Value Creation in the Internet of Things Ecosystem A study of how to control and leverage the value generated by connected devices Master’s Thesis in the Master’s Programme Entrepreneurship and Business Design MARCUS ANDERSSON JOHAN WIKLUND Department of Technology Management and Economics Division of Entrepreneurship and Strategy CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2018 Report No. E 2018: 028 MASTER’S THESIS E 2018: 028 Value Creation in the Internet of Things Ecosystem A study of how to control and leverage the value generated by connected devices MARCUS ANDERSSON JOHAN WIKLUND Tutor, Chalmers: Bowman Heiden Tutor, Essity: Maria Mellgren Department of Technology Management and Economics Division of Entrepreneurship and Strategy CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2018 Value Creation in the Internet of Things Ecosystem A study of how to control and leverage the value generated by connected devices. MARCUS ANDERSSON, JOHAN WIKLUND © MARCUS ANDERSSON, JOHAN WIKLUND, 2018. Master’s Thesis E 2018: 028 Department of Technology Management and Economics Division of Entrepreneurship and Strategy Chalmers University of Technology SE-412 96 Gothenburg, Sweden Telephone: + 46 (0)31-772 1000 Chalmers Reproservice Gothenburg, Sweden 2018 Acknowledgments This master thesis was conducted during the spring of year 2018. It is the final deliverable at the Entrepreneurship and Business Design Master’s programme at Chalmers University of Technology, Gothenburg. This thesis is the result of a study conducted by Marcus Andersson and Johan Wiklund in collaboration with Essity Hygiene and Health AB. We would like to start by thanking Essity for giving us the opportunity to conduct this thesis at their Gothenburg office. We would like to express a special thank to Maria Mellgren who has been our tutor and contact person at Essity. Without her valuable insights and continuous support throughout our journey, we would probably have encountered additional struggles. Further, we would also like to thank additional employees of Essity, especially the IP department for their warm welcoming as well as their sharing of knowledge and insights. Within the academia, we would like to thank our supervisor Bowman Heiden, Co-Director at Center of Intellectual Property, for his support, feedback and guidance throughout the process. Our discussions and meetings enabled us to take this study to the next level. In addition, we would like to thank the ICM faculty of Chalmers School of Entrepreneurship for providing us with the tools needed to tackle the challenges faced when conducting the study. On the business side, we would like to thank our interviewees for their insights, expertise and not least, real life connection. Lastly, we would like to show our gratitude to our families, our critical friends Simon Risberg and Carmen Espejel who ensured the quality of this thesis, and to you, our reader. We hope you enjoy this paper and that the reading is a learning experience that you pass onwards. Gothenburg, Sweden, May 23rd, 2018 Marcus Andersson Johan Wiklund Abstract The internet of things (IoT) is one of the top ten innovations that will have the greatest impact on the economy in the coming years. The IoT is predicted to reshape industries and create new competitive dynamics. For industrial firms, the IoT will enable new value offerings and collaborations. In addition, the IoT will reduce industry borders, increase market competition, and value creation will most likely occur in an ecosystem between several actors. With increasing competitiveness, control of value drivers will be essential in order to stay competitive. Intellectual property rights have traditionally been one of the key elements to profit from innovations. Further, patents have been seen as one of the strongest tools to create control positions. However, the patent law was created before the age of digitisation and patented software has been hard to enforce against infringers. This has led to questions of how firms can create control positions within an IoT context. Therefore, the aim of this study is to provide tools for firms to identify, control and leverage values generated by IoT ecosystems. To fulfil the aim of this study, a case study was conducted. Two cases were chosen, the Tork EasyCube, a smart management system for restrooms, and the Bosch IoT Cloud, Bosch’s overall service for IoT devices. This study identified that the value generated in an IoT ecosystem can be divided into five categories; cost structure, revenue model, economic benefits, control, and competence. Each of these values can be used to attract participants to an IoT ecosystem. In order to leverage the values on the market, a firm has to assess the values from a firm-, customer- and collaboration-perspective. This study showed that to obtain control within an IoT context, firms have to use different IPRs as building blocks in combination with different elements of market power and technical control. A firm can leverage IoT generated values by either controlling the entire value chain or by using a more open business model. By controlling the entire value chain, a firm has high level of both control and potential of monetization. On the other hand, by providing its IoT solution as a more open business model, e.g. a platform, a firm can gain fast scaling and gain lock-in effects of customers. This as platforms enable value creating activities which are based on interactions of actors from different industries. Such offerings are hard for competitors to match. However, increased openness brings challenges. This study points out that a firm has to manage two main challenges. These are ownership of data and the risk of confidential information leakage. Keywords: internet of things, intellectual property, intellectual assets, value driver, control, IoT ecosystem, business model, case study, collaboration, leverage, competitive advantage. Abbreviations API Application Programming Interface AWS Amazon Web Services B2B Business to Business DCU Data Communication Unit DRM Digital Rights Management FRAND Fair Reasonable And Non-Discriminatory IaaS Infrastructure as a Service ICT Information and Communications Technology IoT Internet of Things IPRs Intellectual Property Rights M&A Merge and Acquisition OS Operating System OSGI Open Service Gateway Initiative PaaS Platform as a Service PTC Parametric Technology Corporation SaaS Software as a Service SCU Sensor Communication Unit List of Figures Figure 1. Number of connected devices worldwide. ............................................................................... 1 Figure 2. Illustration of the research process. .......................................................................................... 9 Figure 3. Two opportunities that connected devices and the IoT provide ............................................. 12 Figure 4. Three service offerings enabled by the IoT. ........................................................................... 13 Figure 5. Three phases of the IoT and their value drivers. .................................................................... 15 Figure 6. The IoT value model .............................................................................................................. 21 Figure 7. Illustration of a technology tree .............................................................................................. 22 Figure 8. Intellectual assets categories. .................................................................................................. 23 Figure 9. Porter's five forces. ................................................................................................................. 27 Figure 10. Illustration of the Tork EasyCube. ....................................................................................... 31 Figure 11. Technology breakdown of the Tork EasyCube. ................................................................... 32 Figure 12. The Tork EasyCube’s IoT value model. ............................................................................... 33 Figure 13. Illustration of the Bosch IoT Cloud ...................................................................................... 35 Figure 14. Technology breakdown of the Bosch IoT Cloud. ................................................................ 37 Figure 15. The Bosch IoT Cloud’s value model. ................................................................................... 38 Figure 16. Intellectual assets of the Tork EasyCube and the Bosch IoT Cloud. ................................... 41 Table of Content 1. Introduction 1 1.1 Background 1 1.2 Literature review 2 1.3 Prior research 2 1.3.1 Changing value creation 2 1.3.2 Changing business models 3 1.3.3 Changing control mechanisms 3 1.4 Problem definition 3 1.5 Purpose 4 1.6 Research questions 4 1.7 Scope and delimitations 4 2. Methodology 5 2.1 Research strategy 5 2.1.1 Theory and research 5 2.1.3 Qualitative strategy 6 2.2.2 Triangulation 6 2.2 Research design 6 2.2.1 Case studies 6 2.3 Data collection 7 2.3.1 Documents 7 2.3.2 Case studies 8 2.3.3 Interviews 8 2.4 Research process 9 2.5 Quality of research 10 2.5.1 Reliability 10 2.5.1.1 Internal reliability 10 2.5.1.2 External reliability 10 2.5.2 Validity 10 2.5.2.1 Internal validity 11 2.5.2.2 External validity 11 2.5.3 Objectivity 11 3. Theoretical foundation 12 3.1 The Internet of Things 12 3.1.1 Cloud computing 13 3.1.2 Data-driven value creation 14 3.1.3 Data-driven value chains 14 3.2 Business models 16 2.2.1 Product as a service 16 3.2.2 Platform 17 3.2.3 Value propositions 18 3.3 The IoT value model 19 3.4 Technology canvas 21 3.4.1 Technology tree 21 3.5 Controlling IoT technologies 22 3.5.1 Intellectual assets 22 3.5.2 Rights-based control 23 3.5.2.1 Patents 23 3.5.2.2 Copyright 23 3.5.2.3 Trademarks 24 3.5.2.4 Design rights 24 3.5.3 Trade secrets 24 3.5.4 Contractual control 25 3.5.5 Technical control 26 3.5.6 Market-based control 26 3.5.6.1 Porter’s five forces 26 3.5.6.2 Platform control 28 3.6 Theoretical framework 30 4. Case studies 31 4.1 The Tork® EasyCube® 31 4.1.1 Background information 31 4.1.2 Value proposition 31 4.1.3 Technology tree 32 4.1.4 IoT value model 33 4.2 The Bosch IoT Cloud 34 4.2.1 Background information 34 4.2.2 Value proposition 35 4.2.3 Technology tree 36 4.2.4 IoT value model 38 4.3 Intellectual assets and IPRs of the Tork EasyCube and the Bosch IoT Cloud 40 4.3.1 Intellectual assets 40 4.3.2 Bosch’s patent, trademark and design rights activities 41 4.3.3 Essity’s patent, trademark and design rights activities 41 5. Analysis 43 5.1 Individual case analysis 43 5.1.1 The Tork EasyCube 43 5.1.1.1 Value creation 43 5.1.1.2 Market control 44 5.1.1.3 Contractual control 45 5.1.1.4 Technical control 45 5.1.2 The Bosch IoT Cloud 46 5.1.2.1 Value creation 46 5.1.2.2 Market control 47 5.1.2.3 Contractual control 50 5.1.2.4 Technical control 50 5.1.3 Rights-based control 51 5.1.3.1 Right-based control from a component perspective 52 5.1.3.2 Rights-based control from a system perspective 53 5.2 Results 55 5.2.1 Value creation in an IoT ecosystem 55 5.2.2 Leveraging IoT value through control mechanisms 56 6. Conclusions 60 7. Discussion 62 7.1 Relevance of theory 62 7.2 Practical implications 63 7.3 Limitations of the results 64 7.4 Future research 64 8. References 65 Appendix 72 1 1. Introduction The introduction chapter starts by introducing the internet of things as a phenomenon and how it creates new business opportunities and challenges for firms. Thereafter the problem definition, the purpose, the research questions, and the scope and delimitations of this study are presented. 1.1 Background The internet of things (IoT) is ranked as one of the top ten disruptive innovations that will have the greatest impact on the economy in the coming years (Bisson et al., 2013). Although the IoT term did first appear in year 1999, it is not until recently it has gained increased attention (Wielki, 2017). The IoT refers to the use of sensors and communication technologies, so called automatic identification and data capturing technologies. These technologies are integrated into physical objects. Thereby, physical objects gain intelligent characteristics which enable monitoring of the objects (Aharon et al., 2015, Papert & Pflaum, 2017). The concept of the IoT provides huge opportunities for modern organisations to strengthen their competitiveness. New business models based on connected solutions have been shown to destroy traditional industries where physical products are in focus (Wielki, 2017). Therefore, for industrial firms, which sell non-digitised products, the IoT will mean changes to their existing business models, operations and value propositions. As products become intelligent, new ways of creating value for customers evolve and relationships between industries and customers turns around. For industrial firms to survive, it will not be enough to offer physical products only. Instead, firms have to create relationships with their customers and capture their wishes and needs at the planning stage (Sendler, 2018). That will be possible through digitalisation and with the emergence of the IoT. Sendler (2018) as well as Petrusson (2005) state that the future belongs to a sharing economy and the need of capital will gradually recede. Entry barriers to industries will be reduced and it will become easier for new players to enter and impact existing markets. Figure 1. Illustrates number of connected devices predicted to be installed worldwide from 2015 to 2025 (in billions) (Statista, 2018). 2 To compensate for increased competition, firms will have to change their business models (Sendler, 2018). Instead of having a closed product-centred business model, industrial firms have to develop a more open-oriented business model. A model that favours collaborations and becomes customer- centric. An open business model enables easier connections between devices. Hence, new type of customer offerings can be developed. Such offerings are hard for competitors to copy. However, to realize opportunities and to successfully transform a business model is not easy. The level of success depends on employees’ acceptance and willingness (Buschmeyer, Schuh, & Wentzel, 2016). In addition, an open business model will most likely mean that a firm has to collaborate and be part of an ecosystem (Banerjee et al., 2014). The increased digitalisation of economies brings challenges to identify and control generated values. Patents have traditionally been seen as one of the strongest tools to create a position of control. This as patents allow patent holders to exclude others from undertaking patented activities. However, the patent law was created before the age of digitalisation. In addition, patented software has been hard to enforce against infringers. Therefore, challenges of using patents for controlling software have emerged and increased during recent years (Spinello, 2007). To create a control position in a digitised economy, it is critical that a firm has knowledge of which assets that the firm possesses (Petrusson & Pamp, 2009). In addition, a firm also has to have the capabilities of turning its assets into tradable objects. 1.2 Literature review In the beginning of this study an in-depth literature review was conducted. The purpose of the literature review was to explore what previously had been written within the area of the IoT and how firms can leverage IoT generated values. That meant that the authors were able to identify which concepts that were developed and what had been written before in the field of research (Bryman & Bell, 2015). The literature review brought the authors an understanding of the research topic as well as it influenced the scope of this study. This as the literature review identified previous research and what theory that was available in order to fulfil the aim of this study. 1.3 Prior research The literature review enabled the authors to conclude that there are three main areas of the IoT where prior research has been conducted. These are changing business models, changing value creation and changing control mechanisms. 1.3.1 Changing value creation The phenomena of collecting and analysing data provides the potential to offer business value that goes beyond operational cost savings (Aharon et al., 2015; Banerjee et al., 2014). The IoT creates opportunities for more dynamic industries and new ecosystems. The term “data is the new oil” has widely been used during recent years. The term refers to data as the new way to guarantee growth and 3 profit (Feldmann, Hartmann, Zaki, & Neely, 2016; Rotella, 2012). Data from connected devices constitutes valuable source of direct and accurate information that can be further analysed to decide how to act or react (European Patent Office, 2017). To take advantage of collected data, managers and executives are required to truly embrace data-driven decision making (Aharon et al., 2015). In addition, managers have to understand that the value creation in an IoT environment, will most likely occur in an ecosystem between players located at different operational layers (Banerjee et al., 2014). The emerging changes in value creation makes this subject relevant to further research. 1.3.2 Changing business models Wielki (2017) explains how the IoT enables firms to offer their products as services instead of as physical products only. Turber, Brocke, Grassman, and Fleisch (2014) add to this and state that the use of digital technology in physical products change customers’ perceptions of values. Due to these changes, industrial firms’ business models will have to change. Instead of the traditional firm-centric approach to how value is created, the value creation process in an IoT context is more like an ecosystem (Turber et al., 2014). With the IoT and the changing dynamics of markets, value chains will be disrupted. These factors force existing firms to rethink their methods and business strategies. Industry borders fade and new types of collaborations across industries emerge. The issue of understanding how the IoT’s development affect business models makes the subject highly relevant to be further analysed. 1.3.3 Changing control mechanisms Intellectual property rights, namely patents have traditionally been seen as one of the strongest tools of firms in order to use to create a control position. This as patents allow patents holders to exclude others from undertaking patented activities. However, patented software has been hard to enforce against infringers (Robinson, 2015). Courts often argue that software achieves abstract ideas only. Furthermore, the IoT technology is interactive and collaborative. Therefore, patent infringements are often undertaken by multiple infringers (joint infringements). That makes the legal processes difficult. In addition, when developing software, agile methods are used. Agile ways of working affect the possibility to protect methods through patents (Millien & George, 2016). Altogether, the issues of software patentability, new methods of working brought by the IoT, and increasing challenges for firms to protect and control digitised products, makes the control focus of the subject highly relevant for this study to further examine. 1.4 Problem definition The emergence of the IoT reshapes business dynamics and competition of industrial firms. To compensate for increased competition, industrial firms will have to change their business models (Sendler, 2018). Instead of using product-centred and closed business models, industrial firms have to transform their business models to become more open. This transformation will impact where and how value is created. In addition, increasing connectivity brings concerns of how a firm can protect an IoT 4 solution. Despite prevailing studies conducted on the IoT and connected devices, there is a notable need for an in-depth analysis on how to control and leverage IoT generated values. The literature review did not identify one single theory that could provide a theoretical framework which allowed for an in-depth and holistic analysis of how firms can control and leverage IoT generated values. However, the literature review identified theory that could be combined in order to provide such an analysis. 1.5 Purpose Based on the problem definition, the potential of combining different theory in order to provide an in- depth and holistic approach to how IoT generated values can be controlled and leveraged calls for further research. Therefore, the purpose of this study is to identify theory than can be combined to provide tools in order to identify, control and leverage values generated in IoT ecosystems. 1.6 Research questions To fulfil the purpose of this study, the following main research question has been formulated: MRQ: How can the value generated in an IoT ecosystem be controlled and leveraged? To answer the main research question, the question has been further broken down into two additional research questions. The two sub-questions are as follows: RQ1: What values are generated in an IoT ecosystem? The objective of the first sub-question has been to identify what type of values that are generated in an IoT ecosystem. RQ2: How can a firm leverage the value on the market through different control mechanisms? The objective of the second sub-question is to examine how firms can use control mechanisms in order to leverage IoT generated values on the market. 1.7 Scope and delimitations The aim of this study is to identify theory that can be tested in order to identify values in an IoT ecosystem and how they can be leveraged on the market through control mechanisms. In order to do this, the developed theoretical framework was tested on two cases. This study does not include considerations in relation to personal data storage, regulations or financial matters. 5 2. Methodology This chapter outlines and explains how this study has been conducted, research methods used and how validity and reliability concerns have been managed. The chapter includes the following sub- chapters; research strategy, research design, data collection, research process, and quality of research. 2.1 Research strategy The following sections will highlight this study’s research strategy. A research strategy helps to explain the link between theory and research (Bryman & Bell, 2015). The link between theory and research in this study is highlighted in the following sub-sections of theory and research, qualitative strategy, and triangulation. 2.1.1 Theory and research The purpose of this study includes both identification and testing of theory. This makes the relationship between theory and research both inductive and deductive (Bryman & Bell, 2015). Inductive as different theory is identified in order to create new theory, and deductive as produced theory is tested. This relationship between theory and research in this study can be explained as inductive theory-building (Eisenhardt & Graebner, 2007). The theory used in this study can be explained as middle-range theory. According to Bryman & Bell (2015), middle-range theory is domain specific and vary depending on research purpose. For this study, therefore, theory has been selected according to achieve the research purpose. The theory used in this study is further explained in the section of theoretical foundation. To find relevant theory to this study, the field of research was explored by the literature review. This as an explorative approach toward literature is convenient for conducting research within relative unexplored topics (Denscombe, 2010). As mentioned above, no single framework or theory was found in order to fulfil the purpose of this study. However, by identify and combining different theory, the aim of this study was fulfilled. The combination of theory in this study can be explained as the development of a concept (Lynham, 2002). That means that the authors of this study first constructed a theoretical framework. The framework was then operationalized (Lynham, 2002). For this study that meant that the different parts of the theoretical framework were partly tested and the results were discussed with persons considered to be experts within the respective field of testing. The operationalization part brought insights of how the theory had to be changed. After the operationalization phase, the framework was once again tested in practice in order to verify outcomes with experts within the field of research. This phase concluded that the constructed theoretical framework was applicable (Lynham, 2002). Thereafter, the theoretical framework was fully tested in practice in order to confirm or to disconfirm the usefulness of the framework (Lynham, 2002). That was done by applying the framework to a specific case were the 6 results were discussed with persons having specific knowledge of the case. These persons confirmed the theoretical framework and its applicability. During the testing process, new insights were gained which were incorporated into the framework. Finally, the amended theoretical framework was applied to the reference case once again as well as to a second case. In order to test this study´s theoretical framework, the epistemological orientation has been positivistic (Bryman & Bell, 2015). This as the theory has been tested in order to confirm or disconfirm theory. The ontological consideration in relation to this study´s theory is objectivism (Bryman & Bell, 2015). This as the study´s theoretical framework approaches the the subject matter as if it was an objective entity. 2.1.3 Qualitative strategy In order to test this study´s theoretical framework, this study´s empirical data has been collected by using a qualitative research strategy (Bryman & Bell, 2015). The aim of the collected data has been to develop the theoretical framework and not to generalize the results. Further, this study is based on qualitative data as the study calls for a high level of detail for testing chosen theory. According to Denscombe (2010), that allows the research to reveal more information with higher level of detail (Denscombe, 2010). 2.2.2 Triangulation To allow for a more complete understanding of the studied subject matter as well as to increase the quality of findings (Denscombe, 2010), this study has used different data collecting methods. That allowed for methodological triangulation (Denscombe, 2010). The purpose of using triangulation in this study is to bring forward different perspectives on the researched subject matter. Different perspectives have been achieved by using remarkably different methods. In this study, the high difference between methods has been achieved by collecting data from case studies, documents, and interviews. 2.2 Research design A research design illustrates the frameworks used in a study in order to collect and analyse data (Bryman & Bell, 2015). This study’s research design is based on case studies. 2.2.1 Case studies A case study is appropriate to use in order to gain insights of how and why something happens (Denscombe, 2010; Eisenhardt & Graebner, 2007). This study is based on two case studies in order to inductively build and test theory (Eisenhardt & Graebner, 2007). The first case served as the testing object for operationalizing, testing, applying and confirming the constructed theoretical framework. By focusing on one case as the starting point, a researcher has the potential to build better theory compared to if several cases would have been included. This as the author has the possibility to match the theory to the specific case. The second case was used in order to extend and test the verified 7 theoretical framework. The cases were seen as independent units for analysis as the aim of this study´s inductive theory-building is to create a theoretical framework that can be replicated. By using two cases in this study, it was possible to take advantage of qualitative insights from the first case and transform them into new theory that was tested on the second case (Eisenhardt & Graebner, 2007). Bryman &Bell (2015) add to that as a case study with several cases improves theory building and puts the research in a better position to identify circumstances why theory does or does not hold. Further, a case study research design contributes to create a better understanding of the studied social phenomena (Bryman & Bell, 2015). As the theory used in this study was identified and combined into a new theoretical framework, the theory built in this study is extended from previous research. The purpose of extending the theory in this study is to provide an in-depth and holistic approach towards the research subject matter. From a theory perspective, that enables refinement and improvement of existing theory (Eisenhardt & Graebner, 2007). The two cases were selected based on theoretical sampling. That means that the cases were selected as they were found to be suitable for theory testing and theory building. In addition, the cases were selected because they did represent different examples which can be seen as extremes when compared to each other. That facilitates pattern recognition between the two cases and contributes to increase the replicability of this study´s produced theory (Eisenhardt & Graebner, 2007). 2.3 Data collection The inductive and deductive approach of this study in combination with a case study research design, called for three different methods for collecting data. These are data collected from documents, case studies, and interviews. 2.3.1 Documents Public documents such as company information, reports, conferences and press releases have been used in order to collect background information of the two cases. Further, public information has been used in order to evaluate the two cases patenting-, trademark- and design-rights activities. In addition, public documents have also been reviewed during the literature review. The majority of the literature review was conducted during the first phase of this study. To ensure that the most relevant theory was chosen and to increase the quality and level of objectivity, the literature has been revised and improved when applicable. The majority of the revisions were done before the theoretical framework of this study was established. To ensure that the highest quality of information was retrieved, the main sources of information have been collected from databases provided by Chalmers University of Technology. In more detail, the main databases used were, the Emerald, ScienceDirect, Scopus and ProQuest. Due to the novelty of the research topic, the vast amount of article publications has taken place during the last five years. In order to assess which articles that should be used in this study, two 8 main parameters have been used. First the (1) number of citations and (2) how well-known the author(s) is within the field of research. In addition to academic sources of information, non-academic sources have been used in this study as well. Examples of such sources are white papers developed by management consultancy firms, such as McKinsey and Deloitte and articles published by the European Patent Office (EPO). Both the consultancy firms and the EPO are known of striving for high quality of content. Therefore, it can be argued that their articles and reports are of the high quality needed to be used as sources in this thesis. 2.3.2 Case studies In order to be able to evaluate and test the theoretical framework of this study, two cases were chosen. The first case served as the main object for theory testing while the second case was studied in order to extend the theoretical framework. This to be able to make additional refinements to the framework. The two cases were selected based on theoretical sampling (Eisenhardt & Graebner, 2007). Denscombe (2010) adds to this by explaining that it is essential that cases are selected based on their specific characteristics. To select appropriate cases for this study, the cases had to have the following characteristics. First, the cases had to be practically linked to the purpose. For this study, that means that the cases had to represent industrial-firms which provide IoT offerings. Secondly, the cases were to be considered as typical units of analysis (Denscombe, 2010). Finally, to allow for a broader discussion of how IoT generated values can be leveraged, the cases were also selected to represent different business models enabled by the IoT. For this study that means an open- and a closed-business model. The selected cases are as follows: − The Tork® EasyCube® - an IoT based monitoring/management system for restrooms. By connecting dispensers and logging usage, the overall management of restrooms becomes more intelligent. The product is sold as a product as a service where customers pay-as-they-use plus a monthly subscription fee. − The Bosch IoT Cloud - is Bosch’s holistic cloud-based solution for IoT devices. The Bosch IoT Cloud comprises of three main elements; Software as a Service (SaaS), Platform as a Service (PaaS) and Infrastructure as a Service (IaaS). 2.3.3 Interviews Qualitative interviews have been used to collect data with high level of detail (Bryman & Bell, 2015). For this study, the interviews provided inputs of high level of detail for discussing and confirming this study's results from applying this study's theoretical framework. To enable comparisons of interviews, an interview guide was developed. The interview guide contributed to a flexible interviewing process as well as for retrieving answers to specific questions. This makes the interviews used in this study semi-structured. A semi-structured interview has defined questions but the interviewer can add additional questions during the interview. In addition, respondents are not deemed to answer the 9 questions in a specific manner and do have flexibility in how they deliver their answers (Bryman & Bell, 2015). Notes were taken during the interviews in order to capture the interviewees’ personal perspectives and to keep track of all information exchanged during the interviews. Summaries of the interviews were written within a day after the interviews were completed. That ensured a thorough analysis of the interviews and made it easier to compare the interviews. The interviewees were selected based on the following criteria (1) their experience within the field of research, (2) job assignments in relation to the topic of this study and (3) the interviewees’ availability for interviews. 2.4 Research process This section outlines the overall research process and explains the methods used in each phase of the study. The study can be divided into three main phases. The first covers the background study where the researchers gained a deeper understanding of the issues and challenges of the topic as well as identified relevant theory. The background study, together with the theory building part constitute the foundation of this study's developed theoretical framework. The theory building represents the transition process from phase one into phase two. Phase two continued with theory building and also represents the process of conducted research. The conducted research, first included testing of the theoretical framework by applying it to the Tork EasyCube. By comparing the results from testing the theoretical framework with data gathered from interviews, the theoretical framework was further developed. The theoretical framework was then applied to the Tork EasyCube once again. Thereafter the theoretical framework was applied to the Bosch IoT Cloud. The third and last phase of this study covers the analysis, result, conclusion and discussion as well as finalising the thesis. The research process can be seen in figure 2. Figure 2. Illustration of the research process. 10 2.5 Quality of research People perceive and interpret things differently, it is therefore important to question a study’s gathered material (Patel & Davidsson, 2003). The main purpose of questioning the source of data is to challenge how data has been collected and compiled. If the source of data is of poor quality or is incorrect, the result will be of low quality or false. This phenomenon is referred to as “garbage in, garbage out” (Edvardsson, 2009). This study is a qualitative study and in a qualitative study the concerns relate to words rather than numbers (Bryman & Bell, 2015). It is extremely important to question the sources of data and to use reliable sources only to achieve high quality. To achieve high quality of data, three parameters have been in focus when conducting this study. These parameters are reliability, validity and trustworthiness. 2.5.1 Reliability The consistency of measuring a concept is referred to as reliability (Bryman & Bell, 2015). Bryman & Bell categorise reliability into two main categories, internal and external. 2.5.1.1 Internal reliability Internal reliability refers to how the researcher interprets the collected data (Bryman & Bell, 2015). In more detail, if the result would be the same if the used data was processed again. To reach high internal reliability of this study, both authors as well as supervisors have controlled the data used in this study. In addition, when the authors have gained additional knowledge of the studied subject matter, information and data used have been revised. Finally, to increase the internal reliability, supervisors and opponents have continuously taken part of discussions of how the study proceeded. 2.5.1.2 External reliability External reliability refers to the degree a study can be replicated (Bryman & Bell, 2015). In a qualitative study, replication is often a challenge as settings change over time. Bryman & Bell (2015) refer to this phenomenon as stability. Most of the data used in this study has been collected by using established frameworks. To mitigate the risk of collecting case data of low-quality, only primary sources have been used. In addition, all results from interviews, cases studies and literature have been evaluated and triangulated. All the frameworks used in this study have been accompanied by thorough user instructions. There is, however, a risk that interpretations of guidelines differ among people. To compensate for that risk, both authors have studied the guidelines and the supervisor from academia has been consulted. This to discuss the authors’ intentions of using the frameworks as well as their applicability to this study. 2.5.2 Validity One of the most important criterion of a study is its validity (Bryman & Bell, 2015). Validity refers to the integrity of the author’s conclusions based on the results of a study. For qualitative studies, validity can be categorised into two categories, internal and external. 11 2.5.2.1 Internal validity Internal validity relates to the issue of causality. The issue of causality means whether or not the conclusions and findings of the researcher actually hold water (Bryman & Bell, 2015). For instance, if parameter x causes y, is it certain that x causes variation in y and not something else. To achieve high internal validity, this study has used highly cited and well-known frameworks with clear guidelines. In addition, external experts have been consulted through interviews to validate the conclusions made. The authors are aware of that the framework used for identifying value creators in an IoT ecosystem is relatively new. That could imply validity issues. To mitigate that risk, the background of the framework, its instructions, as well as examples of when the framework has been used have been thoroughly studied by the authors. 2.5.2.2 External validity External validity addresses the concerns of the results applicability to be generalised beyond the specific research context (Bryman & Bell, 2015). For this study, one concern is therefore how this study’s cases have been selected. As two cases only are included in the study, it can be difficult to draw conclusions which are applicable beyond the specific field of research. To mitigate the risk of low external validity, this study’s cases have been strategically selected. The cases have been chosen because they have a clear connection to the purpose of the study. Further, they are considered as typical and appropriate examples for deductively testing this study´s selected frameworks and theory. To reduce the risk of low potential of generalising this study´s conclusions, the conclusions have been analysed and compared with literature, data collected from interviews and the supervisor. In addition, the results have also been discussed with experts within the field of study. 2.5.3 Objectivity Objectivity implies whether or not the researcher’s personal values have interfered when results were interpreted (Bryman & Bell, 2015). Absolute objectivity is impossible to guarantee in business research. However, to mitigate the risk of intrusion when testing the theory, the authors were guided by established frameworks throughout the entire research process. The results generated by using the frameworks have been anchored with professionals and the authors’ supervisors. By thoroughly explaining the research process, the used theory and concepts, the authors’ increase the replicability of this study. Such approach is according to Bryman & Bell (2015) convenient to increase a study’s objectivity. 12 3. Theoretical foundation This chapter presents the theoretical foundation of this study. The chapter starts with an introduction to the IoT and how it affects business environments (3.1). Thereafter, IoT based business models and the concept of value proposition are introduced (3.2). Section 3.3 introduces the IoT value model while section 3.4 explains how value layers of an IoT ecosystem can be disaggregated. Section 3.5 explains how IoT technologies can be controlled. The chapter ends by section 3.6 that describes the overall design of this study’s theoretical framework. 3.1 The Internet of Things The Internet of Things (IoT) refers to the use of sensors and communication technologies, so called automatic identification and data capturing technologies (Wielki, 2017). These technologies are integrated into physical objects. That gives physical objects intelligent characteristics and enables easier monitoring of the objects (Aharon et al., 2015; Chaouchi, 2013; Papert & Pflaum, 2017). Integrated intelligent objects connected to each other create complex networks based on three core elements: (1) smart devices, (2) data aggregation and (3) methods and systems for analysing and storing device data (Fan & Zhou, 2011). Collecting data and performing analysis of data enable business value that goes beyond operational cost savings. In addition, device data creates opportunities to establish new ecosystems and more dynamic industries (Aharon et al., 2015; Banerjee et al., 2014). The IoT and digitalisation reduce borders between industries, companies, and technologies. That leads to new business opportunities, see Figure 3. Figure 3. Illustrates two different opportunities which connected devices and IoT provide (Wielki, 2017). Wielki (2017) explains how the IoT and data collection will enable new value propositions. Firms will be able to offer their products as services instead of selling them as physical products only. Heppelmann & Porter (2014) add to this and explain that the IoT and changing business environments will disrupt value chains and forces existing companies to rethink their methods and processes. For firms to realize the opportunities of emerging business models and the potential of the IoT, firms have to collaborate and become parts of IoT ecosystems (Banerjee et al., 2014). An ecosystem is defined as a complex web of interdependent agents and relationships between components such as technology, industries and people (Shin & Park, 2017). By taking part in IoT ecosystems, firms have greater 13 potential to drive increased business value and to increase data sharing capabilities. By sharing data, IoT ecosystems can attract specialised participants which can use data to generate additional customer value. 3.1.1 Cloud computing The IoT will enable new service offerings. Service offerings in the era of digitalisation are closely linked to cloud computing (Hao & Helo, 2018). Cloud computing can be defined as computed resources located and stored on a provider’s server. The computed resources can then be accessed by customers on demand through the internet (Hao & Helo, 2018). Cloud computing typically falls into three groups of service offerings (1) Software as a Service (SaaS), (2) Platform as a Service (PaaS) and (3) Infrastructure as a Service (IaaS), see Figure 4. Each offering includes different degrees of engagement and contribution by the customer and the provider. Figure 4. Three different groups of service offerings enabled by the IoT (Bosch, 2017; Hasanzadeh, Safari & Safari, 2015). Software as a Service (SaaS) provides customers with access to software services via the internet (Hasanzadeh et al., 2015; Patel, Seyfi, & Jaradat, 2011). SaaS is the most widespread and applicable cloud-based service and includes value propositions such as pay-as-you-use or pay-per-month. SaaS often contributes to reduce costs and fixed-capital of customers. The SaaS provider is responsible for all the parts needed in order to deliver the service to the customer. Customers, on the other hand, only interact with the software service and manages data. An example of a SaaS, is a connected device that alerts customers of what and when to do certain tasks. SaaS provides firms with high computational power and enables firms to gain a greater understanding of its operations and customers (Patel et al., 2011). That means that SaaS often has impact on a firm’s decision making. Especially for adaptation decisions where SaaS is considered to deliver its highest value. 14 Platform as a Service (PaaS) is defined as a cloud-based service that allows customers to develop, run and manage applications on a platform (Farouk, Yousif, & Bakri Bashir, 2015). The platform infrastructure is controlled and managed by the provider of the platform. However, customers and developers may have control of applications deployed on the platform. Infrastructure as a Service (IaaS) is a service offering where the infrastructure such as virtual storage, machines, data-sets and servers are provided to customers via the internet (Hasanzadeh et al., 2015). IaaS facilitates the parts needed in order to deliver both SaaS and PaaS. An IaaS provider manages the infrastructure and offers customers support and services. 3.1.2 Data-driven value creation The term “Data is the new oil” has widely been used and spread during recent years and refers to data as the new way to guarantee growth and profit (Feldmann, Hartmann, Zaki, & Neely, 2016; Rotella, 2012). Data is the main feature of connected devices and is a valuable source of direct and accurate information. That information can then be analysed and used to gain recommendations of how to act or react (European Patent Office, 2017). Data generates new values which can be leveraged and used to gain competitive advantages (Feldmann et al., 2016). The retail industry is one example where data is leveraged to create new values. For instance, by analysing behaviours of customers, retailers can adjust prices and target promotions. In order to take advantage of data collection, managers and executives have to embrace data-driven decision making (Aharon et al., 2015). Aharon et al., (2015) mention seven key areas where data can be used to add value to firms: 1. Analytics 2. Cost reduction and efficiency 3. Realization of physical system 4. Software and data driven innovation 5. Risk reduction 6. New ways of distributing and storing data 7. Business transformation According to Brown, Kanagasabai, Pant, & Pinto (2017), firms which use data strategically and analyses behavioural insights, outperform their competitors. Such firms have been able to achieve 85 percent higher sales growth and 25 percent higher gross margin compared to competitors which do not use data. 3.1.3 Data-driven value chains The IoT will impact value chains and firms that do not respond accordingly are at risk (Brock, Dreischmeier, & Souza, 2013). The IoT enables firms within one sector to become more agile and play important roles in other sectors for delivering their products and services. The fading boundaries 15 between industries and more agile firms, change competitive dynamics of markets (Brock et al., 2013). As new players emerge and more data is captured and analysed, a shift in how value is created is expected to occur. This change is expected to distribute larger shares of the value to suppliers of software and analytics. Aharon et al., (2015) further state that suppliers and installers of hardware, IoT devices and IoT systems most likely will capture less value in the future. For a firm to fully understand the IoT’s effects on value chains, it is essential that the firm understands how the technology and the industry will evolve. The implementation process of the IoT, is based on three main phases. Depending on which phase an industry is within, value will be created by different IoT domains. Figure 5 illustrates the three main phases of IoT and their value domains (Aharon, et a., 2015). Figure 5. illustrates three phases a firm potentially could go through and what the value drivers will be in each of the three phases (Aharon et al., 2015). It is essential for a firm to understand how the firm’s strengths and strategies fit into each implementation phase of the IoT (Aharon, et a., 2015). This in order to achieve competitive advantage and to survive on the market. To achieve a control position throughout the three IoT implementation phases, a firm can become a complete-service provider (Aharon et al., 2015). That strategy enables a firm to expand its position on the market and to become more powerful in the industry. Not all firms have the possibility to become a supplier of complete services. Bughin, Catlin, Hirt, & Willmott (2018) have identified five aspects a firm has to manage in order to succeed when implementing the IoT and related services. First, (1) a firm has to create a clear definition of its digital strategy. Data enables new ways of analytics and those should be used by firms to create new competitive advantages. (2) Understand the economics of digital. On average, digitalisation creates more value to customers than to the firm itself. Digital often cut prices and firms have to understand how to compete and to create novel values for customers in order to stay competitive. In addition, digital rewards first movers and superfast followers. Hence new ways of agile development give the first mover a learning advantage. A first mover often has the possibility to cut costs or to release new versions of its services before followers have launched their first version (Bughin et al., 2018). The first mover’s information advantage, makes it possible to understand how the market will develop. (3) To stay competitive, a 16 firm has to understand the complete IoT ecosystem and its participants and how competitors are expected to emerge. (4) Identify new B2B opportunities. This as the IoT will enable closer collaborations between businesses. Finally, (5) managers have to understand the degree and the pace of change digitalisation brings and how it affects the firm’s current market situation. 3.2 Business models The IoT enables new type of business opportunities and new type of business models. A business model explains how an organisation creates and delivers value (Osterwalder & Pigneur, 2010). Therefore, to be able to fulfil the purpose of the study, IoT enabled business models had to be identified. The following sections presents two types of IoT supported business models. These are: product as a service and platform. The section then continuous by introducing the central elements of a business model and value propositions. 2.2.1 Product as a service One way of capturing IoT enabled opportunities, is to turn a traditional physical product into a product as a service (Northstream, 2017). A product as a service can be defined as: “It is a business model, in which the enterprise sells an integrated package that includes hardware, software, connectivity, maintenance, customer support, installation and other value adding services for a recurring fee. Such a business model innovation centered around service orientation is often referred to as servitization.” (Northstream, 2017 pp. 3) The concept of a product as a service, implies that a supplier provides a solution that has a duration that lasts beyond a single transaction. Instead, the customer is charged based on usage or/and performance. For this kind of business model, it is essential to collect and analyse user data to create value (Northstream, 2017). There are five main business aspects which benefit from using a product as a service. These are offerings, sales, profitability and cash flow, and customer relationship. Collecting user data enables offerings which are optimized according to customers’ needs. Such offerings include tailored prices and specialised offerings towards its customers Thereby a firm has greater potential to differentiate its offerings from competitors on the market (Northstream, 2017). In addition, it becomes easier for a firm to gain new customers since they pay less upfront for gaining access to the service. The relatively long duration of service offerings also enables firms to build deeper relationships with customers which may lead to increased loyalty with customers. For a firm’s profitability, on the one hand, service offerings generate predictable revenue streams. On the other hand, by analysing data from the service provided, the firm can optimize its operations and increase efficiency. Thereby a firm can become more cost efficient as well. Transforming a firm to become service-based is tempting when aiming for outstanding value creation. However, the transition process is challenging and will impact the firm’s organization (Jovanovic et al., 2016; Northstream, 2017). A product as a service is well suited for products with long life-cycles. 17 Hence, product as a service is often suitable for B2B environments. Jovanovic et al., (2016) state three main factors which decide whether or not a firm will become successful by offering its products as products as a service. These factors are: 1. Which functions and roles does the product have with customers. Product as a service is appropriate to offer when the service provides supporting activities to customers’ core value creation processes (Jovanovic et al., 2016). In addition, a product as a service is favourable when the running costs of the product is high in comparison to its initial price of purchase. Further, product as a service is also appropriate in environments where breakdowns of the product lead to extensive costs. 2. Environmental conditions where the product is used. Product as a service is more appropriate in environments which are stable and predictable (Jovanovic et al., 2016). This since a stable environment makes it easier to calculate and predict the product’s need of service and maintenance. 3. Characteristics of delivery systems for products and services. Product as a service has higher potential to become successful when a firm has direct access to its customers and to users of the service provided (Jovanovic et al., 2016). Transforming an industrial firm to become service-based is not easy. The transformation process is dependent on the firm’s employees’ willingness and acceptance of transformation (Buschmeyer et al., 2016). If the process is not facilitated accordingly, the firm may be at risk. This as the transformation process may have an overall negative impact on the firm’s performance (Jovanovic et al., 2016). The process of change means changes of employees’ focus. Instead of only focusing on production or sales functions within the firm, employees of a service-based firm have to frequently interact with customers. There are some best practices in order to successfully transform a firm to become a service-based firm (Northstream, 2017; Jovanovic et al., 2016). First, the firm has to have a clear strategy that is strongly supported by the management team. In addition, the transformation project team has to be cross-functional, including experts within both commercialization and technology. Secondly, the transformation process should be initiated by selecting a few numbers of specific cases and customers. That to facilitate thorough testing and a stepwise adaptation process. 3.2.2 Platform Platform as a business model is used to connect people, resources and organisations to create interactive ecosystems (Parker, Van Alstyne, & Choudary 2016). A platform can be defined as: ”A platform is a business based on enabling value-creating interactions between external producers and consumers. The platform provides an open, participative infrastructure for these interactions and sets governance conditions for them. The platform’s overarching purpose: to consummate matches among users and facilitate the exchange of goods, 18 services, or social currency, thereby enabling value creation for all participants.” (Parker et al., 2016. pp. 11) Platforms can be deployed in all industries where information is important (Parker et al., 2016). The value created on platforms are created by multiple actors. Those actors can be producers and consumers of values simultaneously. On a platform, value is produced, co-produced, exchanged and changed at different locations at the same time. One of the strongest advantages of platforms compared to linear value chains, is their ability to foster faster and efficient scaling (Parker et al., 2016). This as platforms do not have to only use a single firm’s equity or resources to establish the entire value creating system. On platforms, the community decides which products that should be offered. This is different from linear value chains where decisions whether or not to add products are taken by gatekeepers. In addition, platforms enable to unlock new sources of value and provide new sources of supply as platforms provide a greater variety of products to customers. Through faster scaling a platform can achieve network-effects. Network- effects means that the more users that connect to the platform the more value is produced to the platform’s community (Parker et al., 2016). With increased value for the platform’s community, the attractiveness of the platform will increase and more users will join. To achieve network-effects, a platform must offer demand economies of scale. Demand economies of scale means that users increase the value of products and services offered to other users. Demand economies of scale are further strengthened by technological advancements on the demand side of a platform. A platform network is two-sided when both consumers and producers are involved (Parker et al., 2016). A two- sided network can lead to two main network-effects, same-side and cross-side effects. Same-side effects mean that consumers have impact on other consumers and producers have impact on other producers. Cross-sided network-effects are achieved when producers have impact on consumers and vice versa. Both same-sided and cross-sided network-effects can be either positive or negative. With positive effects, the platform can grow and with negative effects, the attractiveness of the platform is reduced. A platform provider can be seen as a broker that facilitates matching of customers and producers (Parker et al., 2016). In exchange, the platform provider charges the users of the platform based on their transactions made on the platform. A firm has to invert to become a successful platform provider. Instead of focusing on internal activities only, the firm has to focus more on external activities. Firms which aim to use a platform-based business model, have to become orchestrators of external resources instead of concentrating on optimizing their products only. 3.2.3 Value propositions The value proposition is one of the central elements of a business model (Osterwalder et al., 2014). The value proposition is the value a company delivers to its customer in order to satisfy their needs 19 (Lindic & Marques da Silva, 2011). Value propositions are created to attract customers and for firms’ own internal use. One the one hand, the value proposition has impact on how the organisation works. On the other hand, the value proposition also describes activities needed in order to serve customers and collaborators in a profitable manner. In the era of the IoT, new value propositions are enabled. 3.3 The IoT value model The evolution of the IoT is driven by two underlying trends (Varmesan et al., 2016); (1) The change of focus from seeing the IoT as a technology platform to see it as a business ecosystem and (2) shift from focusing primarily on the firm’s business model to designing ecosystems as well. In such ecosystem, novel ways of creating values are enabled by connecting actors from different industries. This results in a changing business environment that is much different from traditional linear business environments. With this change, traditional business model frameworks for describing value creation in an IoT environment have shown to be insufficient (Turber et al., 2014). To solve that issue, Turber et al. (2014) conducted research aiming to design a framework that can be used to identify values created within an IoT ecosystem. The research identified four key elements which recur in traditional business model frameworks. The first element is the targeted customer (Who), the second is the value proposition (What), the third is the value chain (Where) and the fourth is the incentive factor (Why). These aspects are fundamental for business models and Turber et al. (2014) found that three of those factors are essential for constructing a new framework that is applicable to the IoT. These factors are as follows: 1. Who - are all the participants of an IoT ecosystem. These actors are seen as collaborators from a service-dominant logic. This as they constitute operant resources within an IoT ecosystem and can co-create value with other actors 2. What - identifies in which layer of an IoT ecosystem an actor provides value within. The aspect of what, makes it possible to identify where the value creation process, cooperation and competition take place. The what dimension can reveal that one actor both cooperates and competes simultaneously with another actor at different layers of an IoT ecosystem 3. Why - addresses the IoT generated value and identifies reason(s) for each collaborator to take part of the ecosystem. The why includes value propositions. These are categorised as monetary and non-monetary values To increase the framework’s level of detail, Turber et al., (2014) disaggregated the IoT into four different activity layers. These are: the device layer, network layer, service layer and contents layers. Based on Turber et al.’s (2014) framework, Vermesan, Bahr, Gluhak, Boesenberg, Hoeer, & Osella (2016) further expanded the framework by adding additional layers of the IoT. Vermesan et al’s (2016) IoT framework includes the following components; 20 - Collaboration and Processing Layer - people and business processes, transformation decisions based on applications and knowledge - Application Layer - dynamic applications, reporting, analytics and processed "smart" data - Service Layer – services, multi-cloud services, analytics, mining and machine learning - Abstraction Layer - data abstraction, aggregation and access - Storage Layer - data integration, accumulation and storage - Processing Layer - edge computing, data element analysis and transformation, analytics, data mining and machine learning. Pervasive and autonomic services are provided through machines in both “autonomic” and “smart” ways - Network Communication Layer - connectivity elements, gateways, communication and processing units, wireless technologies and sensor networks, body area networks, local area networks, cellular and 3/4/5G and LPWAN for delivering information - Physical Layer - devices, controllers, sensors and actuators The values generated in each of the IoT layers are offered as value propositions to stakeholders within the IoT ecosystem (Vermesan et al., 2016). To further develop the IoT framework developed by Turber et al. (2014), Vermesan et al., (2016) broke down the dimension of Why even further. That in order to expand the non-monetary and monetary values into the following five categories of value; - Competence - is the value generated that enables an actor to do something more successfully and/or efficient. - Control - refers to an actor’s aim to gain power of influence to control the value generated in order to achieve economic and/or political goals. The value is linked to market power and refers to a firm's ability to influence the price and output of a market. However, the value of control as used in the IoT value model does not as IPRs enable a firm to exclude third parties from undertaking the same activities or to offer the same solutions. - Economic benefits - a value that is generated that can be quantified in terms of money. For instance, the amount of money that can be saved or generated by collaborating within the IoT ecosystem. - Revenue model - is value created from a revenue stream. - Cost structure - method(s) to manage and determine costs. The IoT value model framework, originally developed by Turber et al., and further expanded by Vermesan et al. (2016), can be seen in Figure 6. 21 Figure 6. The IoT value model that visualizes who the participants are, in what layer the value is created and why the different stakeholders benefit from being within the ecosystem (Vermesan et al., 2016). 3.4 Technology canvas The following chapter describes a method to breakdown technology into technology layers and core components. By understanding a technology’s hierarchy and core components, it is possible to identify the technology’s value-creators. 3.4.1 Technology tree To understand what value a solution may bring and how to protect that value, the different components of the solution first have to be identified (Heppelmann & Porter, 2014). By identifying the value creating components of an IoT ecosystem, the components can be approached from a control perspective. By controlling the components, a firm can protect the value generated from competitors and copycats (European Patent Office, 2017). Thereby, a firm will be able unlock and capture the potential of the IoT and drive value creation and innovation. A technology-tree enables visualisation of a product’s technology hierarchies (Clark, 1985). Identifying a technology’s hierarchy serves two purposes. The first, is to group products and technologies. The second, is to distinguish technology groups by identifying characteristics of technology differentiation. A technology tree can be used to understand where a firm has its current assets and capabilities. In addition, the tree can be used to decide what technology a firm strategically should focus on to increase its future market power (Heiden, 2017). Heiss & Jankowsky (2002) add to this and explain that in order to understand if technologies are of a firm’s strategic interest, the technologies have to be evaluated. 22 According to Heiden (2017), “a technology tree is a foundation for analysis and strategies development in relation to each of its’ underlying fields” (Heiden, 2017 pp. 20). A technology tree enables a firm to map out competitors’ assets in relation to the firm’s own assets. The technology tree can also be used as a management tool. Further, a tool for evaluating a firm’s knowledge within a technology field and which technologies the firm possesses. These insights contribute to identify the firm’s opportunities. It is essential to determine the firm’s strategic position because actors who lack strategic positions are more likely to face higher costs of components and licensing fees. These firms are also subjects to higher risk of getting blocked or restricted in their access to technologies (Heiden, 2017). Figure 7 represents an example of a technology tree. The example is a photovoltaic power generation system. Figure 7. Illustration of a technology tree for a photovoltaic system (Bowman, 2017). 3.5 Controlling IoT technologies Control of IoT technologies can be achieved through different mechanisms. By identifying intellectual assets, a firm can structure them in order to leverage them and create a control position on the market (Petrusson, 2004). In this study intellectual assets are group into the following building blocks: rights- based control (patents, copyrights, trademark protection and design rights), trade secrets, contractual control, technical control, and market control. The following sections will first introduce intellectual assets and thereafter the building blocks of this study will be further elaborated in the following sections. 3.5.1 Intellectual assets Intellectual assets are assets referred to as knowledge assets which a firm can influence (Huggins & Weir, 2012). Intellectual assets are created when knowledge, know-how or learning is documented (Sullivan, 1999). Once documented, a firm can apply different control mechanisms, e.g. intellectual property-rights and/or contracts in order to control the intellectual assets. Thereby intellectual assets become resources that can convey competitive advantages (Petrusson, 2016). Figure 8 shows categories of value-creating intellectual assets suggested by Petrusson (2016). 23 Figure 8. Intellectual assets categories (Petrusson, 2016). 3.5.2 Rights-based control Intellectual property rights (IPRs) can be used to claim ownership rights of intellectual assets (Petrusson, 2016). In this study, IPRs are defined as rights-based control. By using IPRs, an actor can claim intellectual assets and transform them into tangible objects. Such objects can then be transformed into capital and thereby create value for a firm. Further, that means that rights-based control can be used to create value for customers (McConnachie, 1997). IPRs allow creators and owners of the rights to exploit and benefit from developed creations (WIPO, 2017). Depending on the characteristics of a creation, different IPRs can be used to establish a control position for that creation These IPRs are patents, copyright, design rights and trademark protection. Further explanations of these IPRs follow in the sections below. 3.5.2.1 Patents Kranakis (2007) states that “patents are tools for power and control over technology and people.” (Kranakis, 2007 pp. 689). A patent gives the owner of the patent a temporary right to exclude others from making, selling and/or producing the patented object (Heger & Zaby, 2017). The intention of the patent policy is to ensure appropriate returns for an inventor’s research and development activities. To obtain a patent for an invention, different criteria have to be fulfilled (Levin, 2011). First, the invention must include an inventive step. The inventive step is determined in relation to what is known of before and assessed in relation to what is considered to be a person who is skilled in the art. In addition, a patentable invention has to fulfil the criteria of novelty and has to be susceptible of industrial application. A granted patent provides the holder of a patent the right to exclude others from undertaking patented activities for a duration of 20 years from the patent application date. To retain a patent, a holder has to pay patent fees accordingly throughout the patent’s lifetime. 3.5.2.2 Copyright Authors and owners of literary-, musical-, architectural-, audio- and dramatic works, software and databases have exclusive rights to make copies of their work and make the work available to the public (Spinello, 2007). Copyright protection for artistic works gives protection that lasts 70 years 24 after the death of the creator. For databases, the duration is 15 years from when a database has been compiled (Levin, 2011). In general, no registration process is required for obtaining copyright protection. However, to be protectable, the work has to fulfil the criteria of creative effort, originality and individuality (Spinello, 2007). Copyright protected work can rightfully be used by third parties for the purposes of criticism, research, classroom instructions or news reporting. There are, however, limitations to how much of the work that can be reused. According to law, quotations or minor parts of the protected work are fair to redistribute or to reuse. 3.5.2.3 Trademarks Protection of trademarks includes protection of names, symbols and marks which ensure the holder of a trademark an exclusive right to a commercial identity (Spinello, 2007; Levin, 2011). The primary purpose of trademark protection is to hinder unfair competition from free riders taking advantage of other parties’ trademarks (Spinello, 2007). The duration of protection for a protected trademark can be eternal as long as renewal fees are paid and the trademark is used. To protect a trademark, the mark has to be distinctive and graphically representable. A firm can either protect a trademark by direct registration or indirect registration through establishment. 3.5.2.4 Design rights Design rights refer to rights to the visual appearance of a product or prototype (Levin, 2011). The holder of a design right has the sole right to exploit the design. In addition, the holder of a design right can prevent third parties from selling, importing, using and/or exporting products with designs which do not differ in the overall design compared to the protected design (Libecap & Thursby, 2008). To obtain a design right, the design must be novel and be of individual character assessed in regards to what is seen as an informed user (Levin, 2011). The duration of a registered design right varies between 10-25 years depending on jurisdiction. For unregistered design rights within the European Union, the duration of protection is three years. 3.5.3 Trade secrets In comparison to the four IPRs described above, trade secrets are substantially different. Spinello defines a trade-secret as “information used in the operation of a business that gives the owner an opportunity to obtain an advantage over competitors who do not know or use that information so long it’s secrecy is maintained” (Spinello, 2007 pp. 19). WIPO (2018) adds to that definition by stating that trade secrets have to be governed by non-disclosure agreements and if an unauthorized party gets hold of the secret information and use that information, that is unfair practice. As a control mechanism, a trade secret lasts as long as it is not revealed to the public. No disclosure or registration process is needed. Trade secrets do not as patents provide the holder with exclusive rights (WIPO, 2018). Instead the protection gained through trade secrets is highly dependent on trust. Once a trade secret leaks, the competitive advantage gained by keeping information as a trade secret may be lost. Even though 25 companies use non-disclosure agreements for managing trade secrets, trade secrets have been shown to be more difficult to legally enforce compared to patents. 3.5.4 Contractual control Contracts are in legal tradition described as “terms of an offer and acceptance, i.e. in terms of mutual promises to sell and buy.” (Petrusson, 2004 pp. 62). For business people, contracts are often seen as a necessary evil or as a safeguarding function (Haarala, Lee, & Lehto, 2010). Petrusson (2004), however, claims that contracts are essential for claiming property rights. By using contractual tools, a firm can generate business relationships and facilitate transactions. Contracts, in many ways, function as the most important tools to construct structural order for creating and extracting financial values. In addition, contracts are important tools to strengthen a firm’s control. Especially for controlling a firm's products, services and value propositions when IPRs are weak. As an example, a service-based subscription model for a software solution can be governed by contracts (Petrusson, 2004). By using contracts, the provider of the service remains in control of the software and service even though several customers have full access to the service. The IoT, as has been stated, will reduce borders between industries and collaborations between firms will increase (Aharon et al., 2015). These changes lead to concerns of data ownership. For instance, will the data ownership be with the developer, the product provider or the customer? To capture maximum value of data and to avoid misuse of data, management of data ownership is essential. To solve data ownership concerns, contracts should be applied. In general, there are four key data concerns a firm has to manage in order to reach the full potential of digitalisation (Millien & George, 2016). These are as follows: 1. Data in relation to customers a. Raw data – big data collected from smart objects/devices b. Processed data – data generated by analysing raw data c. Input data – data that is entered by the end-user 2. Data in relation to manufacturers a. The manufacturer of smart devices owns the data regardless whether if the devices are sold or leased b. The manufacturer owns the data, but the customer has the right to license-in all the data c. The manufacturer owns the data, but the customer has the right to license-in some of the data d. The customer owns the data, but the manufacturer has the right to license-in the data 3. Data in relation to collaboration partners 4. Data in relation to contribution 26 a. Data in an IoT environment is often generated by one or several customers where different contracts may be in place. It is therefore important to keep track of who contributes with what data. With growing digital environments and as firms become more digitised, firms will need increased software development for their IoT solutions (Millien & George, 2016). The software, will most likely not be developed in-house only by proprietary code. Instead, a firm probably has to in-license software, which often implies open source software. For open source software, it is the licensing terms of the software that determines the proprietary status of the software code and how it is controlled. In the most open setting, a user has to share its software contributions without any restrictions to external parties. Therefore, to reduce the risk that a firm loses its control of a product or a service, it is important that a firm does understand the licensing terms of an open source software before it is used. 3.5.5 Technical control In this study, technical control refers to access control of technologies and restricting use of proprietary hardware and copyrighted works. This type of control is commonly known as digital rights management (DRM). DRM implies a firm’s use of policies, techniques and tools to guide users of digital content into proper use (Subramanaya & Yi, 2006). Moreover, DRM influences the flow of digital content and, in addition, establishes rules of how and what digital content that should be encrypted. That to increase privacy and security protection of users. 3.5.6 Market-based control In this study, market-based control is referred to as market power. Market power is defined as the firm’s ability to influence the price and output of a market (Foss, Foss, & Klein, 2017). Market power can be achieved through different means and strategies. In this study, Porter’s five forces is the starting point for how a firm can evaluate and establish market-based control. 3.5.6.1 Porter’s five forces According to Porter (1980), the competitive edge of a firm and the market competition, are determined by five basic competitive forces, see Figure 9. If a firm uses and creates its business strategy based on the five forces, the firm will become more profitable and strengthen its market power (Porter, 1980). 27 Figure 9. Five forces which drive industry competition (Porter, 1980). Threats of new entrants refers to new capabilities which are brought to the market. These often push down pricing power of old capabilities. Thereby, profitability will be reduced (Porter, 1980). The threat of new entry depends on the barriers to enter the market and on reactions from existing actors on the market. Bargaining power of buyers refers to buyers’ capacity to force down prices and at the same time bargaining for higher quality (Porter, 1980). That can be done in several ways. For instance, having competitors to play against each other, purchasing of large volumes in relation to sales of suppliers and to use standardised products. Threat of substitutes is the degree of competition from another industry that delivers products or services which can be used as substitutes to replace current offerings (Porter, 1980). A substitute may hinder a firm to increase its pricing or altering a product. Substitutes often comes rapidly when development drives competition. This as development causes price reductions or improvement of product performance. Bargaining power of suppliers is the suppliers’ possibility to raise prices or to reduce the quality of the product/services while pricing levels remain stable (Porter, 1980). Suppliers can either be powerful by acting on their own or as a group. Suppliers possess higher bargaining power if their products are differentiated. That as differentiated products lead to higher switching costs for buyers. Rivalry among existing firms indicates strategies existing actors on the market use to create strong market positions (Porter, 1980). One example is when one player changes its price or product and competitors respond by counter moves. The responsiveness makes the firms mutually dependent. If the level of responsiveness is too high, it can have negative impact on the market and its companies. 28 3.5.6.2 Platform control Porter's five forces was created in the 1980 before the digitalized economy emerged. The model is still used. However, since one of the selected cases of this study uses a platform as a business model the authors decided to complement porter’s five forces by including theory of how different elements of platforms can be used to achieve market control. To create market control for platforms, factors such as level of openness, value creation processes, M&As, and data management have to be managed (Parker et al., 2016). Level of openness - It is a challenge to decide a platform’s level of openness to make it successful (Parker et al., 2016). The level of openness regulates how the platform can be used, monetized and how participants or developers should be governed. On the one hand, an open platform may encourage innovation. On the other hand, an open platform could make it harder for the founder(s) to control IP and to monetize from the platform. To regulate a platform’s level of openness, a manager can use different means (Parker et al., 2016). To create exclusive access to essential assets, platform managers develop rules, protocols, and barriers to reduce multihoming of users (Parker et al., 2016). Multihoming means that actors interact on several platforms to achieve similar purposes. Besides direct rules, a manager can incur charges for specific use of the platform. That indirectly makes undesired use unattractive. A platform manager has to regulate the level of openness for all stakeholders on the platform. The stakeholders can usually be categorized into users, developers, sponsors and managers. The roles and responsibilities of the different stakeholders vary between platforms (Parker et al., 2016). For some platforms, the manager of the platform is responsible for the operation and the structure of the platform while another actor is the sponsor of the platform’s technology. A platform sponsor keeps the legal control of the platform’s technology. One actor can both be the sponsor and the manager of a platform. In such case, one single entity has the overall control of the platform (Parker et al., 2016). However, when the sponsor and manager of a platform are two separate entities, the situation is different. Then the manager controls and organize interactions on the platform between customers and producers. The sponsor, on the other hand, controls the platform’s architecture and intellectual property. In such scenario, the manager is positioned closest to the actors located on the platform. Thereby, the manager has major influence on the daily operations of the platform. The sponsor, on the other hand, usually has more economic and legal control of the platform. That means that the sponsor tends to have more long-term control and impact on the platform. The level of a platform’s openness has impact on the quality of the content added to the platform (Parker et al., 2016). To achieve high quality of content and to control what is added to the platform, managers of platforms can employ different types of curation. Furthermore, by using screening and 29 feedback processes, the manager of the platform can decide which actors, developers and activities that are allowed on the platform. Value creation processes - To offer new interactions and additional value on platforms, platform managers can attract external developers by giving them access to the infrastructure of their platforms. Different types of developers can be attracted depending on what value a platform manager wants to add to the platform. Developers can be categorized into core developers, extension developers and data aggregators (Parker et al., 2016). - Core developers create functions which facilitate value creating interactions on platforms. - Extension developers offer additional value creation on platforms, e.g. applications. To facilitate a high degree of openness towards extension developers, some platforms provide application programming interfaces (APIs). An API consists of standardized tools to build software, protocols and routines. - Data aggregators are external developers which improve matching of interactions. Data aggregators use multiple sources of platform information and sell the information for e.g. advertising purposes. Data aggregators usually get platform access through licensing agreements with the manager of the platform. With different developers connected to the platform, the manager of the platform has to decide which parts that should be controlled and by whom (Parker et al., 2016). Mergers and acquisitions - One of the main purposes of M&As in platform business, is to acquire a base of users that overlaps the own platform’s user base (Parker et al., 2016). In addition, acquisitions can also be relevant if features on other platforms are so attractive that they may attract users from the own platform. To avoid that, managers of platforms either have to provide similar features on their own platforms or via partners. If a feature is provided by a partner, a platform manager can gain envelopment effects. Envelopment effects means that one platform gains new users from other platforms. Data management - To create a successful platform, management of data is important (Parker et al., 2016). Data as a competitive force can be used tactically and strategically. Tactically means that data is used for testing. For instance, to find the best position or attributes of a feature. Strategic use of data implies that data is used for analysing the platform's users' activities inside and outside the platform’s ecosystem. By using data analytics, a platform manager can gain deeper understanding of its users and their activities. Hence, a platform manager has the possibility to use the insights to offer and facilitate more efficient value creation on the platform. By controlling strategic data of a platform, it is difficult for competitors to offer similar interactions. That means that platforms can use data to create what Porter calls a barrier to enter the market. 30 3.6 Theoretical framework The theory used in this study has been synthesized into one theoretical framework. To identify IoT generated values, IoT theory, IoT enabled business models, the IoT value model has been included in this study's theoretical framework. The technology tree has been included in the theoretical framework in order to disaggregate IoT values into technology components. This makes it possible to approach the components from an IPR perspective. Thereby it is possible to examine how different building blocks can be used to leverage the value on the market. The building blocks of this theoretical framework are: rights-based control (patents, copyrights, trademark protection and design rights), trade secrets, contractual control, technical control, and market control. 31 4. Case studies The following chapter contains the empirical results of the two cases studied in this study. The chapter includes the following sections for each case; introduction, value proposition, technology breakdown, IoT value model, intellectual assets and IPRs. 4.1 The Tork® EasyCube® The following section presents the case study of the Tork EasyCube. The section starts by giving an introduction to the Tork EasyCube. The section then continues by describing The Tork EasyCube’s value propositions, its technology breakdown and its generated values. 4.1.1 Background information Essity was founded in year 1929 as a forest company called SCA. Over time, the company evolved from being a pure forest company to become a company that also offered personal care and tissue products (SCA, 2018). In year 2017, SCA was divided into two companies. The forest product company SCA and the hygiene- and health company Essity. In the time being, Essity is one of the largest hygiene- and health companies in the world. One of Essity’s largest brands is Tork. Tork includes two main product lines; dispensers and refills (Essity, 2018a). During recent years, in relation to Tork, Essity has introduced a new product on the market, the Tork EasyCube. The Tork EasyCube is a web-based service that offers data driven cleaning processes to its customers, see Figure 10. The Tork EasyCube collects and analyses data from dispensers and devices. Thereby it can suggest when a customer’s area is to be cleaned or when a dispenser needs refill. The Tork EasyCube increases work efficiency and reduce customer complaints. Customers interact with the Tork EasyCube through a web-based application. Figure 10. An illustration of the Tork EasyCube and its three main modules; (1) the sensor technology that measures the number of visitors, levels of refill and the transmits data, (2) the data is collected and aggregated on Essity’s servers and (3) that customer access the data through the Tork EasyCube web application. 4.1.2 Value proposition The main value proposition of the Tork EasyCube is smarter facility cleaning with digital intelligence (Essity, 2018b). The digital intelligence enables a new level of efficiency and effectiveness of cleaning operations. In addition, the user web-application enables users to work together with the Tork EasyCube. The increased efficiency and effectiveness achieved by the Tork EasyCube, have positive impact on manpower and reduce cost. Through the Tork EasyCube’s analytics, customers receive data of when, where and what dispensers that need to be refilled as well as the number of visitors per hour 32 for designated areas. The analytics brings the following benefits; higher customer satisfaction, increased staff engagement and real-time data. The batteries needed for the sensors of Tork EasyCube have a lifetime of five years. The sensors of the Tork EasyCube communicate with local gateways installed at customers’ sites. The communication between sensors and local gateways is carried out by a proprietary protocol of Essity. The gateway then relays the information to Essity’s server via GSM/GPRS (Global Innovation Manager at Essity Hygiene and Health AB, interviewed March 3, 2018). The route of communication, from local sensors to Essity’s server, does not need involvement of customers’ IT-divisions. Keeping the Tork EasyCube and customers’ IT-systems apart, reduces the risk of system breakdowns and complicated integration processes and the data can be kept safely stored. 4.1.3 Technology tree The Tork EasyCube consists of a multi-layer technology infrastructure. The main components of the Tork EasyCube are the device, gateway, server and user interface (Essity, 2018c). Figure 11 illustrates the overall hierarchy of the Tork EasyCube’s technology. Figure 11. Technology breakdown of the Tork EasyCube. The dispenser comprises of a container, refill material and a sensor communication unit. The sensor communication unit’s components are the sensor, the software that defines the functions of the sensor and the battery. The radio protocol used for communicating sensor data from the dispenser to the local gateway has been developed in-house and is proprietary of Essity. The local gateway, through the data communication unit, receives data from sensors, processes sensor data and transmits data to Essity’s server at specific intervals. The gateway consists of a power adapter, a data communication unit and a router. The server includes a database and programs for analytics. The analytic programs provide two types of analysis, (1) analysis for Essity and (2) analysis for customers. Analysis provided to 33 customers are delivered via the web-application (the user interface). The user interface includes the Tork EasyCube Web that is accessible via an URL-login. The URL-login can be accessed from any device with internet connection. For instance, a tablet, phone or a computer. There are two different types of user interfaces, one for facility managers and one for cleaning staff. Each interface displays graphics to match each user group’s needs and requirements (Global Innovation Manager at Essity Hygiene and Health AB, interviewed March 3, 2018). 4.1.4 IoT value model The theoretical foundation of this study describes that new types of values are generated by the IoT. The majority of these values will be be created by multiple actors in an ecosystem (Turber et al., 2014). Such an ecosystem will be structured differently depending on the characteristics of the IoT solution. The following section is the result of applying the IoT value model framework to the Tork EasyCube. The section visualizes and explains the different values created in each IoT layer of the Tork EasyCube. Based on the technology breakdown of the Tork EasyCube, nine different value layers were identified, see Figure 12. Due to confidential information, suppliers have only been categorized into software suppliers and suppliers of physical goods. Figure 12. 2D visualization of the Tork EasyCube’s IoT value model. Refill - The first value layer is the refill layer. Refills are offered by Essity to customers. Customers pay-as-they-use. Thereby Essity and Essity’s suppliers receive monetary value from a revenue model (Global Innovation Manager at Essity Hygiene and Health AB, interviewed March 3, 2018). Dispenser - Dispensers generate monetary values for both Essity and its suppliers when dispensers are sold. In addition, the dispensers bring economic benefits to Essity as buyers of dispensers are likely to buy refills from Essity as well. SCU and visitor logging - The third and fourth value generating layers are the SCU-layer and the visitor logging layer. The sensors and the technology within the sensors are delivered by two types of suppliers, (1) suppliers of physical components and (2) suppliers of sensor technology and sensor software (Global Innovation Manager at Essity Hygiene and Health AB, interviewed March 3, 2018). Both types of suppliers receive monetary value when their products are bought by Essity. Essity, on the other hand, obtains value defined as competence. This as sensors provide Essity with customer 34 data that can be used to gain insights and knowledge of customer needs. That data can later be leveraged to strengthen customer offerings. Gateway and Server - The gateway and the server do not provide customers with any direct value. However, the hardware needed for the gateway and the server are provided by a third party. That party receives monetary value when the hardware is sold