Business Models: Energy Optimization.
Author:
Gert-Jan Beeke (S2689847)
Supervisor: Dr. David Langley
Co-assessor: Dr. John Dong
Abstract
To better understand how IoT technologies embedded in goods enable innovation in service systems, this study investigates how to design a service-dominant business model which leverages the affordances of the IoT. Using a formative design science approach, we established an IoT S-D business model by which optimal energy efficiency of heating systems makes economic sense, also to the business party. We established that IoT affords detailed coordination and data-based actions, which in turn affords a pattern of mutual value creation between business and consumers. However, evaluation indicates that successful implementation of our business model is highly contingent on the institutional logics dominating in both the industry and marketplace. For future research, we propose a model that uses IoT S-D offerings to leverage contradictions in institutional goods-centered logics and legitimize service-oriented alternatives.
June 20, 2016
Total word count: 19.775
MSc Business Administration, Change Management
University of Groningen,
INTRODUCTION
Today’s world is characterized with an ever increasing digitalization of previously non-digital things such as cars, thermostats, light bulbs, pet-collars, watches and so forth, which through a network of sensors and actuators interact with their environment creating digital ‘smart’ platforms such as smart buildings, smart cities and smart energy grids (Wortmann and Flüchter, 2015; Yoo, 2010). This pervasive digitalization phenomenon is known by many as the Internet of Things (IoT) and may have disruptive implications about how business is conducted (Barrett et al., 2015). This paper seeks out how to create value based on the potential uses of IoT and the bulk of sensory data it produces. To investigate new business model logic, this paper investigates a specific model which allows businesses to increase their profits and improve environmentally sustainability by a continuous improvement of their goods energy-conversion efficiency.
In effect of the big data ‘revolution’ made possible by the IoT, Barrett et al. (2015) argue that it is relevant to conceptualize the service provided by applied resources as the common denominator of all economic exchange, rather than a traditional goods paradigm based on simply selling higher quantities of goods and products. In this service-dominant (S-D) logic, a ‘thing’ is seen as merely a mechanism that provides a service (Vargo and Lusch, 2004, 2008). Thus, a service is not the ownership of a digitized lawnmower, but the prospect of having a well-mowed lawn. The main premise of S-D logic is to compete on all available operant resources such as skills, information and knowledge throughout the larger ecosystem in which goods are applied (Vargo and Lusch, 2004, 2008). The S-D logic is highly appropriate for application to the IoT, which produces an extreme amount of sensory data that allows for real-time and data-driven improvements of the service. An example of the IoT and S-D logic combined is provided by Rolls Royce’s aircraft engines, which are embedded with sensors for real-time data analytics in order to continuously improve on performance, safety and maintenance costs (Barrett et al., 2015; Neely, 2008). In effect, rather than following a conventional business model that competes on price per sold product, Rolls Royce now leases their engines by selling ´power by the hour´ and is, therefore, able to compete on value captured by sensory data that helps improve fuel consumption efficiency, and prevent major downtime by monitoring maintenance issues (Neely, 2008).
A similar service-oriented value proposition might also be appropriate in other domains. However, the change in Rolls Royce’s organizational structure, tasks, and roles of actors (including those of clients) indicates a disruptive reconfiguration of rules and norms within an industry. Thus, in order to truly leverage new IoT technologies, conventional goods-centered businesses may have to radically change their socio-technical system (Lyyntinen and Newman, 2008; Seidel et al., 2013). Yet most conventional businesses have limited experience, knowledge and resources on how to appropriate and leverage the IoT with their operational systems centered on their traditionally non-digitized goods (Witchalls, 2013). Management literature provides few pragmatic answers on how a business should approach the IoT in terms of governance and management in order to create value for the business (e.g. Baradwaj et al., 2013; Barrett et al., 2015; Yoo et al., 2010). Also, research is primarily concerned with issues and developments of the technical IoT aspects, whereas the potential disruptive implications for organizations remain underdeveloped in research (Yoo, 2010).
Accordingly, this paper addresses the gap argued in Barrett et al. (2015, p. 139) about “how might digital technology embedded in products enable innovation in service systems?”. By following those authors’ proposition of information as the key component of value creation in today’s digitized context, this paper’s main research question is: How to design a service-dominant business model which leverages the affordances of the IoT?
To do so, this paper adopts a design science approach to develop design theories for new IoT business models. Specifically, we focus on one potential area of disruption through IoT technology: energy conversion and optimization. In a goods-dominant market, a firm benefits from being able to advertise energy efficiency in relation to its products, but the actual performance is less important, particularly as it is rarely measured. Recent controversy surrounding the efficiency claims of some car manufacturers, including Mitsubishi (Soble, 2016), highlights this inconsistency. In an IoT S-D offering, the actual, real-time energy conversion of goods, such as central heating boilers, can be monitored, reported and adjusted, opening up wholly new commercial propositions.
presents design theories (i.e., if you want to achieve Z in situation X, then do Y) complemented thick descriptions (i.e., how to interpret and apply the design theory) in the form of an IoT S-D business model demonstration. This model is both formed and tested through evaluation with experts. This formative design approach allows iterative refinement of the business model, reflection upon its underlying design theory, and establishing questions for future research.
To management and practitioners this research contributes by providing prescriptive-focused knowledge of a heuristic nature (Van Aken, 2004, 2005) that allows effective appropriation of the IoT in other domains. Witchall’s (2013) international survey amongst 779 senior business leaders illustrates the top barriers to increase the use of the IoT: lack of skilled and knowledgeable employees, a lack of senior management commitment, and knowledge and products or services (are perceived as) having no obvious IoT elements. Hereby businesses run a considerable risk of falling behind competition, considering the rapid speed of digitalization, its disruptive potential in a broad range of industries, and the learning curve typically required for human agents in order to use new technologies effectively (Strong and Volkoff, 2010; Witchall, 2013). This research also contributes to IS literature concerned with digitized businesses models (Bharadwaj, 2010; Yoo et al., 2010), the development of IS design (Carlsson et al., 2010), the effects of digitalization in social life (Yoo, 2010), the development of green IS systems (Melville, 2010; Seidel et al. 2013; Watson et al. 2010), and service innovation in IS (Barrett et al., 2015).
LITERATURE REVIEW
Thus far there is no agreed upon definition which summarizes what the IoT encompasses as its field of application is as numerous as diverse, and the corresponding information and communication technologies are rapidly developing (Wortmann and Flüchter, 2015). The IoT is often referred to as a grand emerging phenomenon (Yoo, 2010) or a vision (Atzori et al., 2010; Gubbi et al., 2013). Therefore before describing the main design theory based on S-D logic, this paper first discusses background information on what the IoT constitutes and the theoretical lens of affordances that is used to conceptualize the phenomenon.
IoT constituents
To specify the IoT in piecemeal components, Porter and Heppelman (2014) describe the IoT as a phenomenon which consists of products with physical, smart, and connectivity components as its key constituents. Smart components amplify the capabilities of the traditional physical components (e.g. the engine block, tires and batteries of a car), while connectivity amplifies the value of smart components and enables it to exist outside the physical product itself (Porter and Heppelman, 2014).
Taken from another angle, the IoT can be seen to emerge from the combination of three perspectives, being the things, the internet and semantic technology (Atzori et al., 2010). By the narrow things perspective, the IoT involves physical objects being embedded with technology which thus enables things with computing capabilities and internet connectivity. Effectively, formerly non-digitized objects become uniquely addressable, traceable and remotely controllable (Atzori et al., 2010). The thing itself and its external environment become increasingly embedded with interconnected sensors which may process all sorts of data back and forth. Accordingly, the internet perspective focuses more broadly on the network of interconnected things and its infrastructure such as radio towers, servers and databases and standardization of communication protocols (Atzori et al., 2010). This perspective emphasizes that the IoT as an open information system consisting of loosely coupled, heterogeneous, and interconnected elements (i.e., things and sensors), rather than merely a single and closed device with additional computer capabilities.
In effect, what differentiates an IoT environment from a non-digitized environment is that tasks become data or information intensive. Through semantic technologies and algorithms, raw sensory data can be processed into actionable information. Yet, in order to reach its full potential, it often requires additional human knowledge of the domain and its specific context (Barnaghi et al., 2012; Rowley, 2007). Thus, in order to leverage the big data generated by the network of interconnected things, smart semantic middleware technologies offer beneficial algorithms and convenient solutions (Atzori et al., 2010), but it also requires developers to consider the social realities in which the thing is deployed and the cognitive appraisal and interpretation of the users. In the next section, this paper uses the theory of affordances to further conceptualize this interplay between IoT technology and socially constructed realities (Markus and Silver, 2008; Zammuto et al., 2007).
Theory of affordances
To capture the interplay between the social reality and technology, the theory of affordances offers a relevant and useful perspective (Markus and Silver, 2008; Zammuto et al., 2007). Gibson (1979) developed the idea of affordances to explain differences between species on how an object is perceived in terms of the possibilities the object affords for action. To illustrate the general concept, one can consider a hammer. Most people will see a material object shaped in a particular form and which affords them the ability to hit a nail into a wall, while a dog might only see a toy to chew on. If one is working outside in the wind, the hammer may be recognized by some individuals as an effective paper weight. Realization of affordances thus depends on the material features and individual’s cognitive capabilities, but also differs through time and space (Yoo, 2010).
In the IT context, Markus and Silver (2008) argue, “affordances are a type of relationship between the object and a specified user (or user group) that identifies what the user may be able to do with the object, given their capabilities and goals”. More formally, they define affordances “as the possibilities for goal-oriented action afforded to specified user groups by technical objects”. It is important to note affordances may arise from physical hardware and non-physical software, because what is important about the object’s materiality not only the matter but also the form in which it is shaped that endures across time and space (Leonardi, 2012).
Leonardi (2011, 2012) refers to affordances as the coming together of the material agency (i.e., ways in which a technology acts that endures over time and space, and which usually needs to be provoked by a human) and human agency (i.e., coordinated human intentionality, often in response to the material agency) in sociomaterial practice. Following Leonardi (2012), sociomaterial practice are shaped and reshaped by a social sub-system with socially constructed institutional logics such as rules, norms, values, culture and hierarchy. Thus, in effect of affordances enacted, there is a recursive shaping between the human agent (e.g. their intentions) and materiality vis-à-vis institutional logics (Leonardi, 2012).
Hereby, affordances offer a conceptual middle ground between material determinism and radical human constructivism (DeSanctic and Poole, 1994). Recognizing both arguments, the critical realist view argues technology has materiality which exists regardless of human agency and therefore causal potential does exist, but whether or not this potential is realized is moderated by other conditions including human choice (Markus and Silver, 2008). In other words, affordances create potential but not both necessary and sufficient conditions for goal-oriented actions (Markus and Silver, 2008; Seidel et al., 2013).
Accordingly, to construct business models that leverage the affordances of IoT one also needs to consider the social and cultural context in which the technology is implemented. Through a case study, Seidel et al. (2013) present a process model which backtracks how transformational affordances of organizational sense-making and sustainable work practices emerged and became realized. Using theories of affordances, they argue an overarching theme (in their case: environmental sustainability) must be incorporated consistently throughout all parts of the sociotechnical system in order to achieve such transformational goals. This illustrates that despite their emergent nature, affordances can be guided by clever design decisions of both the material and social environment.
can be converted into actionable information. In a similar line, Porter and Heppelman (2014) argue that the IoT objects will generally develop in four areas, where each builds on the preceding one: First, the IoT allows for monitoring changes internal or external to the object and may notify actors accordingly. Next, the control over the objects functions will allow for personalization of the users experience. Third, the monitoring and control allow for algorithms that optimize the value-in-use, and lastly, by a combination of all three, the object operates autonomously which allows for self-coordination with other systems, automatic enhancements and self-diagnosis.
But what does this afford us in daily life? Burton-Jones and Grange (2013) illustrate an overarching high level affordance of information systems that is to use data to represent a relevant part of reality, which can be utilized by agents to execute informed actions. Following the IoT semantics perspective, if a sensor monitors a change in condition x, it may send the information to the user, a machine or another relevant agent by which they can take a timely and informed decision for action. By doing so, activities can be done more effectively than without these representations of reality present. Hence, following this understanding, the ability to make informed actions based on representations of reality is what makes a now-digitized object ‘smart’.
In a similar line, Zammuto et al. (2007) argue modern information technologies affords to coordinate organizations based upon information instead of upon hierarchy, which traditionally has been the basis for structuring organizations. In effect of improved information coordination capabilities, they illustrate five pragmatic affordances of IS, being visualization of entire work processes, real-time creation of new modules of services or products, mass and virtual collaboration, and to simulate what-if scenarios. In an IoT context, the potential to create a ubiquitous network of connected things, people, buildings, cities, businesses and so forth, affords informed coordinative capabilities at an unprecedented level of detail. To summarize, the smart and connected capabilities of IoT-goods may thus afford a digitized world in which data directly coordinates the world around us based on well-informed and data driven decisions. Effectively, and to provoke thoughts, IoT then metaphorically acts as the lens that directly shapes and mediates our experiences; much like a character experiences the world in a role-playing game (Yoo, 2010).
Service-dominant logic
In line with the above discussion about data richness, informed actions and information coordination as overarching IoT affordances, this paper follows Barrett et al.’s (2015) proposition that information becomes the primary component in order to create value through the IoT. In Barrett et al. (2015), the idea of information as the main driver of new potential uses is linked to the S-D logic which – as discussed in the introduction - conceptualizes service as the fundamental basis of economic exchange (Vargo and Lusch, 2004, 2008).
In addition to exploiting the affordances of a system, S-D logics may also positively complement the symbolic expression of an IoT system (i.e., what the spirit of the system represent to a specific user group; Markus and Silver, 2008). By monitoring, storing and inferring data which reflects reality, the IoT will have a pervasive impact on privacy (Weinberg et al., 2014). Also, the data safety issues on collection, errors, secondary use and unauthorized access drive scepticism (Milberg and Burke, 1996). How people elaborate on whether to use a system may differ amongst groups and individuals due to differences in attitudes and intentions (Angst and Agarwal, 2009), culture (Bellman et al., 2004), and the trade-off between personalization and increased experience in exchange for more information (Chellappa and Sin, 2005). In general, people become increasingly sceptic as a technology becomes ubiquitously present, but with the right antecedents people also are more likely to develop a pragmatic stance towards the technology, i.e., they examine the benefits and privacy risks and decide to trust the organization or seek legal oversight (Harris Interactive and Westin, 2002; Bala and Venkatesh, 2016). Thus, if some stakeholders feel disadvantaged (e.g. loss of control, perceived threat, no true benefit) by the IoT system, they are likely to resist. While not solving all issues related to privacy and safety (which is outside this papers’ scope), the S-D logic does present an economic paradigm which incentivizes ethical behaviour between stakeholders involved: a win-win solution. It also underscores that businesses cannot simply add potentially privacy infringing sensors in their products without offering the potential for truly increased value-in-use in return.
Constructing IoT business model logic
Synergizing previous literature offers a starting point for IoT-business model design theory. Based upon their value proposition, a business applies their intangible operant resources to transform tangible operand resources (e.g. raw material) into ‘final’ goods, which then serves as a mechanism that affords the user value-in-use. In turn, the IoT allows the user to produce information-through-usage (e.g. based on real-time sensory data feedback) that generally moves upstream from the value-in-use back towards businesses (figure 1). Hereby, the IoT affords the business the ability to remain involved with their goods
post sales by proactively monitoring, and potentially improving and personalizing some critical
performance indicators throughout the goods lifespan. In response to the sensory data, the business may apply its knowledge, skills and other operant resources to improve on performance, prevent future maintenance costs, or even inform the user how to use the goods more effectively. Accordingly, the IoT offers a solution to create a recursive loop of operant resources such as skills, information and knowledge in the exchange between a business and its customer base (i.e., the user), which allows a mutually beneficial improvement of both the value proposition and the value-in-use, or, put differently, an ongoing improvement of the service. To summarize the previous as a rule for the construction of a business model, this paper formulates the following design theory.
Design theory: In order for a business in an IoT setting to achieve value creation, they (a) apply the S-D
logic to formulate specific value propositions based on the goods they offer, and (b) coherently leverage the sensory data originating from value-in-use for improvement of their value proposition and in consequence value-in-use.
Figure 1. How value proposition and value-in-use become mutually
beneficial in an IoT context with an S-D relationship
An assumption made in the design theory above is that the business may have more and better understanding of their goods than their customer base (i.e., the business has better operant resources available). This may certainly be true for more specialized and technical aspects of goods such as the operational functioning of a car engine or central heating system.
Accordingly, businesses propositions commonly include maintenance- or warranty-agreements for their goods, which essentially shifts responsibility to the business to guarantee their goods do not malfunction and deliver service over an agreed period of time. By using IoT-infrastructure and following the design theory, this may be realized at relatively low costs and high accuracy.
Similarly, but more radical, is to shift responsibility for other relevant performance indicators from the user to the business. A business’ value proposition may include relevant performance specifications such as the production of emissions, noise, light or heat or other quantifiable effects of using the good. Depending on its relevance and context, users may value a performance guarantee or future improvements on a specific performance indicator throughout the actual usage period.
One generally important performance indicator of goods – which this paper will focus on - is its energy efficiency (i.e. how efficient fuels such as gas, water or electricity are converted in a desired output, such as heat). Amongst many other appliances, this may be highly relevant for selecting a central heating boiler which is commonly used for 10 to 20 years and significantly impacts utility expenses and the environment. The IoT may monitor energy efficiency, and the business may act accordingly based on our design theory. Goods may be advertised within a scale of energy efficiency (e.g., through mandatory energy labels). In accordance, a business proposes a contract that their central heating boiler requires a set amount of fuel over a fixed amount of time or use-hours, in order to heat a volume of water for building heating. Users may value the certainty that their fuel costs will be optimized, and may even decrease over time. Alternatively, utility costs can be included within a pay-per-use or power-by-the-hour business model. In this case, the business takes over responsibility for total cost of ownership (i.e., material costs and operational costs) while the user pays a fee in exchange for usage, regardless of utility costs. Hereby, an energy efficiency improvement directly increases the business’ profit margins, thus providing them extra stimulus for ongoing efficiency innovation.
This presents a first direction to utilize the IoT and create commercial value propositions for energy efficiency optimization; however feasibility of this direction and the underlying design theory remains unexplored. The remainder of this research sets out to investigate the feasibility for business models based on improving the energy efficiency of a buildings central heating system (including the boiler). By doing so, it also validates the underlying logic of our design theory.
METHODOLOGY
This paper follows a design science (DS) method in order to contribute a design theory and propositions on how businesses can create value using the IoT. DS is a suitable method when considering the novelty of the IoT technological phenomenon, the speed and diversity in its development, and overall its disruptive potential for change. DS is particularly useful when extending knowledge on fuzzy, ambiguous and complex issues where foresight is required (Denyer et al., 2008; Romme, 2003). Accordingly, DS is advocated by multiple IS researchers to contribute the understanding of IoT use by businesses (Baradwaj et al., 2013; Barrett et al., 2015; Yoo, 2010; Yoo et al., 2010).
Theoretical positioning
Van Aken (2004, 2005) argues the core of DS is to develop knowledge that can be used by a well-defined group of professionals (i.e. individuals who have received formal education in their field such as doctors, architects, engineers, lawyers) who solve real-world field problems using their skills, creativity and scientific design knowledge. Hereby, DS complements descriptive-focused knowledge that explains the empirical truth (e.g. often in hindsight with a quantitative law-like model) with prescriptive-focused knowledge that intents to develop and test alternative solutions for a problem (Van Aken, 2004). This does not mean that DS produces narrow scientific knowledge for a specific managerial problem, but knowledge of a heuristic nature for a class of managerial problems which practitioners can apply in their specific domain (Van Aken, 2004). Thereupon, DS fits this paper’s goal of developing knowledge for management on how to utilize the IoT effectively.
(e.g. Hevner et al., 2004; Gregor and Jones, 2007). However, in a DS research which views service as the fundamental basis for exchange (i.e., S-D logic), the distinction between goods and services are of lesser importance (Kimbell, 2011). As argued before, the critical realists view – which this paper follows - finds a middle ground between hardline determinism and social relativism by acknowledging that technology produces the potential for causal relations, but that whether or not this potential is realized is moderated by other factors in social reality (e.g. DeSanctis and Poole, 1994; Markus and Silver, 2008; Volkoff et al., 2007). Accordingly, and despite the centrality of the IoT-technology of this paper, the end product research is not an IT artifact but, more appropriately, a set of interventions in the (sociotechnical) system (Carlsson et al., 2010; Denyer et al., 2008; Kimbell, 2011), which in this paper takes the shape of a design theories complemented with business model demonstrations.
Specifically, this paper follows the sociotechnical DS method of Carlsson et al. (2010), who propose a method for developing IS design knowledge with methodological suggestions from both IS design science (e.g. Hevner et al., 2004; Gregor and Jones, 2007) and management design science (e.g. Denyer et al., 2008; Romme, 2003; Van Aken, 2005). The approach of Carlsson et al. (2010) aims for “practical design knowledge and theory that can be applied by individuals to design and implement IS initiatives”. These interventions come in the form of design theories in a sociotechnical system where IT artifacts – such as IoT objects - are critical – but not exclusive - means for achieving the desired outcome of the intervention (Carlsson et al. 2010). A design theory is expressed as “in problem situation P and context C, to achieve outcome O, to act according to design proposition(s) D (adapted from Bunge 1967; Carlsson et al., 2010)”. Design activities and data collection
This paper follows Carlsson et al.’s (2010) four generic key activities to conduct DS research. While these imply linearity, in reality DS involves creative and iterative thinking in which the researcher frequently moves back and forth between activities for fine-tuning. The specific path we followed is briefly discussed below. To help structure our report, we use the publication schema of Gregor and Hevner (2013).
Unlike most DS research, we started with a generic solution (i.e., the IoT) but without a specific predefined field-problem. To make an analogy with penicillin: there was first the discovery of the ‘solution’ in 1928, but it took another 10 years before its potential as an antibiotic for a vista of diseases was properly understood. Similarly, we know there is an interesting new technological phenomenon emerging, but our understanding how to use it for solving real world problems and creating business models logic remains less developed.
Accordingly, we started with Carlsson et al. (2010) activity of reviewing extant theories, knowledge and data in order to ground the research in previous studies and knowledge. This ultimately resulted in our design theory at the end of our literature review that links S-D logic to the IoT, and which in this paper serves as the underlying fundament for an IoT business model construction.
Having established a specific field-problem and design theory, our third activity is to extend and refine the design-theory and knowledge by designing a business model artifact with three value propositions alternatives. This results in the ‘business model demonstration’ chapter below, which also functions a ‘thick description’ to further aid readers’ understanding of applying our design theory in practice (Carlsson et al., 2010). In order to construct consistent business models based on our design theory, we made use of the business model canvas (table 1) of Osterwalder, Pignuer and Tucci (2005). We started from the value propositions and then moved forward by considering what these would imply for each specific box. By considering the different and interconnected components of the canvas, it allows to iteratively develop increasingly consistent and detailed business model logic which demonstrates the functioning of our design theory. We note this activity happened in sequence with the fourth activity.
Key partners Key activities Value proposition Customer Relationship Customer Segments Key resources Channels
Cost structure Revenue stream
Table 1. Business model canvas (Osterwalder et al., 2005).
The fourth activity is to test the applicability of the design theory and knowledge using a normative approach, which is to take interviews with relevant experts in order to find a proof of concept (Carlsson et al., 2010; Gregor and Hevner, 2013). We took a total of seven interviews with eight people from a variety of backgrounds (see table 3 in evaluation chapter). These interviews are part of a formative evaluation, which is an iterative approach during our business model development with the goal of detecting and eliminating usability problems (Theofanos and Quesenbery, 2005). It is a less-developed approach contrasting summative evaluation where a relatively complete design is tested against quantitative criteria (Theofanos and Quesenbery, 2005). Gregor and Hevner (2013) also argue that with formative testing, the final evaluation cannot necessarily become as full and in-depth as in behavioral research projects where the artifact was already developed by someone else. For instance, we experienced that due to the iterative development and the variety in interviewee’s backgrounds, it was less ideal to use a standardized questionnaire. Instead, by asking tailored questions we argue it allowed interviewees to make more relevant contributions from their own professional experiences. We justify the approach as it allows co-creation of the business model artifact from a rich range of perspectives, and therefore it benefits our goal of developing relevant design knowledge.
Our interview-approach is to pitch and discuss the energy efficiency proposals with a variety of IoT practitioners, business developers and other relevant experts in semi-open interviews. A pitch was prepared, but as noted above, the interviews were kept open to allow participants share their own professional experiences. Results have been translated from Dutch to English, transcribed and coded for evaluation purposes (Appendix D). For deductive codes, we use the four components of the sociotechnical equilibrium model of Lyyntinen and Newman (2008) (i.e., technology, structure, task and agents). This allows reflection on our designs feasibility in terms of technology and whether these are realistic in social practice. Hereafter we can discuss how these social and technical components influence each other. The results of evaluation allow incremental improvements of the initial design, and to discuss the implications of the design theory on a more abstract level along with suggestions for future research.
BUSINESS MODEL DEMONSTRATION
Value propositions
As a major opportunity for energy optimization, we found variety of sources suggesting boilers are commonly installed with factory default setting and without embedding it in the wider central heating system, resulting in less than optimal energy efficiency. While this is most time and cost efficient for the installation party, it increases the total cost of ownership. For instance, while a boiler is fit highest efficiency, the radiators water levels might be poorly balanced resulting in rooms overheating or not heating at all. Additional maintenance costs, additional heaters installed or inefficient energy usage. Another example is improper installment and configured of a boiler, resulting in unnecessary energy deficiencies when converting gas into heat. Appendix A presents technical details of the specific optimization opportunities we identified.
Conventional business model offer no incentive to optimize for energy efficiency of heating systems, while the effects are less known to users in general. Opportunities for personalization and optimization of commonly installed boilers, along with the hydraulic balancing of the radiators, have been reported (e.g. Ruch et al., 2014; Dentz et al., 2013, Appendix C) which offer potentially significant reduction in fuel consumption (e.g. gas and electricity) while extending the boilers’ lifespan and reducing maintenance costs to the boiler and system. Those innovations are also beneficial to the climate-comfort, which is linked to healthier environments and productivity improvements of people in the room. Professional optimization is also argued to be safer, as erroneous interventions may cause hazardous emissions and legionella bacteria.
Thus, here lies the opportunity for business models to improve the energy efficiency of currently installed boiler and heating systems. When proactively monitoring and comparing the quantity of fuel input with the output of heat produced by the heating system (i.e. the mechanism that convert input into output), there lies the opportunity to continuously improve on this performance metric (figure 2). Collecting data is thus required for fuel input and heat output, and optional for relevant aspects of mechanisms and value-in-use. Hereby we note that if replacement of mechanisms such as a boiler or installing additional materials such as automatic valves, heat buffers, solar panels, or heat pumps makes economic sense by reducing total cost of ownership while improving output of heat and consequently the quality of value-in-use, it may push both the optimization business party and users to consider these options.
Figure 2. Business model basic functioning.
Based on the business opportunity above, we developed three value propositions alternatives that enable an economic stimulus for business parties to pursue energy optimization:
therefore offers an incremental change to the status quo. In itself, it does not offer a true economic incentive for the businesses to go beyond agreed efficiency levels, unless the contract is re-evaluated periodically. Accordingly, it may be combined with value proposition 2.
2. Data brokerage and improvement offerings: Collected energy efficiency data is distributed to relevant energy optimization parties (i.e., by the user themselves or with their informed consent), which allow these optimization parties to make reliable, data-based offerings to the user for additional energy savings. This proposition may be coupled with the efficiency contract guaranteeing the user a quantifiable improvement with predictable return on investment. This still requires proactive behavior of the user in order to initiate innovations; therefore we also consider a third proposition alternative below.
3. Heat-as-a-service: We consider a complete package deal as a third proposition: The business party takes full responsibility for the total cost of ownership (i.e., the cost of material, fuel, maintenance and all other operational costs), while the user only pays for the heat (i.e., the output in Gigajoule) that they demand and use. It only requires the user to specify how much heat they want and how fast they want it.
To explicate, the main difference is that proposition 1 and 2 still allow users to keep (partial) ownership over the materials and therefore more autonomy, control and responsibility. Changes can be proposed by any party but are initiated by the user after re-evaluating the contract. These propositions might be more appealing to customers valuing some degree of self-determination. Conversely, the third proposition puts all material ownership to the business by default. This is appealing to customers that only care about output, and wants to minimize their part in how this is realized. Accordingly, any efficiency changes are directly reflected in the business’ profit margins.
Business model canvas.
We continue to clarify the functioning of these value propositions by using components Osterwalder et al. (2005) business model canvas. An overview is presented at in table 2, which also includes heuristic rules for constructing similar IoT S-D business models. The demonstration is evaluated in the next chapter.
Key partners - User is a key partner. Make or buy: - Material manufacturing - Installation - Maintenance - Fuel supply Key activities Key performance indicator: Energy efficiency (conversion of fuel to heat) Value propositions Incremental path 1: Efficiency contract 2: Data brokerage and improvement offerings Integrated service 3: Heat-as-a-service Customer relationship - S-D logic. - Value co-creation - Long term orientation. Customer segments Those that lack sufficient coordinative
capabilities and other resources for energy optimization.
E.g. utility buildings, multifamily homes, private home owners. Key resources
Data: Input (e.g. m3 gas)
and output (Gigajoule).
Operant: Optimization
knowledge and skills.
Operand: Fuel, heat
producing goods. Channels - IoT technology. - Human mechanics and engineers. Cost structure
- Re-evaluate cost allocation of total cost of ownership between user and businesses.
- If the user is an (intermediate) fuel supplier, they are paid accordingly in a heat-as-a-service offering.
Revenue streams
- Potentially disrupts business models of partners. - Premium pricing options for convenience and safety. - (Periodic) fee for energy efficiency contract.
- One-time charge per improvement (value prop. #2) - Pay per gigajoule (Heat-as-a-service);
Table 2. Business model canvas for IoT S-D offerings, specified on energy efficiency.
Key activities. At the core of all three value propositions is a party that specializes in energy optimization
measuring the systems production of thermal energy (i.e. heat)1. From here on, parties may seek to
personalize and improve efficiency through their preferred method of choice. It is up to the specialization party to decide how to do this most efficiently. For instance, possible options are instructing and educating a user, sending out a plumber for physical inspection, installing additional technologies such as automated mixing valves, or, ultimately, through smart automated algorithms. We note this train of thought follows the consecutive four areas argued by Porter and Heppelman (2014) of monitoring, personalization, optimizing, and algorithmic automation.
We also note that the party specializing in energy optimization may also have other (complementary) key activities. For instance, in an efficiency contract offering, it is a logic match for plumbers to include an energy optimization with a maintenance contract. In a heat-as-a-service offering, a party may be deeply involved with most value-chain activities (e.g. manufacturing, installment, maintenance, and fuel-supply) that together form the total cost of ownership.
Customer segments. Our propositions may be relevant to a broad range of building owners that use
conventional heating systems, but lack the coordinative capabilities and resources such as time and knowledge to improve energy efficiency of their heating system. Accordingly, we argue to target a broad segment which may include utility buildings, multifamily homes and private home owners.
Revenue stream. We perceive cost savings on fuel as a direct potential source of revenue for our value
propositions. Our value propositions directly compete for revenue with fuel companies that use a business model based on selling large quantities of fuel. This disruptive change may also influence tax incomes. It also negatively affects business models that rely heavily on malfunctions in order to sell repair-hours and additional components. Conversely, these businesses may profit if they change their model to proactively improving energy efficiency. We found that explicating whose business model is disadvantaged helps predict oppositional forces from (former) industry partners. We return to this topic in the evaluation. To indicate the potential of our proposition in terms of gas savings: in the Netherlands, the average household uses 1440 m3 of low caloric natural gas a year, which in 2016 costs €925 (Milieucentraal, 2016). In utility buildings and multifamily homes, the energy usage is even higher per square meter and commonly less efficient due to the increasing complexity of the system of heat supply and demand (e.g. consider a building with a hundred rooms with different heat demands), and thus the cost of coordinating heat distribution efficiently. Although their scope is not restricted to the central heating system, Energy Service Companies (ESCo’s) commonly report 30% to 40% reduction in energy costs in utility buildings and multifamily homes (ESCo netwerk, 2016).
Additionally, the three propositions offer longer term benefits which may justify a premium price. These additional benefits include convenience, less maintenance costs, longer lifespan of boiler, higher comfort, healthier environment, safety improvements, increased productivity, and decreased emissions.
The specific revenue stream differs per value proposition:
1. The efficiency contract proposition most likely receives a fixed (periodical) fee and does not directly benefit from any additional efficiency improvements, unless the contract is updated periodically. Alternatively, it can be combined with the improvement offerings proposition below. 2. In case of improvement offerings, the user pays for each individual energy-saving offering they
accept. Based on the IoT data, the investment for additional goods or maintenance should have a transparent and predictable payback time, meaning that personalized and more reliable offerings can be made. The data increases bargaining power of the user, whom can more accurately consider their return on investment. Accordingly, optimization parties may consider a no cure-no pay proposition, i.e., if the efficiency improvement is not realized there can be refund of costs. 3. In a heat-as-a-service offering, revenue is generated through each quantity of ‘heat’ delivered to
the user. In this offering, every saving on fuel, material and maintenance that is used as a mechanism to produce a Gigajoule of heat has a direct effect on the business’ profit margin.
1 Gigajoule is the dominant metric for heat usage in Europe (ACM, n.d.). Alternatively, the US uses BTU.
Cost structure. First, our propositions require an additional investment in Iot technology for monitoring
the fuel input and heat output on top of conventional costs for boiler heating system and maintenance. Additional investment costs may depend on the method of choice. For instance, the optimization party may rely on human labor, technological solutions such as automated valves, or smart automated algorithms to improve efficiency. However, on long term these investments should be economically justifiable. The return on investment can be made relatively predictable based on the data captured by IoT sensors.
The important question of cost structure is which stakeholder pays for what: In conventional business models, the investment costs in material and labor are predominantly the responsibility of the user. They invest upfront for material, installment, fuel, maintenance, additional components and other operational costs. In our proposition, a specialized party takes responsibility for either a proportion or the entirety of the heating mechanism and fuel supply. The user may prefer autonomy in deciding which material to invest in following the first two propositions. Alternatively, in heat-as-a-service, the business party requires a large say in all specific material used in order to maximize efficiency, reduce fuel use and other costs and, consequently, maximize their own profit margin.
Effectively, in heat-as-a-service, if the user still holds responsibility for providing fuel to heating system, than the optimization party becomes also a customer of the user. For instance, the user has bought gas from a gas company and ‘sells’ this fuel to the heating-system which then converts it into the desirable product of heat. While a user may pay 25€ per Gigajoule of heat, they receive 0.64€ for each m3 of gas used by the heating system2. Thus, the reduced gas usage per Gigajoule of heat causes a higher profit
margin for the business party.
Customer relationship. The relationship between the user and the business is best characterized by the
S-D logic. The user and business work together to reduce the total cost of ownership while improving comfort. Hereby, the user co-creates value by offering building space and use-data and potentially fuel to the business party, but also exchanges valuable information and payments to the businesses. For instance, the user may agree to offer room for a larger boiler, additional heat pump, or solar panel in return for a guaranteed improvement of quality and costs.
The relationship should be oriented on long term. Both user and business are beneficiaries of their relationship. Also, optimizations may generally require a return on investment of multiple years. Ideally, the user perceives the situation as always having an engineer helping them to optimize their heating system, especially with an efficiency contract or in the heat-as-a-service scenario. But even if the user agreed to a single time improvement investment (value proposition #2), they still gain from a guarantee that the investment pays off in the long haul.
Key partners. As argued above, the user co-creates value, may provide fuel, and has influence on total cost
of ownership by either agreeing or declining optimization offerings. Therefore it must be explicitly noted that, in addition to being a customer, the user is also a key partner of the optimization party.
In determining other key partners, we argue the optimization party is confronted with ‘make or buy’ tradeoffs. This is related to the cost-allocation problem noted in the cost structure. For all value propositions, the service starts by monitoring of energy efficiency. From here on, the optimization party may choose to outsource or specialize in additional activities such as material manufacturing, human labor, and fuel supply. Especially in the heat-as-a-service offering, one may benefit from a unified value chain where interdisciplinary functions cooperate towards the same goal of lowering total cost of ownership and improving quality.
Communication channels. In IoT offering, the digitized object functions as the main channel for the
majority of communication. This communication may include real-time performance indicators of energy efficiency based on the sensory data post installment. In our propositions, we propose the data sensors forward the achieved energy efficiency to a display (e.g. a phone or thermostat), and notifies relevant users and businesses if contingencies occur.
investments are proposed or unusual defects occur, the sensory data may preferable be interpreted by knowledgeable engineers that understands the context of the user. Hence, the IoT and the sensory use-data it produces are mechanisms for optimization, but not a goal in itself.
We emphasized post installment to indicate that it remains open how to communicate and market the proposition towards new potential customers. During interviews, we experienced that despite the significant potential savings, our interviewees have argued strongly that, in their experience, the vast majority of their customers have previously been indifferent towards energy efficiency offerings. We return to this topic in the evaluation and discussion.
Key resources. The value proposition requires three types of resources, being operand and operant
resources and the sensory data. Operand resources include the tangible resources such as the fuel, boiler, piping, valves and other material that produce heat. It also involves the IoT infrastructure, sensors and actuators. Operant resources are the human knowledge that is needed in order to optimize energy efficiency for each user. In line with S-D logic, this optimization knowledge is a key asset when multiple parties start competing on energy efficiency.
We found that sensory data is an individual class of resource. While data represents actionable information, it first needs to be interpreted by a technological or human agent. As noted earlier, data is thus only as valuable as the interpretation of the human operator or algorithm, or the combination of both. In order to optimize efficiency, one requires input of fuel supply and output data of heat demand. In addition, the optimization efforts may likely benefit from data of the mechanisms functioning and the users’ value-in-use experience.
EVALUATION
In order to provide evidence that the design is indeed useful, we continue by outlining the evaluation process of our business model demonstration. As noted before, we used a formative evaluation approach which allows to work iteratively during business model design to detect and eliminate problems (Theofanos and Quesenbery, 2005). This is at the expense of more in-depth evaluation techniques used in summative and behavioral studies where pre-existing models and questionnaires can be tested with qualitative metrics (Gregor and Hevner, 2013; Theofanos and Quesenbery, 2005).
The iterative approach constrained us in the number of interviews to be conducted within a limited time span. Data collection started after developing design theory, selecting an application field, initial value propositions (appendix B) and first iterations in the business model canvas. Still, the approach did allow for (indirect) co-creation of the business model design from a rich range of perspectives, which we found is more relevant during development of heuristic IoT design knowledge for future business models. Hereby we also emphasize that the data gathering through interviews was only part of the full design experience in which we moved from the black-box in IoT business model knowledge towards our business model demonstration, and that all accumulated IoT-design knowledge resulting from this innovation path is our central objective.
Notwithstanding, we accumulated sufficient feedback to evaluate our design in strength and weaknesses. A chronological overview of participants, their background and full codification of their view points is located in appendix D, and a brief overview of participants is listed in table 3. During coding, we found themes that overlap with the four features socio-technical equilibrium model (i.e., task, actors, structure, and technology) of Lyytinen and Newman (2008). We found comments have been made either favoring or criticizing our propositions based on those features, which proved useful for further deductive coding.
A IT expert. Pilot interview (interview notes), and reflected on technical feasibility via email. B Lector on energy and disruptive technology
C Business developer of utility company
D Two business developers of boiler manufacturer (interview notes) E IT engineer of an advanced building heating algorithm
F A start up entrepreneur using the IoT and a service concept for selling durable ‘washes’ (i.e., a washing machine is leased, but the product sold are washes).
Interviewee viewpoints
To start with the positive, our interviewees do not worry for the technical feasibility of our propositions (i.e., whether the IoT technology can truly afford what our business model intents). Interviewee A argues the IoT is ready for monitoring and interventions in general. Interviewee B refers the IoT as an “enabler” for monitoring performances. And interviewee C said for the technology to become operational “all ingredients are there”. Accordingly, interviewee D, F and G have both put IoT technologies in action that provides customer support during use. And interviewee E is developing a heating algorithm that is from a technological viewpoint more advanced than our proposition.
To our business model task, during design, we made incremental changes by including the entire heat system as the ‘mechanism’ we intend to optimize for energy efficiency, rather than solely focusing on the boiler. Interviewees B, and D noticed that our early scope that only included the boiler was to narrow. While theoretically possible, interviewees D argued it is better to conceptualize the entire heat system as various components and factors influence each other’s efficiency. Interviewee B noticed that involving the entire heat system (i.e., the boiler and radiators in a conventional home) would lower transaction costs for optimization when mechanic needs to visit a site. Also, interviewee E demonstrated that in modern heating system, the boiler can be turned off 90% of the time by which it mostly serves as a backup, and that a variety of other more durable appliances can provide heat on those moments.
It is this integrated viewpoint that offers the potential of our business model to create better margins and better products. As interviewee F puts it: “in the (value-)chain you can save much: Water usage, infrastructure, detergent production and distribution, machine design and distribution, you can save so much cost on maintenance cost”. While F talks about washing machines, the same argument equally applies on other appliances such as boilers, radiators and other components that form a building’ heat systems: an integrated view on all components that afford value-in-use while costs are minimized.
As argued in ‘customer relationship’ in our design chapter, we notice that in order to truly capture those benefits typically requires a long term orientation. In accordance, interviewees D argue that mechanics (i.e., their industry partners) need to learn to cross sell (i.e., to sell additional products to a customer), but that they currently are not ready to do so: “Mechanics know how to use an adjustable wrench, and that is it. They are not salesmen.” The cause for this inability to sell additional products and services (i.e., to commercially market our proposition) is argued the lack of long term orientation by customers, and that there is no incentive to think otherwise. Interviewee C argues “The problem is that customers know the issue [system deficiencies] but do not experience it as such. Even mechanics cannot sell the proposition to enter ones house for four hours and make everything better.” They also believe marketability is the major challenge we face. Interviewee A argued for tangible outputs, not long term fuzzy cost savings. Interviewee B worries about transaction costs, as even a single site visit by a mechanic of installing IoT equipment has a long return on investment. Interviewee D strongly advocated that the customer is not interested in savings. “They just want comfort, no matter at what costs. If they feel cold, they just turn up the thermostat and if that costs money or what happens next, that is of no importance”. Even interviewee F, an advocates of durable systems and change, acknowledges that their “customers are not really concerned with energy saving. They care about cold beer and quality washes”. Similar to interviewee D, their customers want “comfort” above everything, and only secondary “get the perception that everything is done at the most sustainable way” (Interviewee F).
Still, interviewee G advocates that a new market is slowly forming where customers want to focus on their core business and outsource secondary tasks such as building heating. Both F and G aim to break the market trend by which secondary activities (i.e., activities not within ones core specialization) become commodities and business must pursue what G refers to as ‘volume contracts’. G advocates that if “you are secondary […] you add little value, so you have to be cheaper cheaper cheaper. With a performance contract you can better explain your value and proof you are not a commodity. In a volume contract, there is little to gain. The price has to go down but there is no room for improvements anymore.”
business model. They note the heating-system industry field is well-matured with strong but balanced forces of industry partners, such as mechanics and grid operators, which they depend on for sales. On a similar note, interviewee A argued his experience that businesses rather compete than collaborating over data ownership. Interviewee E argues that “At release the installation is configured and if you are lucky, than people will look at it again and some stuff is reconfigured, but generally it is never changed again”. Interviewee C problematizes the trust the customer required for new monitoring and billing methods. Interviewee G summarizes the cultural issue:
“A business is used to think that in the end it is about the lowest price. So yes, if that is not the case you have to work harder because ‘that is what matters’. A customer looks at the business and thinks, ‘yes, nice a long term partnership, but I am just a customer and you just want to profit of me’. How often do you see small contract prints, for instance if your phone is broken down, that the issue is not within the contract? ‘Sorry this is outside the insurance’. For everything you do, you must pay extra. A lot of those images we have are based on the past in a traditional market.” To counter a cultural change is needed. Interviewees D argue they incrementally push the industries cultural boundaries using a long march approach where they move two steps forward, and take one step back if any partners feel threatened by their move. To do so, they also invest in education of mechanics. They argue if partners can convey a convincing story of safety risks or energy efficiency opportunities, they do have an opportunity to cross sell additional components and services that are beneficial to users. Interviewee F and G propose another approach by testing new service business models in order to legitimize their use to both customers and other businesses. Interviewee G also argues the performance contract is not a goal itself, but a tool to force a change in the relationship between the user and business by which they establish trust, recognize each other’s goal, and establish a long term win-win relationship.
DISCUSSION
Evaluation indicates that interviewees agree our business model design is theoretically possible in terms of technology and task, but in practice the dominant institutional cultural logics form barriers preventing long term, value co-creating relationships for business to consumer (B2C) and the business to their wider industry partners (B2B). We continue to discuss these findings below.
The central role of value co-creation in S-D logic business model combined with IoT technology can – in theory - afford goods-centered businesses to pursue advanced sustainable practices. For instance, our design contributes to savings on fossil fuel, maintenance, and material waste by economically incentivizing the use of durable goods over non-reliable and disposable goods. We note that such value co-creation is closely related to customer-business integration, which has long been recognized as one of the strongest associations with performance improvements in supply chain management (Frohlich & Westbrook, 2001). It creates customer loyalty and enables best practices by creating an optimal mix-and-match of goods and services (Enkle, Perez-Freije & Gassmann, 2005).
However, the institutional gap put forward by the interviewees is a major red flag. Institutional theories argue that behaviors are patterned and reproduced because social norms have become taken for granted (Scott, 1987). Such cultural barriers contribute most to the implementation risk of new IT (Scott and Vessey, 2002). If one starts implementing our business model vanilla in a conventional G-D company, there is likely a major gap between the generalized, abstract conception of our business model as designed and the specific performative practices most actors such as employees, incumbent industry partners and customers believe in (Orlikowski, 2000; Volkoff et al., 2007). Normative authoritative industry standards are imposed structures that are most difficult to change without broad consent (Soh and Sia, 2005), and pressures actors to conform to institutional norms and practices (He and Baruch, 2009; Greenwood and Suddaby, 2006).
Lusch, 2004), and may likely contributed to a degree of cynicism and distrust between two parties that perceive each other in having competing goals (e.g. “a business just wants to profit of me” – interviewee G). Just recently has the dominance of G-D logic been problematized of inhibiting the transformation from G-D logic to S-D logic (Evardsson et al., 2014; Skalen and Edvardsson, 2016; Vargo and Lusch, 2014).
To understand why these B2B and B2C relationships are similar and thus both constrained by the same institutional structures, we need to consider the S-D logic on an abstract level. As displayed in figure 3, the ninth foundational premise of S-D logic argues that all social and economic actors are resource integrators, and thus each actor integrates resources from a range sources that form their specific, subjective value-configuration (Vargo, 2009). Accordingly, we should also not differentiate in a B2B market and B2C market, but consider they are both equally the same (i.e. resource integrators). This also means that, conversely, each actor is also a beneficiary when they exchange goods, payment, use data or other resources that are valuable to them in the value-configuration space (Vargo, 2009). In other words, each actor – either user or business - is just another node in a network-to-network exchange playing field.
Figure 3. Value co-creating network-to-network exchange (Vargo, 2009)
Our evaluation indicates two distinct approaches interviewees pursue for overcoming institutional barriers. The organization to which interviewee D belongs has obtained a well-established central role in a matured industry and moves forward by incremental steps. In turn, new entrants, such as interviewee F and G, positions themselves at the periphery of their industry and aim to brute-force change with a ‘just-do-it’ mentality in order to prove the legitimacy of their alternative service business model. Peripheral players are less aware or constrained by institutional logics, and often disadvantaged by the prevailing institutional arrangements that privileges elite central players (Greenwood and Suddaby, 2006). In turn, central players may become increasingly aware of the contradictive sub-optimal effects their status quo produces, which motivate them to abandon their comfortable established practices and increase their openness to adopt alternative logics (Greenwood and Suddaby, 2006).
Both positions have pros and cons to force change. Central players have legitimacy, economies of scale and more resources to change, but are both constrained and advantaged by established institutional practices, while new peripheral entrants can more easily abandon institutional logics, but have fewer resources to do so and are commonly perceived by the market as less legitimate and to involve more risk (Greenwood and Suddaby, 2006). Central institutional entrepreneurs - such as interviewee D - are aware of the current technical opportunities that afford advanced sustainable business models, but institutional norms and interdependencies constrains their movement speed (Greenwood and Suddaby, 2006).
are under increasing pressure to put price over quality. Second, Greenwood and Suddaby (2006) argue if multiple actors bridge their boundaries, awareness of incompatible goals between the different actors become imminent, which thus creates awareness of alternative logics. This again relates to the benefits of co-creating value in S-D logic (Vargo and Lusch, 2004). Third, the technological scope (what is possible) vis-à-vis the institutional jurisdiction constrains central players to adapt, but also creates resource asymmetries between regulators and entrepreneurial organizations (i.e., if resources for change exceed those of regulators), which causes to weaken those constraining institutional effects and grows openness to alternative ideas (Greenwood and Suddaby, 2006). In figure 4, we summarized the contradictions that institutional logics may produce in the social subsystem.
Path for institutional change
From a sociotechnical systems’ perspective, institutional structures is one component in the reciprocal interplay between materials, institutional structures and individual/small groups of actors (e.g. DeSanctis and Poole, 1994; Leonardi, 2012; Volkoff et al., 2007). While dominant institutional structures may constrain change, it does not dictate it. In addition, the novelty of IoT technology and contradictions in system-performances offer opportunities to drive change.
As visualized in figure 4, we acknowledge that the G-D logic is dominant in most goods-centered industries, but propose that these cause various contradictions which can be leveraged by IoT technologies and best practices to trigger sensemaking processes, revisions of social and material practices, and, in effect, renegotiations of dominant institutional logics. We continue to clarify our model below.
