material covered videos, PowerPoint presentations with audio, web pages, and scientific and popular articles. The course could be studied independently, online, within the assigned timeframe, and the tasks required collaboration with peer students. The assignments consisted of group discussions, playing games, a competence test, a game evaluation, and an assignment where students brainstormed their own SGs. A literature research task was added to the final implementation. The basic course structure with levels is illustrated in Figure 3. Figure 3: Basic Structure of the Planned E-learning Course on SGs. Level 1 started with the definition of general terms related to games, gamification, and SGs. After the completion of an assignment at level 1, level 2 opened automatically. This part of the course consisted of deeper information about games, such as genres, different types of SGs, and game design and production, as well as how to act as a product owner or client in a SG project. In level 3, the content was more oriented toward SGs, especially regarding attitudes and motivation. 4.4 Iterative Course Development: Implementation, Observation, and Reflection The course was built to the Moodle LMS and it proceeded from level to level according our storyboard. Some parts, like visual layout, were not possible to realise in the Moodle. The total study period was planned to be 2.5 months for 3 ECTS (European Credit Transfer and Accumulation System), which equates 81 hours of HE students’ work. The teachers’ role was active at the beginning, in the middle, and at the end of the course. Before releasing the course, feedback was given by colleagues, and some changes were made. The course had three implementations with observation and reflection phases. The first implementation of the course in Finnish was realised as open online studies during summer 2016. A part of the material was delivered in English. The professional backgrounds of 21 enrolled students varied greatly. Among participants, there were, for example, nursing, healthcare and sports students, librarians, IT specialists, teachers, and physiotherapists. Eight of the students completed the course on time (38%, see Fig. 5), and 15 of them gave feedback, which is described in Table 3. The second implementation of the course was realised between autumn 2016 and winter 2017 on a new learning platform, a Moodle-based LMS, especially designed to serve open education and cross-institutional studies. This course was targeted to a multidisciplinary audience. In total, 28 participants signed up, and 25 of them completed the course (89%, see Fig. 5). Table 3: SGs course implementations during 2016-2017. Implemen-tation of the open online course on SGs (ECTS) Parti-cipants enrol-led (no.) Participants that finished the course/ passed (%) Positive feedback Critical feedback and suggestions Summer 2016, (3 ECTS) 21 38% Peer discussions Positive attitude towards gaming Finding new interesting games to play Clear and beneficial assignments Communication Outlook of the course Need for more teacher support Activating interventions Clearer and approachable materials Autumn 2016 – Winter 2017 (3 ECTS) 28 89% Course content More material in Finnish. Better timetabling of assignments A chat channel for real-time peer communication The removal of technical problems Summer 2017 (local language) (5 ECTS) 156 29.5% Suitable for studying SGs The assignments were considered useful and helpful Peer work Guided tutorial More feedback from the teachers Peer work Gamified tools (collecting coins) Summer 2017 (English) (5 ECTS) 102 11.8% Suitable for studying SGs Clearer information about assignments Help in technical issues Table 3 sums up the main feedback from the course. The course was stated to be suitable for studying SGs and the positive attitude related to gaming was mentioned as one positive remark. Students needed more clearer instructions in some assignments, better or free timetable of assignments, and reported technical problems specially at the beginning of the course. Teacher’s presence was wished (activating interventions, feedback, support) even when it had been stated that the course was independently studied on-line course. Peer discussions were seen both as positive and negative issue. A few game companies were interviewed and their viewpoint regarding the content required to educate adequate SG experts were asked. Certain main categories emerged from companies’ responses: game technology; business-related topics, such as budgeting, sales, and marketing; service design; and understanding the entire nature of game development, which involves multidisciplinary teamwork. The course was refined to bring closer subject area and game development experts. The third implementation (see Fig. 4) took place during summer 2017, one course in Finnish and one in English. The course was designed as a self-learning, online course, and it was upgraded to 5 ECTS. It was refined based on the feedback given by former course students, especially related to timetabling and technical problems. A chat for peer communication was also created. Besides the existing level structure, new tools for gamification were added: a progress bar that showed the activity completion, a top active student list and a stash that meant collectable coins (six) hidden in the course area. Figure 4: Final course on new LMS (part of the user interface). In total, 156 participants signed up to participate in the course in Finnish, and 46 (29.5%) of them completed the course (see Fig. 5). As well, 102 students joined the English course, and 12 (11.8%) of them finished with an accepted grade. Some of the students signed up to take both of courses and completed only one of them. Because courses were offered freely in national level for HE or professionals without any fee as summer studies, the interruption rate was quite high. Table 3 describes the main feedback. 21 28 156 102 8 25 46 12 SUMMER 2016 AUTUMN 2016 SUMMER 2017 LOCAL SUMMER 2017 ENGLISH Signed up Completed Figure 5: Participants, that signed up and completed SGs courses in 2016-17. 5 Discussion and Conclusions The purpose of the study was to identify how to build an e-learning course on SGs for multidisciplinary students. Earlier knowledge revealed the need for experience in several disciplines, including understanding behavioural (Baranowski et al., 2013), pedagogical (Durairaj & Umar, 2015; Shaikh et al., 2017), and expertise (Kemppainen et al., 2014) -related aspects. Additional challenges were related to the online environment (see Gomez et al., 2010) and using two languages. To answer the main research question related to developing an e-learning course and to identify a way to educate multidisciplinary audiences about SGs, an action research project was initiated. The solution was sought by first analysing state-of-the-art knowledge and then planning the course. After that, content for the course was defined, and finally, an iterative approach was applied in the development of the three implementations. Social constructivism theory was used as a pedagogical approach in designing an e-learning course on SGs. The social interactions and dialogue between students were based on peer discussions and chats. Knowledge construction in a social setting was one of the main ideas of the course, and crowdsourcing was used by taking advantage of the multidisciplinary knowledge of the students. Social constructivism theory emphasises the importance of feedback, and in this course, the feedback method was peer feedback. The idea of intentional learning was the basis in planning the independent conduct of the course (Durairaj & Umar, 2015; Shaikh et al., 2017). The heterogeneous student group online brought up new challenges for the content: different backgrounds and starting levels on the topic of SGs had to be considered. This gap was covered by maintaining joint interest and goals in learning (see Shaikh et al., 2017). The feedback from students revealed that the discussions were considered fruitful, though some activating interventions were proposed. Positive attitude changes toward gaming were reported. Assignments and material received mixed feedback: some of the students felt they were clear and beneficial, while some needed more structured content and help with assignments. A teacher’s presence, such as for support, communication, and feedback, was expected, even though the courses were noted to be studied independently. These findings support the suggestions of Yunus and Salim (2008). As a summary, the study revealed nine attention points: 1. Carpe Diem method (Salmon et al., 2014) supports multidisciplinary work in planning an online course. 2. The schedule in online course must be set but not too tight. An asynchronous model for online courses gives students more freedom (see also Duraij & Umar, 2015). 3. Assignments and timetables have to be as clear as possible. 4. Peer discussions are an important part of e-learning, and they support social constructivism theory (Durairaj & Umar, 2015; Shaikh et al., 2017; Marsh & Ketterer, 2005). Enabling chat supports discussions. 5. Peer evaluation does not replace the feedback and presence of a teacher. Using too much peer work can be seen as forced. 6. Students who are more advanced in the topic can utilise their expertise in discussions and propose new ideas and materials to their peers. 7. Gamification integrated into course structure, such as levels and a progress bar, can be considered motivating, if they work well. Gamification without connection to course structure (e.g. collecting coins) can be seen as irritating. 8. Even in an independent online course, a teacher is needed: to help with technical issues, to clarify assignments if needed, and to activate discussions (see Tait, 2014). 9. The iterative development and incremental improvement of an online course according to feedback is a good approach (see Fig. 6). Figure 6: Phases in action research study on developing an e-learning course on SGs. The current work continues within the multidisciplinary education on SGs. One expansion will be courses targeted to specific professional areas, such as nursing and healthcare. In these holistic courses on SGs, the developed online course will work as a background material, and it will be completed with hands-on SG prototyping workshops. Acknowledgments The research for this paper was financially supported by the European Social Fund (ESF) and Niilo Helander Foundation. The developed courses can be found from www.segabu.fi References Backlund P., Engström H., Berg Marklund B. & M. Toftedahl. (2017). Developing games for non-leisure contexts: Identification of challenges and research gaps. In 2017 9th International Conference on Virtual Worlds and Games for Serious Applications (VS-Games), Athens, 2017, pp. 15–22. doi: 10.1109/VS-GAMES.2017.8055806 Baranowski T., Buday R., Thompson D., Lyons E.J., Shirong Lu A. & Baranowski J. (2013). Developing Games for Health Behavior Change: Getting Started. Games for Health Journal: Research, Development, and Clinical Applications 2(4), 183–190. Baskerville R. (1999). Investigating Information Systems with Action Research. Communications of the Association for Information Systems: Vol. 2, Article 19. Available at: http://aisel.aisnet.org/cais/vol2/iss1/19 Baskerville R. & Myers M. (2004). Special issue on action research in information systems: Making research relevant to practice. Foreword. MIS Quarterly 28(3), 329–335. Coghlan D. & Brannic T. (2014). Doing action research in your own organization. 4th edition, Sage Publishing. Deterding S., Dixon D., Khaled R., & Nacke L. (2011). From game design elements to gamefulness: defining "gamification". In Proceedings of the 15th International Academic MindTrek Conference: Envisioning Future Media Environments (MindTrek '11). ACM, New York, NY, USA, 9–15. doi: https://doi.org/10.1145/2181037.2181040 De Troyer O. (2017). Towards effective serious games. In 9th International Conference on Virtual Worlds and Games for Serious Applications (VS-Games). Athens, pp. 284–289. doi: 10.1109/VS-GAMES.2017.8056615 Djaouti D., Alvarez J., Jessel J.-P. & Rampnoux O. (2011). Origins of Serious Games. Serious Games and Edutainment Applications. Springer-Verlag London Limited. Durairaj K. and Umar I.N. (2015). A proposed conceptual framework in measuring social interaction and knowledge construction level in asynchronous forum among university students. Procedia – Social and Behavioral Sciences, Volume 176, pp. 451–457. https://doi.org/10.1016/j.sbspro.2015.01.496 Gomez E.A., Wu D., & K. Passerini K. (2010). Computer-supported team-based learning: The impact of motivation, enjoyment and team contributions on learning outcomes. Computers & Education 55(1), 379–390. https://doi.org/10.1016/j.compedu.2010.02.003 Kali Y., Levin-Peled R. & Dori Y.J. (2009). The role of design-principles in designing courses that promote collaborative learning in higher-education. Computers in Human Behavior 25(5), 1067–1078. https://doi.org/10.1016/j.chb.2009.01.006 Kemppainen J., T. Korhonen T. & T. Ravelin T. (2014). Developing Health Games Requires Multidisciplinary Expertise. Finnish Journal of eHealth and eWelfare 6(4), 200–205. Available at: https://journal.fi/finjehew/article/view/48213 Korhonen T. & Halonen R. (2017a). On the development of serious games in the health sector - A case study of a serious game tool to improve life management skills in the young. In Proceedings of the 19th International Conference on Enterprise Information Systems (ICEIS 2017), Volume 3, pp. 135–142. doi: 10.5220/0006331001350142 Korhonen T. & Halonen R. (2017b). Serious games in healthcare: A systematic mapping study. In Proceedings of the 30th Bled eConference, Bled, Slovenia, pp. 349–368. https://doi.org/10.18690/978-961-286-043-1.25 Korhonen T., Halonen R., Ravelin T., J. Kemppainen J. & Koskela K. (2017). A multidisciplinary approach to serious games development in the health sector. In Proceedings of the 11th Mediterranean Conference on Information Systems, Genoa, Italy. MCIS 2017 Proceedings. 30. Available at: http://aisel.aisnet.org/mcis2017/30 Lewin K. (1946). Action research and minority problems. Journal of Social Issues 2(4), 34–46. 10.1111/j.1540-4560.1946.tb02295.x Loh C.S., Sheng Y., & Ifenthaler D. (2015). Serious games analytics: Theoretical framework. In Serious games analytics: Methodologies for Performance Measurement, Assessment, and Improvement, 3–29. New York, NY: Springer International Publishing. 10.1007/978-3-319-05834-4_1 Marsh G.E. & Ketterer J.J. (2005). Situating the Zone of Proximal Development. Online Journal of Distance Learning Administration, VIII(II). Nancarrow S.A., Booth A., Ariss S., Smith T., Enderby P., & Roots A. (2013). Ten principles of good interdisciplinary team work. Human Resources for Health, 11:19. http://doi.org/10.1186/1478-4491-11-19 Onah D.F.O., Sinclair J. & Boyatt R. (2014). Dropout rates of massive open online courses : behavioural patterns. In 6th International Conference on Education and New Learning Technologies, Barcelona, Spain. Published in: EDULEARN14 Proceedings pp. 5825-5834. Ricciardi F. & De Paolis L.T.D. (2014). A comprehensive review of serious games in health professions. International Journal of Computer Games Technology Vol. 2014, Article ID 787968, 11 pages. http://dx.doi.org/10.1155/2014/787968 Ryan R.M., Rigby S. & Przybylski A. (2006). The Motivational Pull of Video Games: A Self-Determination Theory Approach. Motivation and Emotion 30(4), 344–360. http://dx.doi.org/10.1007/s11031-006-9051-8 Salmon G., Gregory J., Dona K.L. & Ross B. (2014). Experiental online development for education: The example of the Carpe Diem MOOC. British Journal of Educational Technology 46(3), 542–556. http://dx.doi.org/10.1111/bjet.12256 Salmon G. & Wright P. (2014). Transforming Future Teaching through Carpe Diem, Learning Design. Education Sciences 4(1), 52–63. doi:10.3390/educsci4010052. Shaikh U.U., Umari U., Karim S. & Asif Z. (2017). Re-Thinking Vygotsky: Applying Social Constructivism to Asynchronous Online Courses utilizing the Power of Crowdsourcing. In Proceedings of the 21st Pacific Asia Conference on Information Systems, Langkawi, Malaysia. PACIS 2017 Proceedings, 233. Available at: http://aisel.aisnet.org/pacis2017/233 Susi T., Johannesson M. & Backlund P. (2007). Serious Games: An Overview. Technical Report HS-IKI-TR-07-001. Available at: http://his.diva-portal.org/smash/get/diva2:2416/FULLTEXT01.pdf Tait A. (2014). From place to virtual space: Reconfiguring student support for distance and e-learning in the digital age. Open Praxis 6(1), 5–16. http://dx.doi.org/10.5944/openpraxis.6.1.102 Yunus Y. & Salim J. (2008). Framework for the evaluation of e-learning in Malaysian Public Sector from the pedagogical perspective. In Proceedings of the International Symposium on Information Technology (ITSim), Vol. 3, 8 pages. IEEE. Zyda M. (2005). From visual simulation to virtual reality to games. IEEE Computer 38(9), 25–32. https://doi.org/10.1109/MC.2005.297