How to commercialize IoT? : exploring the potential of the Internet of Things to create value through business

(1)

How to commercialize IoT?

Exploring the potential of the Internet of Things to create value through business

models

Author: R.D.R. Bergen Master Thesis

Business Administration – Entrepreneurship & Innovation Management

Supervisor: Ieva Rozentale MSc Second reader: Prof. Dr. Peter van Baalen

Student number: 11148055 August 19th 2016

(2)

Statement of originality

This document is written by Régy Bergen who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been

used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(5)

I. Abstract

The Internet of Things interconnects and combines physical objects with sensors, data

management and communication methods, through the Internet. Industry boundaries are changed since this new technology creates value for customers and develops new business models. New technologies are commercialized through business models. However, it is still unclear how IoT captures value. Little academic attention is provided on this topic. Therefore, this study focuses on how businesses exploit a commercial value from IoT through their business models.

Qualitative research was conducted to examine business models of eight company cases related to IoT. Findings reveal that since there are currently no standardized techniques for capturing value form IoT, companies have to make specific choices regarding the elements in their business models. These choices predominantly entail: the purpose of value propositions, a company’s resources and partnering company’s resources, and the types and methods of revenue streams.

(6)

1. Introduction

During the past four decades some quite interesting changes in the nature of doing business have occurred. Emerging economical markets, rapid growing new technologies, and industry

developments have resulted into new opportunities for businesses. The Internet has played an important role in the technological progress, and in developing new ways of doing business (Timmers, 1998). New environments have emerged, in which the Internet is a source of

information, and considered to enable economical exchange between supply and demand. Hence, the Internet has transformed industries (Teece, 2010). As a result, businesses entered online markets to create value for customers and to capture value from delivering products or services (Teece, 2010). An evolution has emerged in the Information Technologies (IT) industry, where IT has now become an integrated part of products and services (Porter & Heppelmann, 2014). This entails machines becoming ‘smart’, in the sense that they are embedded with sensors, data management, and improved measuring techniques. The Internet has expanded this technological development, because machines have now become interconnected with it. These revolutionary advancements are referred to as the ‘Internet of Things’ (IoT) (Porter & Heppelmann, 2014). The idea behind this concept is to interconnect physical objects, such as machines or mobile devices, with sensor technologies, actuators, data communication technologies, and the Internet (Atzori, Iera & Morabito, 2010; Fleisch, 2010; Gubbi, Buyya, Marusic & Palaniswami, 2013).

According to Cisco (2013) the Internet of Things plays a major role in the global technology market, as they estimate that it will generate $14 trillion in profit in the upcoming decade. This reflects the financial potential of IoT. Furthermore, The World Economic Forum (2015) in collaboration with Accenture questioned 250 market leaders about the market potentials of IoT. Respondents answered that IoT unleashes potentials for value creation,

(7)

disrupting industries, and developing new business models. These perceptions lead to questions stated by The World Economic Forum (2015): what will be the implications for new IoT business models, and how do businesses create value from IoT business models. A concluding remark underpins that IoT is perceived as a new technology that changes industry boundaries. However, it is difficult to create value from a new technology, because value remains undisclosed until it is commercialized through a business model (Chesbrough & Rosenbloom, 2002). Thus, it is of interest to examine business models for IoT, considering that the commercialization of IoT is expected to occur through business models. In particular, little academic attention has given to the former and the latter. Therefore, it is still unclear how to commercialize IoT through business models. For the above-mentioned reasons this study examines: value creation from IoT through business models. Accordingly, the main research question is as follows:

How do businesses exploit a commercial value from the Internet of Things, through their business model?

1.1. Theoretical and practical relevance

In academic literature little attention is given to the topic of IoT, because it is still a relatively new topic, in as such that empirical evidence is almost absent. To date, IoT covers the area of technology since it is viewed as a revolutionary technology. In academic literature IoT is predominantly approached from the discipline technology (Gubbi et al., 2013). However, IoT also contributes to the business domain via: creating value for customers, developing new business models and disrupting industries. This area of research is currently disregarded in academic literature. Therefore, this study contributes to exploring the topic of business models for IoT, since this have not yet received considerable attention.

(8)

Furthermore, it is interesting to explore the business side of IoT, since 72% of the 250 market-leader respondents stated that IoT is expected to disrupt not only business but also industries, and 79% of them expect it to occur within the upcoming five years (World Economic Forum, 2015). Hence, the managerial and practical relevance for this study relates to identifying opportunities on how to commercialize and create value for IoT through a business model. Specifically, the profit speculations mentioned above advocate a thorough examination of this field.

1.2. Research aims and design

The overarching objective of this study is to obtain new insights and explore how a new technology, namely, the Internet of Things, is capitalized through a company’s business model. This objective can be achieved by: (1) analyzing companies’ business models that are used to commercialize IoT, and (2) exploring the role that IoT has in value creation processes within these companies. To investigate these aims, a qualitative research was carried out consisting of 12 in-depth interviews. These interviews were held with five co-founders of businesses in software-hardware, IT, healthcare, and consumer goods, as well as, five IoT representatives of businesses in software-hardware, IT, telecommunications, and aviation. In addition, two IoT professionals were included who elaborated on the general topic of business models for IoT, without specifically referring to their company’s business model. Data was analyzed with a software program ATLAS.ti, and these analyses were used to answer the research question.

(9)

1.3. Structure thesis

The succeeding chapters include information on the research context. The next chapter contains the theoretical framework, which presents theoretical discussions on the main topics of: business models, the emergence of IoT, and IoT business models. Following these discussions, attention is given to a theoretical synthesis, in which sub questions are presented. The third chapter

elaborates on the methodology of this study, which is continued by the chapters presenting the results, discussion and conclusions.

(10)

2. Literature Review

In this chapter the main theoretical concepts within this study are discussed. First, definitions of business models are discussed, and topics related to definitions on business models are allocated. Followed by the topics: business models for new technologies, virtual markets and e-commerce, and the Internet of Things.

2.1. Business models definition

Although, various scholars have examined the concept of business models, no consensus exists on the definition, rather dispersed perspectives of the concept (Morris, 2005; Shafer, Smith and Linder, 2005; Zott et al., 2011). Table 1 depicts several definitions and perspectives on business models. They are obtained from several syntheses of business model definitions, conducted by different scholars. Definitions in table 1 are selected based on their occurrence in peer-reviewed journals, and the number of citations.

Table 1: An overview of selected business model definitions.

Authors: Definition of business model:

Timmers (1998, p. 2) ‘’A business model is an architecture for

the product, service and information flows, including a description of: the various business actors and their roles, the potential benefits for various actors, and the sources of revenues’’.

Amit and Zott (2001) The business model is a unifying unit of

analysis to capture value creation that arises from multiple sources. The business model portrays the design of content, structure, and governance of transactions to create value through the

(11)

Chesbrough and Rosenbloom (2002) The business model provides a coherent framework with embedded technological

components as inputs, and converts them through customers and markets into economic outputs. Also, it is considered to facilitate a connection between technological developments and economic value creation.

Magretta (2002) The business model tells a story by

answering who the customer is, what the customer values, how the business makes money, and what the underlying economic logic is that explains how the business delivers value to the customer at an appropriate cost.

Morris, Schindehutte and Allen (2005, p. 727) ‘’A business model is a concise representation of how an interrelated set of decision variables in the areas of venture strategy, architecture, and economics are addressed to create sustainable competitive advantage in defined markets’’.

Osterwalder, Pigneur and Tucci (2005, p. 3) ‘’A business model is a conceptual tool containing a set of object, concepts and their relationships with the objective to express the business logic of a specific firm. Therefore we must consider which concepts and relationships allow a simplified description and representation of what value is provided to customers, how this is done and with which financial consequences’’.

Nine building blocks are based on four pillars: (1) value proposition; (2) target customer, distribution channel, and relationship; (3) value configuration, core competency, and partner network; and (4) cost structure and revenue model.

(12)

Al-Debei and Avison (2010, p. 373). ‘’An organization’s understanding of its business model could be viewed as novel strategic-oriented knowledge capital that is crucial for business organizations in an emerging, turbulent, and digital business environment’’.

The main dimensions are: value

proposition, value network, value architecture, and value finance.

Teece (2010) A business model reflects the company’s

hypothesis about the customers’ needs and how the company makes profit from meeting the needs, hence, it reflects the questions on what the

customers want, how they want it, what customers will pay for it, and how the company gains profit from it.

This table shows that varied perspectives on definitions of business models exist, resulting in both differences and similarities. Chesbrough and Rosenbloom (2002) believe that business models facilitate a connection between technological developments and economic value creation. While Magretta (2002) refers to business models as telling a story about who the customer is and what their values are, in order to profit as a company from the delivered value. Teece (2010) adds to Magretta’s views that a business model portrays a company’s hypothesis on the

customers’ needs. According to Timmers (1998), business models constitute a firm’s architecture for product, service and information flows. Similar to this view, Morris et al., (2005) also refer to an architecture. However they consider architecture to be part of business models, alongside economic and strategic characteristics of a business. This architecture encompasses the internal processes and infrastructure of a business to create value (Morris et al., 2005).

(13)

Comparable to the view of business models as architecture, is the perspective of Amit & Zott (2001), who construct business models as a unifying unit of analysis. The authors argue that business models are an important source of value creation for a firm, suppliers, partners and customers. They add the notion of Internet business models, and refer to ‘e-business’, which is conducting business over the Internet, and they argue that firms exploit business opportunities through the Internet (Amit & Zott, 2001). Furthermore, other scholars perceive business models as a framework or a conceptual tool (Chesbrough & Rosenbloom, 2002; Osterwalder et al., 2005). In this regard, a coherent framework is considered to connect technology with economic value. Osterwalder et al., (2005) ague that business models link strategy, technology with the organization. Al-Debei and Avison (2010) argue that a business model is a conceptual tool with four main dimensions, namely, value proposition, value network, value architecture, and value finance. These dimensions are similar to the four pillars of a business model framework, argued by Osterwalder et al., (2005). The authors describe the pillars as: product, infrastructure

management, customer interface, and financial aspects. Additionally, Osterwalder and Pigneur (2010) extend this framework to a blueprint, which describes how an organization creates, delivers, and captures value. They define the blueprint as a Business Model Canvas, which includes nine building blocks to cover the four pillars, as presented in table 2 (Osterwalder & Pigneur, 2010). The building blocks describe the components within a business model (Osterwalder et al., 2005). The first pillar entails the building block value proposition. The second pillar contains building blocks: key resources, key partnerships, and key activities. The third pillar consists of building blocks: customer relationships, customer segments, and channels. The fourth pillar covers the financial aspects.

(14)

Table 2: Definitions of the nine building blocks within the Business Model Canvas (Osterwalder & Pigneur, 2010, p. 20-40).

Pillar: Building block: Definition:

Products Value Propositions ‘’The bundle of products and services that create value for a specific Customer Segment’’.

Infrastructure management

Key Resources ‘’The most important assets required to make a

business model work’’.

Key Partnerships ‘’The network of suppliers and partners that make the business model work’’.

Key Activities ‘’The most important things a company must do to make its business model work’’.

Customer interface Customer Relationships ‘’The types of relationships a company establishes

with specific Customer Segments’’.

Customer Segments ‘’The different groups of people or organizations an enterprise aims o reach and serve’’.

Channels ‘’How a company communicates and reaches its Customer Segments to deliver a Value

Proposition’’.

Financial aspects Cost Structure ‘’All costs incurred to operate a business model’’.

Revenue Streams ‘’The cash a company generates from each Customer Segment (costs must be subtracted from revenues to create earnings)’’.

The section above addresses different views of definitions on business models, however, within these views similarities can also be found. For instance, several authors argue that the overall objective of doing business is value creation. In this sense the purpose of business models is referred to as identifying who the customers are and how to create value for them (Amit & Zott, 2001; Chesbrough & Rosenbloom, 2002; Magretta, 2002; Teece, 2010). When defining business models, financial characteristics, internal processes and a company’s infrastructure are often addressed. More specifically, the financial characteristics of a company refer to what customers

(15)

discussed before, in some perspectives an architecture is used to represent the internal processes or infrastructure of a firm (Morris et al., 2005; Timmers 1998). Moreover, a framework

encompasses a combination of value creation, infrastructure management, customer interface, and financial aspects (Al-Debei & Avison, 2010; Osterwalder, Pigneur & Tucci, 2005).

2.1.1. Classifying topics on business models

Several scholars have created syntheses of topics that are related to business models. For

instance, Morris et al., (2005) argue that these topics involve: value creation, core competency of a company, strategically positioning a business in a marketplace. While Zott et al., (2011) argue that topics on business models relate to: e-business and information technology in organizations, strategic issues to create value, and innovation and technology management.

Thus, with respect to the research question in this study, academic literature is of interest in the areas of: e-business, information and technology, and value creation. In particular, this study acknowledges a definition of business models that focus on creating value, and delivering value to customers, in order to capitalize on technology. In this regard, economical value of technology is only revealed through a business model, because business models can bridge technology and economic value creation (Chesbrough & Rosenbloom, 2005; Osterwalder et al., 2005). Therefore, this definition can be adopted to explore how business models commercialize a new technology. In the proceeding section, literature is discussed with regard to the research question on the topics: business models for a new technology, virtual markets and e-commerce, and Internet of Things.

(16)

2.2. Business models for new technologies

This section discusses that a business model mediates between technology and economic value creation. For a company that creates new technologies or ideas it is important to understand that technology by itself does not have a single objective value, in fact, the economic value remains undisclosed until it is commercialized via a business model (Chesbrough, 2010). There is no single independent value for technology, since it has the tendency to be developed in different ways, and a customer can value a technology to a degree that it offers a solution to a problem (Chesbrough & Rosenbloom, 2002). According to Zott and Amit (2010), a business model offers opportunities to create and capture value for all involved parties, in as such that it considers fulfilling customer’s needs and at the same time it generates profit. In this regard, business models can be helpful to understand how businesses capitalize value from a new technology.

Furthermore, evolution in technology and the emergence of the Internet contributed to the developments of: new businesses, new industries, and new information and communication methods (McGrath, 2010; Teece, 2010). Therefore, the Internet has developed an area for companies to construct an array of new and various business models (Chesbrough &

Rosenbloom, 2002). As a result, profound changes have occurred for businesses on how they operate and create economical exchange (Amit & Zott, 2001). However, Teece (2010) argues that the growth of Internet raises questions for a company on how value is created for a customer, and how value is captured from delivering new information services through the Internet. Thus, it is of interest for this study to see that developments in technology and the Internet have resulted into businesses operating in new established markets. However, it is unclear how online businesses create value from delivering products and services. Regarding

(17)

these notions, literature is further discussed in the next section, on virtual markets and e-commerce.

2.3. Virtual markets and e-commerce

This section reviews advances in technological and innovation settings, with an emphasis on the emergence of virtual markets. According to Amit and Zott (2001, p. 495), ‘virtual markets are settings in which transactions are conducted via open networks constructed on both fixed and wireless Internet infrastructure’. In this regard, the Internet is not just a digital source for information, it is also a method to create economical exchange. Electronic commerce relates to conducting business via the Internet, which is complementary to traditional businesses

(Timmers, 1998). In essence, trading goods, products, and services via the Internet created new opportunities for existing and new companies. The evolution of virtual markets resulted into new trading mechanisms and transaction methods for businesses (Amit & Zott, 2001). Where

traditional markets conduct commerce offline, within the virtual markets business is conducted online. For example, a traditional buyer and supplier transaction method occurs on a physical location, whereas in virtual markets no physical locations are required. Therefore, within these online environments, companies require new methods to create, deliver, and capture value (Teece, 2010). These methods are developed through technological innovations, according to Chesbrough & Rosenbloom (2002). Technological innovations require a business model in order to commercialize the potential of technologies (Chesbrough & Rosenbloom, 2002). Thus, the Internet is an innovative technological advancement in creating economic value within virtual markets, and it is interesting to understand how firms construct new business models in newly created market settings. Complementary to technological developments in online markets, is the

(18)

rise of the Internet of Things (IoT). This technological and innovative concept, as well as, business models for IoT is discussed in the proceeding section.

2.4. Internet of Things specifications

In the preceding sections several theoretical approaches to business models, virtual markets and e-commerce have been discussed. This section discusses specifications of IoT, and the type of business models for IoT that are addressed in academic literature.

In the past two decades, an evolution has occurred in information and technologies (IT) in which physical products and machines, which were composed solely of mechanical and electrical parts, gradually have become complex systems combining software, hardware, and even Internet connectivity (Porter & Heppelmann, 2014). IoT is referred to as the third IT-driven wave, according to Porter and Heppelmann (2014). In this regard, they differentiate IoT from the first and second waves respectively, by arguing that both waves brought the rise of automated systems and the Internet. Hence, IoT addresses both the first and second wave, while also including sensors, software, and cloud and data services to analyze data. In as such that, IoT becomes an integrated part of products (Porter & Heppelmann (2014). These views are in line with the definitions of IoT, which are presented hereafter.

‘’The Internet of Things (IoT) is a novel paradigm that is rapidly gaining ground in the scenario of modern wireless telecommunications’’ (Atzori et al., 2010, p. 2787).

‘’In the Internet of Things paradigm, many of the objects that surround us will be on the network in one form or another’’ (Gubbi et al., 2013, p. 1645).

‘’The basic idea of IoT is that virtually every physical thing in this world can also become a computer that is connected to the Internet’’ (ITU, 2005, as cited in Fleisch, 2010 p. 127).

(19)

In essence, IoT involves a system of technologies that connects machines or devices with sensors, actuators, and computers to the Internet, in order to communicate and gather data without human involvement (Atzori et al., 2010; Gubbi et al., 2013).

In 1999, Kevin Ashton introduced the term IoT in the context of Radio-Frequency

Identification (RFID) tags, to identify physical objects’ traceability and status of location (Atzori et al., 2010). Hereafter, a broader context of IoT emerged, which originated from a ‘smart

environment’. In the same decade Mark Weiser introduced this term, and refers to an interwoven world that consists of physical products connected to a virtual world, but they are operating almost invisibly in our daily lives (Gubbi et al., 2013). Such an environment consists of multiple connected technical appliances that communicate almost autonomously with each other. Mattern and Floerkemeier (2010) add ‘smart objects’ to the description of a smart environment, and argue that it has the potential to revolutionize the utility of products and services. Fleisch (2010) acknowledges the idea of smart objects working autonomously with their surroundings, but he adds that there is a difference between IoT and the Internet. Where IoT uses low energy consumption of computers, the Internet uses high capacity of computers. And, he continues by referring to IoT as an almost and complete exclusion of humans (a machine-to-machine communication method), whereas with Internet services humans are the users. Porter and Heppelmann (2014) argue that smart connected products have three core elements, namely physical components, smart components, and connectivity components. In essence, they refer to hardware, software, and network connectivity. Atzori et al., (2010) divide IoT into three visions. Firstly, the Internet oriented vision, which is the network connectivity via an Internet Protocol (IP). Secondly, the things oriented vision, which are technologies such as sensors, devices, and systems to enable traceability or visibility of an object. Lastly, a semantic oriented vision, which

(20)

is analysing data and making sense of what knowledge and information is obtained from data. Gubbi et al., (2013) add to this view that the usefulness of IoT is enabled when all three

interconnect. To conclude, IoT entails physical objects such as devices, sensors, machines or any other computer appliances are used almost invisibly and autonomously in our surroundings, and they are made smart because of their new functionalities. Namely, they are interconnecting information, data, communication, and Internet. So, IoT requires products and technologies to function differently or in a new way within businesses in order to create value. Hence, a business model is a necessity to flourish the potential of such new technologies, as argued by Chesbrough and Rosenbloom (2002). Therefore, in the following section business models for IoT is

discussed.

2.4.1. Internet of Things business models

Gubbi et al., (2013) highlight that IoT contains potentials for businesses. However, physical objects obtain information from the environment, while using the Internet to communicate messages to a server, where data is analysed for end-users. As a result, these physical objects use a different set of technological applications, and not merely mechanical applications.

These notions indicate that businesses and industry boundaries are often restructured, because IoT products combine physical, smart and network components (Porter & Heppelmann, 2014). Additionally, smart and connected products allow radical technologies to open up a spectrum of new business models for capturing value. As discussed earlier, business models can link

technologies and economic value creation (Chesbrough & Rosenbloom, 2002). In this regard, a business model framework can be adopted (Osterwalder et al., 2005). In line with these notions,

(21)

business models for IoT are discussed in academic literature. However, existing empirical evidence is rather limited.

Dijkman, Sprenkels, Peeters and Janssen (2015) stress the importance of developing business models for IoT, and argue that the Business Models Canvas helps a company regarding business model design. The authors argue that this framework: designs the business model as a blueprint of the company, is frequently used in business environments, and is cited more than other business model frameworks in literature (Dijkman et al., 2015). Prior to this research the authors conducted a literature survey on business models for IoT, which showed that the area is relatively unexplored (Dijkman et al., 2015). A sequential exploratory research approach was used to obtain empirical data, within a wide scope of industries. These industries were:

agricultural, energy, healthcare, smart home, smart buildings, supply chain, and transportation (Dijkman et al., 2015). A maximum of three interviews were conducted per sector. Results from 72 surveys, showed that value proposition is considered the most important building block as revealed by a 95% confidence interval, at <0.01 level of significance. In addition, data from 11 interviews showed that value proposition, customer relationships and key partnerships are considered relatively more important than the other building blocks (Dijkman et al., 2015). Further research may be required, because the study by Dijkman et al., (2015) is considered to be an explorative research in a field of study that has received relatively limited attention.

Therefore, results may not have been entirely conclusive.

In contrast to the previously mentioned empirical research, Westerlund, Leminen and Rajahonka (2014) argue that businesses in IoT require a change of focus on business models. They refer to developing ‘ecosystem business models’. In their view, they adopt business models in an ecosystem, which consists of multiple interconnected stakeholders, for instance, suppliers

(22)

of physical objects, services, software, and data, or an open source user community. Within this ecosystem perspective, not the objectives of a single company are emphasized, but the entire system of connected businesses is considered (Westerlund et al., 2014). Ecosystem business models comprise of a set of pillars, which is identical to the pillars from the Business Model Canvas by Osterwalder and Pigneur (2010). The former pillars are anchored in an ecosystem, and therefore they include methods of creating and capturing value for all stakeholders in an ecosystem, instead of value creation for a single business (Westerlund et al., 2014). However, the commercialization of IoT faces challenges regarding innovation immaturity, and IoT business model design (Westerlund et al., 2014). IoT business models primarily lack standardization, structure, governance, stakeholder roles, and value creation (Westerlund et al., 2014). To

overcome these challenges, a proposition is made to focus on business model design tools. In this regard a business model framework, such as the Business Model Canvas could be adopted. However, they argue that existing frameworks, due to their focus on building blocks of business models, provide a fragmented picture. Therefore, they ‘fail to explain the dynamics between the components’ of business models (Westerlund et al., 2014, p.9). Nonetheless, these are

perceptions, and empirical findings would be required to evaluate them.

To conclude, regarding empirical evidence of developing business models for IoT, it is of interest to see that Dijkman et al., (2015) argue to adopt the Business Model Canvas. While Westerlund et al., (2014) agree with the notion to adopt this framework, however they also emphasize that it does not entirely capture the dynamics between business model components. Both views are of interest to this study because the former view is based on empirical evidence, while the latter somewhat criticises the adopted framework method.

(23)

2.5. Theoretical framework synthesis

In conclusion, considering the previous theoretical discussions, it becomes clear that various definitions of business models exist within academic literature. Both differences and similarities have been discussed. The evolution of technologies and the emergence of the Internet resulted into online environments, where business is conducted. New methods to create, deliver, and capture value are developed through technological innovations in virtual markets (Teece, 2010). The Internet of Things emerged as a revolutionizing technology to create new businesses and economical markets. Additionally, IoT entails a system of technologies that are interconnected through, for example, machines, devices, sensors, data gathering systems, and the Internet (Gubbi et al., 2013). A challenge for companies with IoT products and services is to capitalize value from IoT (Westerlund et al., 2014). Crucial for companies that deploy new technologies is unlocking the technological potential in such a way that it is a merit. Technology itself does not come with a certain value. A condition to benefit from it’s potential and to create economic value is through a business model (Chesbrough and Rosenbloom, 2002). Pioneers in conducting

research on business models for IoT have adopted a framework by Osterwalder and Pigneur (2010), which is the Business Model Canvas (Dijkman et al., 2015). The topic of business models for IoT is relatively unexplored. Hence, the main research question investigates and explores how a radical technology, such as IoT, is commercialized through a business model. Therefore, the Business Model Canvas is adopted in this study, to bridge technology and economic value creation. To structure the main research question in this study, several sub-questions are posed. The first sub-question investigates IoT products or services of companies.

(24)

The second sub-question investigates what elements of a business model are crucial for capturing value. The first and second sub-questions are linked to the four pillars of the Business Model Canvas. Furthermore, a last sub-question explores challenges for IoT business models, since it is still unclear how to capitalize on a radical technology such as IoT. Consequently, the

sub-questions within this study are posed as follows:

- How are businesses’ value propositions related to IoT?

- What elements of a business model are crucial for capturing value from IoT? - What are the challenges in developing business models for IoT?

2.5.1. Conceptual framework

Figure 1: conceptual framework

A visual representation of the discussed main concepts within this study is illustrated. As mentioned in the first part of section, business models bridges technology and economic value creation. This study investigates IoT, which a new technology, and economic value creation related to it. Therefore, business models are viewed as an interrelated concept between the former and latter, which is depicted by the dotted lines.

Further explanation is addressed in the proceeding section methodology, on how this study is conducted, and how to answer the main research question and sub-questions.

(25)

3. Methodology

In this chapter, an outline is provided on how this study on business models for the Internet of Things is carried out. The first part of the chapter explains in depth the research approach and strategy. The second part outlines the research design, case selection, research population and sample. The third part describes the operationalization of constructs, in which the characteristics of business models that are investigated in this research are specified. The fourth and last part comprises of the data collection and data analysis.

3.1. Research approach and strategy

Seeing that not much prior literature is available on our topic, we are seeking new insights, asking new questions, and examining new phenomena, tasks fundamental to explorative research studies (Saunders & Lewis, 2012). The most appropriate way to conduct an explorative study is by conducting a literature study, conducting interviews, and interviewing experts or professionals in the field of the subject (Saunders & Lewis, 2012). These three methods are used for

investigating and answering the main research questions of this research study. This study focuses primarily on comprehending the core of IoT related business models, and how

companies with IoT applications behave. If limited knowledge on phenomena exists in literature, an inductive research approach that allows theory to emerge from data can be a valuable starting point (Siggelkow, 2007). Cases from an inductive research approach can also contribute to enrich existing theory (Siggelkow, 2007). According to Saunders and Lewis (2012), an inductive

approach places emphasis on a close understanding of the research context. Qualitative inductive research investigates environments in their natural occurrence and uses social actors’ perceptions and explanations to understand phenomena (Rynes & Gephart Jr., 2004). Hence, induction

(26)

possesses a more flexible structure, in which the development of theory evolves as a result of analyzing collected data (Saunders & Lewis, 2012). Therefore, in this study a qualitative

inductive approach is adopted in the sense that it is informed by existing literature and theoretical thinking, but stays open to additional, or alternative explanations that are acquired from case studies (O’Reilly, 2012). The following section entails an elaboration on the research design and the type of case study.

3.2. Research design and case analysis

Case studies are often used for the above-mentioned research approaches, making it is applicable to this research. In this study the boundaries of the context are not clearly evident, which is an essential notion for case studies (Yin, 2009). Multiple case studies have stronger designs for theory building than single case studies (Yin, 2009). Therefore, a multiple case study analysis is conducted to illustrate how business models with IoT applications are characterized and used by companies. This design requires determining whether results of one case also occur in other cases, and subsequently to what extent evidence is transferable to other cases (Saunders & Lewis, 2012). In this regard, Miles, Huberman and Saldaña (2014) argue that in essence, generalizing evidence from one case to another is the basis to match underlying theory. Therefore, case studies are not necessarily about testing stated theories, which is referred to as probability, rather they are about discovering explanations and the development of theories, which is a non-probability type of research (Eckstein, 1975, p. 133, cited in Thomas, 2011).

Furthermore, design possibilities can be considered when case studies are used (Thomas, 2011). This study builds on a parallel study of multiple cases, where all the studied cases occur simultaneously. The appropriate number of cases that are required in a study depend on the

(27)

2014). A suggestion is to have a minimum of five richly researched cases with appropriate

information (Miles, Huberman & Saldaña, 2014). Information is appropriate when it enables data saturation and establishes data analysis (Creswell & Miller, 2010). For cases with highly

complex data, more than ten cases can become inconvenient or difficult in managing data (Miles, Huberman & Saldaña, 2014).

3.2.1. Unit of analysis

A unit of analysis defines cases and for this study they allow the research question to be answered (Yin, 2009). In this study, a company is referred to as a case, and several companies form the units of analyses. To obtain more detailed distinction about the units of analyses, the object and subject are described (Thomas, 2011). The object in this study is referred to business models for IoT, which is the phenomenon to be explained. The subjects in this study are IoT professionals and innovation managers that do the explaining about the object.

3.3. Research population and sample

The research population defines a set from which a sample is drawn (Eisenhardt, 1989). The research population for this study is the entire set of companies that have their core business in IoT, and they have smart products or services with technologies to interconnect physical objects such as devices, sensors, data, and network connectivity. These companies are based in the Netherlands. The sample from this population set consists of companies operating in the

software-hardware or IT sector that offer IoT products or services. As mentioned in the literature section, IoT is perceived as a radical technology. Therefore, the sample of this study entails companies involving products and services in a technology sector. Hereafter, information is

(28)

provided on how cases are selected. Followed by pointing out the method that is used to target respondents, and the final sample that participated.

3.3.1. Case selection

The cases are specifically selected to create different perspectives on the same subject. This purposive sampling method is used to actively search for a select number of cases that will help to answer the research questions (Saunders & Lewis, 2012). This method is used to understand what is happening with specific phenomena (Saunders & Lewis, 2012). Therefore, this method is adopted in this study to understand specific phenomena regarding the commercialization of IoT through business models. As a result, cases are selected upon specific criteria, which entails that companies must involve a value proposition based on products or services related to IoT.

Initially, cases are screened to identify which companies are appropriate and, which companies can be selected for the sample. Cases were selected based on resemblances in characteristics of companies’ offered smart products or services, which are related to IoT. This predominantly entails products and services with embedded software and hardware that are interconnected with machines or devices, sensors, data management, and Internet.

Furthermore, Creswell (2005) argues that a ‘purposeful maximal sampling’ technique can be used to show various perspectives on the investigated phenomena. Therefore, in this study cases are also selected that are not operating in the sector technology, but have resemblances in characteristics of the offered IoT products and services.

(29)

3.3.2. Target respondents

After screening companies, 22 companies were identified in the industries of: IT/ICT, software, hardware, telecommunications, smart-industrial, healthcare, consumer goods, and aviation. 18 companies were contacted via an introductory email about the purpose of this research, and a request to participate. They were targeted via webpages of IoT events in the Netherlands that mentioned the companies’ involvement in such events, LinkedIn, and via networking. The remaining four companies were not targeted because of the limited time for this research study. After the companies were targeted, 13 out of these 18 companies did not respond to the email. Another email was sent within approximately two weeks, with a request to participate. However, the companies did not reply. Two companies declined the offer to participate, because they considered it important to maintain their business information private. Companies that approved to participate were contacted via e-mail to schedule a date, time and location.

The final sample that agreed to participate in this research consisted of eight cases and 12 informants. Representatives of eight companies were either co-founders or IoT managers. Two out of the eight companies allowed two participants. Besides this, two IoT professionals were requested to participate in the research, however, they only elaborated on business models for IoT in general, and not specifically on their company’s business model. The companies operate in sectors of finance and telecommunications, and current businesses do not entail IoT products or services. An overview of the participating companies in their respective sector is illustrated in table 3. The number of representatives is also mentioned. Besides this, persons were contacted per e-mail, referrals from business network, and via networking. Furthermore, the final sample for this study has some variety in cases.

(30)

Table 3: Overview of participating companies.

Company: Sector: Number of

participants:

Contact per:

Sensefox Software-Hardware or IT 2 E-mail

Sodaq Software-Hardware or IT 1 E-mail

Ordina Software-Hardware or IT 1 E-mail

Koning & Hartman Consulting / Smart-Industry 2 E-mail

Louis & Ralph Consumer goods 1 Network

event

KPN Telecommunications 1 E-mail

Qorvo Software-Hardware or IT /

Healthcare

1 Referral

Schiphol Group Aviation 1 Referral

(Innovation manager) ABN Amro (Innovation manager) Banking / Finance 1 Referral Teleena (Board of member) Telecommunications 1 Referral 3.3.3. Ethics

The participants were requested to participate in an interview, and making it a voluntarily act. Furthermore, the interview protocol containing all questions was sent to all participants a couple days before the appointment. The idea behind this was to create a transparent overview of the interview questions. Confidentially and anonymity was agreed upon, in the sense that it was made clear that information would not be directed to participants or the organization.

(31)

3.4. Operationalization of constructs

In this section, the methods are explained on how concepts of the Business Model Canvas are measured. During interviews questions were posed about four main topics, which are identical to the four pillars the Business Model Canvas. Figure 2 presents an overview of nine building blocks that are structured into these four pillars. The interview questions were posed to gather information of a company on the following topics: the products and services, resources, network of partners, activities, type of relationships a company has with their customers, the

communication channels that are used, and financial aspects such as costs and revenue streams.

(32)

3.5. Data collection

In-depth interviews were conducted with a semi-structured interview guide, which is presented in Appendix 1. Semi-structured interviews were used to explore the perceptions of the

participants on business models for IoT, and without directing them explicitly towards an answer. For this reason, open-ended questions were applied in the interviews, in which respondents were allowed to disseminate their views according to what they consider to be important information, contributing to the research. In essence, the respondents had the chance to be experts in their field (Leech, 2002). The interview guide was based on interviews that were conducted by Dijkman et al., (2015). In the appendix of their study, an interview protocol is presented containing questions that were posed in their research. These questions were based on research done by Osterwalder and Pigneur (2010). Additionally, some questions were included. For this study, the interview protocol was used of Dijkman et al., (2015). Eight questions were added to this interview protocol, and 20 questions were deleted. Additional questions were considered important to answer the research questions for this study, while the deleted questions were considered irrelevant for this study.

Two pilot interviews were conducted, one with representatives of a company, and on with an IoT professional. Based on these interviews, adaptions were made for the interview guide in order to select and determine the best possible outcome of the questions, as suggested by Saunders and Lewis (2012). After the first pilot interview one question was deleted because it appeared to be irrelevant, and two questions were slightly changed because some respondents found them difficult to answer. Data was also obtained from companies’ representatives, and IoT professionals. The interview guide from the professionals was similar to the one used for the companies’ representatives. However, three questions were deleted containing specific

(33)

information about the IoT products or services of a company. The interviews were recorded and notes were taken. After each interview a summary was created, within approximately three hours. All the conducted interviews were transcribed.

Also, information was collected during the biggest IoT fair of Europe, in Utrecht. This contributed to understanding the technical side of IoT for the researcher, in terms of the usage of

Raspberry Pi, Arduino, Microsoft Azure. During the event an interesting remark was noted on

IoT, because the technology side of IoT was highlighted, whereas the business side, or business model topics were avoided.

3.5.1. Interview structure

The structure in conducting interviews and posing questions were identical for all 12 interviews. Most of the interviews were scheduled in a separated room in order that the interviews were not interrupted. 11 of the 12 interviews were conducted in Dutch, because that is the native language of the research and informants. While one interview was conducted in English. During the interviews the purpose of the study was once more explained, and this was followed by an introduction of the company. A semi-structured interview protocol was applied, in which the role of the researcher was only to pose questions and avoid certain interaction with the informants. Hence, researcher bias was explicitly avoided by means of not providing answers to questions, or participating in a discussion (Miles, Huberman & Saldaña, 2014). Furthermore, after posing a question the floor was given to the informants. Occasionally, informants went off topic, however, they could finish the sentences and questions were gently asked to bring them back on topic.

The style of questioning that is used is two folded. On the one hand, a grand tour questioning style is applied. On the other hand, example questions are used. The former asks

(34)

respondents to provide a verbal tour on the topic they know well (Spradley, 1979, as cited in Leech, 2002). Additionally, this style benefits the questioning best since it let respondents to talk in an open and fairly focused way (Leech, 2002). The latter is a more a specific questioning style (Spradley, 1979, as cited in Leech, 2002). This means that respondents were asked to provide examples, and to respond in their own words. The two styles complement each other on a more overview-level and in-depth level of questioning (Leech, 2002).

In conclusion, a consistency in collecting data was a priority, because data from multiple different companies and their representatives was obtained. Therefore it was important to engage participants from different sectors in a semi-structured way of interviewing. Hence, identical interview guides are used for the companies’ representatives and IoT professionals. For the IoT professionals a somewhat adjusted version of the interview guide was applied, which was a more general version on the topic commercializing IoT through business models.

3.6. Data analysis

During data collection phase, preliminary analysis was performed from field notes to structure the discussed themes and topics. These notes were used when data was analyzed. Also,

transcripts were used for analyzing data. Furthermore, the collected data from multiple cases were analyzed into two cycles of coding. Codes are labels with a symbolical meaning for words or short phrases that describe information collected during the interviews (Miles, Huberman & Saldaña, 2014). Both field notes and transcripts were used for the first cycle of coding. A second cycle of coding is used to search for patterns in data. Besides, a content analytic summary table is created, which is a matrix display that brings together all related data from multiple cases into a single form of exploratory analysis (Miles, Huberman & Saldaña, 2014).

(35)

The first cycle of coding consists of creating themes, and linking specifically the phrases of an interviewee to a theme. This type of in vivo coding method is adopted to place phrases in quotation marks (Miles, Huberman & Saldaña, 2014). In this regard, a theme is referred to as one of the nine building blocks from the Business Model Canvas (Osterwalder & Pigneur 2010). As a result, nine tables were created to illustrate direct citations of the informants. In addition, the quotations consist of answers to the posed questions during the interviews. Respondent’s quotations that are crucial or used for the results section are also illustrated in Appendix 2.

A second cycle of coding method is applied to search patterns in the data, and hence, in this cycle pattern coding is applied (Miles, Huberman & Saldaña, 2014). A theme is subdivided into categories. These categories consist of a symbolic word, and they were developed from the interviewee’s quotations. Therefore, no initial categories were used, however, some categories are identical to the categories created by Osterwalder and Pigneur (2010). This method lays groundwork for linking themes and categories from each case. Furthermore, some of the categories are mentioned more often than others, which does not mean that the categories indicated less frequent are less important. Additionally, since the research sample consists of various types of companies, it may be that a perspective of one company does not occur in another company. Categories that are counted more than once presented in table 4. A consequence of creating patterns and displaying the findings’ categories in a table is that

underlying phenomena are identified. All eight cases are randomly ordered, to ensure anonymity of the participants. Besides, after the first and second cycle of coding, another session of

searching for patterns was applied to see whether important information was missing, and all relevant information was derived.

(36)

4. Results

In this explorative study qualitative research was conducted, to obtain data from eight companies during 12 interviews. Data was analyzed and findings were structured according to the nine building blocks of the Business Model Canvas. Table 4 presents an overview of these patterns that are found in data. The first part of this chapter contains a short explanation of this overview. In the remainder of the chapter, findings of this study are discussed in more detail. They are divided in three sections, in line with the sub-questions of this study.

(37)

Table 4: Overview of findings depicted in categories. Value Proposi-tion Key Re-sources Key Partner-ships Key Activities Customer Segments Customer Relation-ships Channels Cost Structure Revenue Streams Case 1 Battery Data base-station Open source Measure-ments Sensors LoRa tech-nology Know-ledge Human Physical Collabo-ration Co-creation B2B Buyer/ supplier Niche market Early adopters Dedicated personal assistance Web sales Crowd-funding page Hardware Software Project Buyer / Supplier Billing method

Case 2 Data base-

station Measure-ments Sensors Battery Wireless LoRa tech-nology Know-ledge Time Money Develop-ment Physical Human Collabo-ration Co-creation Buyer / Supplier Network Platform B2B Buyer / Supplier Niche market Dedicated personal assistance Co- Develop-ment Network Direct Referral Hardware Invoice/ billing Subscrip-tion fee Pay per use Case 3 Optimize Combine things Predictive main-tenance Software Open source Network LoRa tech-nology Sensors Human IT Time Money Know-ledge Physical Collabo-ration Co-creation Identify needs B2B Service Dedicated personal assistance Network Research and Develop-ment Pay per use Pay per service / device usage, per sent amount data Flat fee Billing

(38)

Value Proposi-tion Key Resources Key Partner-ships Key Activities Customer Segments Customer Relation-ships Channels Cost Structure Revenue Streams Case 4 Software Services Sensors Human Know-ledge Use cases / business cases Collabo-ration Co-creation Mass market B2B Service B2C Partner company Service Service fees Subscrip-tion fees Buyer/ supplier Billing Case 5 Combine things (data sensors) Measure-ments (real time) Know-ledge Physical Human Time Money Collabo-ration Co-creation Platform Connect things Identify needs B2C B2B Mass market Auto-mated services Network Direct Research and Develop-ment costs Hardware Software Produc-tion Sharing collected data Case 6 Connect things Network LoRa tech-nology Data services Battery Know-ledge Hardware Software Physical Collabo-ration Collabo-ration Network B2B Dedicated relation-ship Own network Events Network R&D Variable costs Fixed costs Subscrip-tion fee Per data Per amount of messages Data services Billing Case 7 LoRa tech-nology Wireless Battery Physical Hardware Money Buyer / Supplier Collabo-ration App develop-ment Technical develop-ment B2C Mass market Auto-mated services Crowd-funding Web sales Partner stores Hardware Software Variable costs Fixed costs Buyer / Supplier Sharing collected data Case 8 Hardware Software Combine things Predictive main-tenance Optimize LoRa tech-nology Sensors Human Use case / Business case Collabo-ration Buyer / Supplier Collabo-ration Co-creation App develop-ment Technical develop-ment B2B Mass market Co-creation Buyer / Supplier Direct channels Own network Hardware Variable costs Fixed costs Pay per use Billing

(39)

Value Proposi-tion Key Resources Key Partner-ships Key Activities Customer Segments Customer Relation-ships Channels Cost Structure Revenue Streams IoT profes-sional 1 Software Measure-ments (real time) Sensors Combine things Human Sharing know-ledge Collabo-ration Connect things B2B Buyer / Supplier Niche market Early adopter Buyer / Supplier Service costs Buyer / Supplier Flat fee IoT profes-sional 2 Connected products Human Software Use case / Business case Sharing know-ledge Collabo-ration Mass market B2B B2C Direct channels Produc-tion Hardware Subscrip-tion fee Pay per use Sharing collected data

(40)

Research findings were assembled and categorized, table 4 shows the patters that were

discovered in respondents’ answers. The Business Model Canvas was used as a reference point for adopting the themes (nine building blocks), therefore, no other themes were developed. However, within these themes new categories emerged from respondent’s findings, since open questions were posed. At a first glance, some categories stand out because the respondents mention them most frequent. In the first pillar, products and services, two categories stand out, namely the use ofsensors and LoRa (Long Range) technology. In the second pillar,

infrastructure management, three categories are repeatedly emphasized, namely, human

resources, collaboration and co-creation. Within the third pillar, customer interface, one category is highlighted, which is the B2B market. The last pillar, financial aspects, reveals that costs mainly involve hardware, and revenue streams entail methods of: invoice or billing, and pay per use. Since the sample of this research study is rather small, the frequency of the categories by itself contains little value in terms of explaining findings. Much more meaning can be discerned when looking into the explanations of respondents. Therefore, findings are discussed in more detail in the following sections.

4.1. Findings related to value propositions

This section describes findings regarding the first pillar, value proposition. In particular, interview respondents were asked to disseminate on: (1) IoT products or services of the

company, (2) what customer problems and needs are addressed, and (3) the uniqueness of their company’s IoT products or services.

(41)

Results show that companies develop and offer a variety of IoT products and services. Companies operate predominantly in the business-to-business (B2B) markets, while the remaining companies operate in the business-to-consumer market (B2C). In general all

businesses create technological solutions for their customers, through connecting and combining sensor systems, data measurement and communication methods. These notions are represented in the following quote.

‘’The question is: what is the definition of IoT? In my opinion, it is about connecting and combining things. The Internet is a method of communication for IoT’’ – (Case 8).

The combination of the addressed technological solutions is referred to as an IoT product or service, which in these cases are used for: measuring ecological environments such as water or air quality, measurements in healthcare such as the well-being of elderly people, industrial measurements that predict maintenance of machines or devices, and traceability and location of a physical object. Almost all cases indicate to use similar technological applications, which are based on: software, hardware, network connectivity linked to the Internet, and tools for

managing, analyzing and measuring data. In this regard, one company stands out because they solely offer communication techniques within IoT. This company offers Internet connectivity for firms that apply IoT products or services, however they do not offer an IoT product or service themselves. This Internet connectivityis referred to as ‘LoRa technology’, and it is utilized by almost all of the cases. Alternative Internet connectivity’s are available, however, LoRa

technology is low cost and low in energy usage. Therefore companies mainly utilize it, which is also indicated in the quotations below.

(42)

’’The Internet connection is LoRa (Long Range), and to be more specific LoRaWAN (Long Range Wide Area Network). It is comparable with WiFi, which is 10 – 15 meters of distance. LoRa is 10km. It has a low energy usage, and a lower bandwidth. At this moment you use cables to send data, or you have to use a SIM-card. (…) LoRa is cheaper and wireless’’ – (Case 2).

Furthermore, all companies indicated that measurements are an essential part of IoT. In particular, measuring real time data is viewed as indispensable within IoT, as indicated by the two quotations below. Data is not merely measured ad-hoc, but more frequently, such as once or twice a day. This creates new dynamics regarding: insights into measurements of physical objects, customer needs, and value creation for customers.

‘’ A bank provides a loan to companies, which is based on estimations and forecasting, and through IoT that involve sensors, measurements are not based on estimations because companies have real time data. So, firms know what they have produced during a day and what amount of money is required to produce it’’ – (IoT Professional 1).

Real time measurements may even predict a certain impact, such as when machines require maintenance, and this means that a company can pro-actively act upon what type of maintenance have to be carried out.

‘’A smart system and software are used to obtain different data streams, and to create real time measurements. A certain impact can be predicted’’ – (Case 5).

(43)

Besides, findings also indicate that almost all firms tend to offer multiple technological applications to create an IoT product or service. Six of the eight companies offer products or services that comprise a combination of: software, hardware, network connectivity linked to the Internet, and tools for managing, analyzing and measuring data. These companies indicate that they manage either all of these applications for their customers, or the majority of these

applications. Nonetheless, cases six and seven indicate to only develop a single product or service solution. While one of these two companies creates an IoT consumer product, the other develops Internet connectivity specifically for IoT.

Above all, findings suggest that companies have different purposes for offering IoT products or services to their customers. Some cases suggested that the purpose is to obtain

insights into business cases. Other cases mentioned that in order to successfully apply IoT related services; there has to be a business case present. In other words, there has to be a problem

situation for which these IoT services offer a solution. Cases one, two, three, and six highlight that by applying IoT, they gather real time data which offers them profound insights into current business processes. As a result, these insights can predict maintenance of machines or other mechanical appliances, which can result into reducing costs. One IoT professional agrees with these notions. Thus, the usage of IoT in a business is to make processes more efficient, smarter and cheaper. Conversely, cases four, five, seven and eight indicated that a company needs to understand a customer’s problem, in order to offer IoT services as a solution to that problem. One IoT professional agrees with this notion, and adds that firms often fail in developing a successful business in IoT, because they do not have developed a business case. Thus, these findings show that the majority of the cases have similar technological applications to develop IoT products or services, however, they have different purposes for customers.

(44)

In conclusion, based on the above-mentioned findings, this section answers sub-question one. Value propositions within business models are related to the product or service offerings of a firm. Results indicate that in the majority of the cases products or services consists of a combination of various technological solutions and/or applications. However, even though companies made use of similar technological solutions and/or applications in their value

proposition, two different purposes were addressed. These involved; (1) measuring data to obtain insights into business processes, and (2) establishing a business case that offers solutions to customers’ problems accordingly. Furthermore, another important finding reveals that measuring real time data is fundamental when companies utilize IoT. These types of measurements offer real time insights, which can enable companies to take accurate decisions or even predictions regarding business processes. In the succeeding section perceptions regarding crucial elements in business models for capturing value from IoT are discussed.

4.2. Findings related crucial elements of business models

In this section, findings highlight fundamental elements within business models to create value from IoT. Respondents elaborated on the topics of the four pillars of the Business Model Canvas. Ten respondents reported on questions concerning their own business model, and two IoT

professionals elaborated on elements of business models for IoT in general.

In general, findings show that: (1) businesses tend to operate in the B2B market segment, (2) customers are often perceived as partnering companies because they collaborate and/or co-create on the process of developing IoT products or services, (3) several resources are used from partnering companies, and (4) a variety in cost structures and revenue streams are utilized.

(45)

All of the cases indicated that collaboration is essential to firms for several reasons. For instance, more than half of the cases mentioned that collaboration is vital for combining the technological and business domains. This is illustrated below:

‘’We build partially the technical side, and someone else manages the market’’ – (Case 2).

Another reason for collaborating with partner firms is because firms do not own specific techniques for developing IoT products or services. For example, they do not possess resources on software or hardware. Hence, these resources require a company to outsource, so that an IoT product or service can be developed. An example is presented from case seven.

‘’We buy hardware from a company. (…) For software we choose a company that can supply our demands’’ – (Case 7).

Fundamental to collaboration is co-creation. Six cases mentioned that companies tend to

combine technological and business resources with their partnering firms. In fact, this means that specific knowledge is used from at least one other company for developing an IoT product or service. An example of co-creation is presented below:

’’We have technological knowledge, and the other firms have professional knowledge on software’’ – (Case 1).

Furthermore, findings show that various resources are perceived as essential to develop IoT products or services. All cases, except for cases six and seven, show that human resources are fundamental to firms. Respondents indicated that knowledge on technology is required to develop products or services, and to generate a market. Therefore, human resources are required to determine boundaries and possibilities of developing IoT products or services. An ideal combination of resources is knowledge on the technological and business domains. Thereby

How to commercialize IoT? : exploring the potential of the Internet of Things to create value through business