Business models as scientific models: The question of perspective?

1 Introduction

In 1998 [1, p.1], the sociologist Zygmunt Bauman stated that “globalization is on everybody’s lips; a fad word fast turning into a shibboleth, a magic incantation, a pass-key meant to unlock the gates to all present and future mysteries”. Almost in the same period, the term ‘business model’ started to be used extensively among practitioners and theorists in the scientific management field. Emerging from e-business [2], the business model concept has evolved into an apparent pass-key that can open a passage to business success. The concept of business model sky-rocketed in national and European policy documents in association with innovation and social entrepreneurship. The idea was that ‘out there’ exists ‘the right’ business model. A company only needs to pick up from the list of business models the right one in order to be effective and competitive [3]. When the company is able to choose the right business model and employ it appropriately, this will give it a kind of competitive advantage [4] or will allow development of a more socially and environmentally responsible company [5]. At the same time, the wrong one destroys its market and competitive position regardless of its technological sophistication or product quality [6].

The business model concept has been theoretically elaborated [7], and today is in wide-spread use. Based on our literature review, Massa, Tucci and Afuah [8, p.76] have identified “three interpretations of the meaning and function of ‘business model’ as 1) attributes of real firms, 2) cognitive/linguistic schemas, and 3) formal, conceptual representations of how a business functions.” However, they also point to a lack of agreement on a single definition of what a business model is. It seems there has been a continuing discussion [7] to which earlier works [2, 6] also point. Moreover, different definitions provide a range of elements that create a business model concept. Seelos [9] argues that the concept became popular with internet expansion in the 1990s. The business model was related to and emanated from e-businesses. Seelos [9, p.1–2] also states that “since then, various authors have positioned the term business model as a key concept in understanding organizations but at the same time have voiced concerns that the rapid proliferation of the concept has created an abundance of meanings”. Maucuer and Renaud [10] provide a bibliometric analysis of business models’ literature pointing to future trends and also to past developments. They point out three significant weaknesses of the literature to date. Firstly, the literature is based on transdisciplinary approach; secondly, the analysis is too straightforward, based on ‘word by word definitions’ analysis’ rather than on meanings that lie behind the words, and thirdly content –related specifics have not been sufficiently considered. Their analysis also sheds light on the lack of practical perspective and application of the business model concept in the real life of companies [11]. The underlying assumption is that the whole discussion about business models has possibly emerged among scientists and practitioners because of the potential economic interest. It seems that the power-knowledge lies in advisors, consultants, and as well as academics as they have the ‘authority’ to say what should be done in order to create value for owners (companies) and customers [12]

The abundance and proliferation of ideas have led to what could potentially be called mainstream discussions on business models published in different special issues on business models by high-quality journals such as Long Range Planning, Strategic Organization, International Journal of Innovation Management and R&D Management. However, there have also been interesting discussions published in conference papers [13] and scientific journals [14] approaching business models through narratology. Besides research and conceptual developments of the concept, the real-life of organizations has also been affected, among other things by consultants and business-people ‘helping’ organizations to ‘draw’ and develop the ‘right’ business models. The ‘right’ model has also been sought by theorists who have tried to capture law-like regularities underlying the business models similar to the one discussed by Tsoukas and Cummings [15] for the field of organization studies. The latter implies that business models share characteristics that belong to scientific models. Lately, Cuc [16] has conducted extensive bibliometric study on business model literature. He also presented the “business model research co-occurrence of keywords density” [16, p.10) in which co-occurrence between business models and scientific model does not exist. This is the identified gap in current literature on which this article aims to shed light.

According to Porter [17], a company creates, delivers, and captures value within its activities that are supposed to be unique enough to provide it with some form of competitive advantage. This idea, in a more or less modified form, is followed by many mainstream conceptualizations of the business model. Considering a more kaleidoscopic view of a business model than this traditional one is, the nature of business models can be sought within two relations. One is the relation reality-model, and the other one is the relation model-manager. If we discuss and theorize the first relation, then we need to answer the question: Does a business model represent a ‘reality’ of a company? If we theorize and focus on the second relation model-manager, then we need to bring into discussion the significance of context in which a business model is embedded together with the manager.

The purpose of this paper is to explore and discuss business models. The paper is guided by two research questions. RQ1: Are business models scientific models? RQ2: What value business models as scientific models have? To reach this objective, we used a narrative research approach using elements of grounded theory analysis.

The literature in the field of business models is extremely broad and the discussions touch on their contribution to organizational theory in general as well as specific theories in the field of strategic management, marketing, sustainable development, and competitive advantage. In the literature on business models, however, we do not find theorizing about their possible predictive value as a scientific model. In this paper, we provide key definitions for understanding the business model as a scientific model and we address the issue of its predictive value by use of relevant philosophical sources. The theoretical contribution of this paper is about informing an existing theory rather than constructing new knowledge. From a practical point of view, the article’s contribution is in understanding the actual limitations of the business model and the conditions under which the latter can be properly applied.

Although this article is conceptual [18], the discussion is enriched by previous studies on business models. The emerging theory in this article is inductively built on the praxis, informed by relevant theoretical and empirical articles and addresses the identified gap in current publications. The paper has some limitations that need to be considered. Firstly, there are few conference papers included in analysis even though there is an assumption that new ideas and perspectives emerge as conference papers first. Secondly, although there is a body of philosophical sources that could be relevant for better understanding the concept of scientific model, this paper focuses on business. Thirdly, this is a conceptual paper that uses some data from different empirical studies; however, the focus is not empirical results but on better understanding of the concepts under discussion.

In the following sections, we first present the business model concept before we discuss business models as scientific models. Finally, we provide an argument to support the proposition that business models can be considered as scientific models that are, however, embedded in specific organizational, cultural, and technological contexts. In conclusion, we propose some directions for future research.

2 On business models

There has been an extensive literature review on business models conducted in scientific journals and by different authors. Shafer et al. [19] identified 12 definitions and 42 different business model components or building blocks in the period 1998–2002. Osterwalder, Pigneur and Tucci [20] surveyed a sample of 62 participants who were asked about what they understand to be a business model, producing 54 definitions. Almost every participant has his/her definition/understanding of a business model. Baden-Fuller and Haefliger [21, p.419] state: “The business model construct has become attractive to many academics, taking on its momentum as is evidenced by the fact that, in the three years since publication, the Long Range Planning [22] special issue on business models attracted more than 150,000 downloads and more than 3,500 Google Scholar and more than 500 ISI citations”. As well as these published papers, Massa, Tucci and Afuah [8] point to lively academic discussions in the last five years at the Strategic Society conferences, Academy of Management Annual Meetings, and DRUID conferences. Lanzolla and Markides [23] say that while the business model construct may not be entirely new, it can still provide new theoretical insights in strategy, complementary to Resource Based View and Market Positioning.

The business model can be considered, a set of activities or routines [24], a tool [25–31], as a system [32], a way, or a method of doing business [33]. Baden-Fuller and Morgan [22] define business models by analysing how business models act. In their view, business models “act as various forms of model: to provide means to describe and classify businesses; to operate as sites for scientific investigation, and to act as recipes for creative managers” [22, p.156]. They have influenced researchers world-wide. Baden-Fuller and Morgan’s [22] paper “Business models as models” has been cited 2077 times, according to Google Scholar (14 April 2022). Other papers written by the same author and discussing the relationship between business models and models were also cited very often. Baden-Fuller and Haeflinger’s paper from 2013 has been cited 1350 times (Google Scholar on 14 June 2022), and the paper by Baden-Fuller and Mangematin [34] has been cited 657 times (Google Scholar on 14 June 2022). This interest in considering the business model as a particular type of model could probably be related to the general interest in understanding models in social and natural sciences as indicated by 1655 citations (14 June 2022) on Morgan and Morrison’s [35] book “Models as mediators: Perspectives on natural and social science models”.

Apart from this abundance of mainstream or key literature, there is another view on business models called the narrative perspective. According to organizational and management literature [36–41], a narrative approach could be a possible approach to theorize business models within their contexts. The narrative approach is based on context-dependent knowledge. It is dynamic flexible, and oriented toward action. A business model understood as a narrative is a recurrent, re-told, and re-constructed story that describes and explains how a company works. Business models as narratives represent the most fundamental sense-making form of company’s experience [36]. Narratives “not only give meaningful form to our experience. They also provide us with a forward glance, helping us to anticipate situations even before we encounter them, allowing us to envision alternative futures” [36, p.380].

Haggege and Collet [13] discuss different views on business models. They point to ontological, systemic, choice/consequence, and narrative perspectives. They see the ‘narrative perspective’ as the emerging one. Considered as narratives, business models are explanatory. They seem to include all the players and the sequences by which a value proposition will be delivered. In that case, we need to consider the influence of culture and tradition within particular national and business contexts in order to understand business models properly. Business models are not abstractions that aim at representing bounded realities or pieces of reality that are embraced in some general principles detached from the activity, the actor involved and the surrounding context. This decontextualized and agentless reasoning was evident at the time of the e-business boom when business models were regarded as boundless and de-contextualized. Recent developments have been increasingly recognizing the context and cultural dimensions embraced in organizational research and business models themselves [42, 43].

This recognition has also led to the multiplicity of business models [22] and to the understanding that a business model is a ’practical thing’ that has useful and analytical value to entrepreneurs and managers running a business [44]. At the same time, the concept of business models has also been of interest to organization scientists, mostly to classify businesses [45] and of interest to business consultants who looked for another source of competitive advantage. However, there is little, basically no, literature about what and how relations between business research and practices possibly affect the development and dissemination of business model theories. Gabriel [46, p.142] proposes that using organizational theory in practice “tactically means being selective, eclectic and flexible in one’s uses of recipe books from which to draw ideas for creative bricolage”. Similarly, another theorist [47] have argued against the ‘applied toolkit’ model. There is considerable literature that sees the theory applied to business practice as little more than ex-post-facto justification, consultancy consumerism, or mere caprice. The “chasm” that Czarniawska [48, p.8] spoke of does not seem to have diminished, which indicates the need to explore the implication of business models as scientific models.

The conceptual discussion is grounded in current literature using methods of comparison and compilation as well as a critical review. The following data bases were used: Google citation, ScienceDirect and Scopus (using keyword ‘business model’ and ‘business model innovation’), special issue of Long Range Planning Journal, and selected articles from Journal of Business Models. Also, Google Scholar platform was used to identify conference papers etc. that could open different streams of research. Two basic criteria were used: the business model had to be discussed in a non-trivial and non- marginal way; the business model had to be related to an organization.

This paper is also grounded in results of the EU cross-border research; from these projects the bibliographic references were also drawn [49–52]. The project targeted the development of a business model canvas that would shed light on existing business models as well as set grounds for companies’ future developments on their internal innovative and strategic potentials and external market opportunities.

Secondly, the compilations of narratives from these projects were used and based on the principles of grounded theory [53]. The aim was to explore theoretical issues of business models, and, if possible, to identify an ideal type of business model for each branch. Although narratives have been used in management research in the last 25 years [54], long before that they had emerged in the social sciences [55]. Specifically, for business models, we have not identified a bibliographic source and authors that would deal with narratives in business models creation and analysis. A structured narrative [56] enables ‘discussion’ between participants in the study and the canvases they have created. Such a structured narrative is the sense-making process, and translation of current status of the situation of a company into a meaningful future. What counts is the process rather than the desired stage. In the company’s dynamic life, a narrative about the future through Canvas is an unstable yet enriching process that also contributes to organizational culture. Narratives, when told and/or written, are the organization, not only about the organization. The use of structured narrative shed light on one of the critical questions: “Are business models scientific models that have explanatory as well as predictive power?” Based on empirical data from the studies, presented theoretical discussion is built, and the basis for a new interpretation of the nature of business models is constructed.

3 Business models as scientific models

Past research [57–59] has shown that managers seek out scientific models in the belief that they will provide accurate representations of their conditions and reliable means of gaining a competitive advantage. Bailer-Jones [60, p.59] states that scientific models, with some success, “tell us something about certain features of the world”. She draws her arguments from examples in the ‘empirical’ (or ‘natural’) world. While these models cannot produce definite truth or falsity, they do provide useful heuristic devices.

When a model is understood as an interpretative description of reality, it can make access to phenomena or objects easier. A model is not a trivial description. It gives a description that is an interpretation in that the description goes beyond what “meets the eye”, e.g., by exploiting a theoretical background that is relevant to interpreting the phenomenon” [60, p.61].

Bailer-Jones’s understanding of models implies a positivist standpoint in understanding social phenomena. Hughes [61] states that within social science-inspired by positivism the main questions revolve around theories that could attain the certainty of theories in natural sciences; verification and falsification as fundamental criteria for distinguishing scientific from non-scientific statements; and the sovereignty of the empirical.

Current studies on business models focus on explaining, understanding, and developing typologies and taxonomies to build a ‘scientific’ foundation for business models [22, 62–65]. The classifications of models and theories are built within the positivist paradigm with its assumptions about truth and the nature of scientific problems. Consequently, they conform to the conventional wisdom about having the right business model amid a multiplicity of wrong business models where the ‘right’ business model is often associated with a competitive advantage, efficiency, and effectiveness [66]. The problem with such claims is that business models seem to be empirical descriptions of a world that is unstable, subject to different discursive practices and perceptions, and dependent on value judgements about right and wrong as well as good and bad. The operations of a business world are highly contextual and subjective. They can rarely if at all, be representations of a sovereign empirical and tangible world. Instead, narrators build and construct their companies’ business models. When narrators/managers adapt business models to the firm’s language, identity, intention, and context, they change it from a business tool into a specific time-context strategic character of the firm. Business models emerge as constructs in time–space–actor relationships. Each business model constructed has its own life. That means that they cannot be replicated, at least not in their entirety in any other moment or any other network of relationships. This fact entails a question about what business models scientifically represent. Bailer-Jones [60, p.61] states that “the term representation points to facets of scientific models that cannot be captured in an account exclusively relying on true propositions, while an analysis in terms of true and false propositions will highlight what more is required for representation to take hold”. Representation moves the model concept away from fact. From a structuralist perspective, a “representation” does a job that scientific models cannot do because such models exclusively rely “on true propositions, while an analysis in terms of true and false propositions will highlight what more is required for representation to take hold” [57, p.61]. To a large degree, the scientific rigour of truth and falsity is irrelevant in considering representational models for four reasons:

“Models do not entail propositions.

It is not surprising then that Godfrey-Smith [67, p.734] considers modelling to be the “deliberate detour through merely hypothetical systems” that it incorporates. Models will be less general than a theory because they are not intended “to describe a phenomenon in its entirety. They set themselves smaller, manageable tasks and focus only on selected aspects of phenomena” [60, p.66]. Such models “can be moulded and adjusted to address the real and concrete empirically observed situation” [60, p.62]. These adjustments are further simplified in order to obtain a practical result [60]. Accuracy of the model is less important than its capability to perform certain tasks [68]. Representational models are not complete. They are selective and partial. Such models include some true but also some false propositions as they are simplified. Basically, reductionism and false propositions are incommensurable features of representational models [60].

The pragmatic approach provided by Knuuttila [69] is a far looser conception of representational models. In his view, the structural model is “isomorphic with respect to some real system or a part of it” [69, p.164]. This moves away from the notion of models as containing explicit and implicit propositions. That is, models can provide various representational means to convey “diverse kinds of information, much of which cannot be readily, if at all, propositionally presented” [69, p.166], through processes of affordance. Thus Knuuttila [69, p.166] criticizes models for failing to “pay enough attention to the expressive and inferential power of the iconicity of signs”. Affordances are “dispositional properties” that provide the “characteristic background circumstances, triggering circumstances and manifestations” [70, p.949]. They can be represented as “X has affordance property A (at time t relative to an organism O in circumstances C)” [70, p.956]. Thus, it is a “subjunctive conditional” which renders it in practice as:

“At time t, if background circumstances C were the case, then (if a set of triggering circumstances T were the case, then a manifestation M involving X and O would be the case with probability p” [70, p.956].

It is this contingency of representational models, which some might consider a weakness when compared with scientific fact, that makes them useful, says Knuuttila [69]. The “cognitive value” emerges from the interaction with them: “models give us knowledge not because they happen to represent their target systems more or less accurately but because they are purposefully constructed to allow inferences of various kinds” [69, p.169]. We cannot study all the complexity of the real world. Therefore, we use models that allow practitioners to “simplify or modify the problems scientists deal with”. Affordances are built into models as specific model assumptions [69, p.170].

Managers’ desire for discovering scientific knowledge that supports the development of the right business model draws the attention of scientists who try to explain the business (social) world with natural science tools. However, business models are not universally right or wrong because they lose usefulness when de-contextualized. Only when they are contextualized do models become exemplary cases, which are, no doubt, interesting for scientists and managers, and become known as the Ryanair business model or the Amazon business model or the Nespresso business model [3]. This is what Knuuttila [69] describes as presupposed activity of competent and informed agents. The “representational force”, according to Suárez [71, p.768], is “the capacity of the source to lead a competent and informed user to a consideration of the target”. This “relational and contextual property of the source is fixed and maintained in part by the intended representational uses of the source by the agents” [71, p.768]. In the case of a business model, it could be argued that contextualization of the company is a continuous process of the agents who need to make relational and contextual properties reflexive and processual.

Some “typical” business models can serve as ‘archetypes’, but it is questionable whether they are of any use for practitioners or theorists, since the time, the context, the actors, their assumptions, and their dispositions can never precisely replicate ‘the moment’ when that idiosyncratic business model was first explicated. When a business model is represented in ‘canvas’ or other diagrams, the unique circumstance of that particular ‘case’ (the company) needs to be explicated in order to be fully understood. Currently, when we give comments on the model, these commentaries express proposition [60, p.64] which, hence, are capable of being true or false. The question is whether this is a central issue, and if so, how it is related to the predictive power of business models.

4 Predictive power of business models

The question of predictive theories, predictive models and predictions is important for understanding business models as scientific models. Prediction and explanation in social science emerge from the natural sciences. However, in social sciences, the notion of predictive theory and prediction is questionable because they are based on invariable assumptions about highly variable human and societal phenomena. To a certain level, this point is addressed by Seelos [9], and Morgan and Morrison [35] when they describe different “types” of models. Morgan and Morrison [35] point to various tasks that models perform. There are two significant groups of models: empirical and theoretical models. Maps, case studies, role models and similar, are empirical models. Theoretical models are used in the sciences [35].

It is the type-role relation that seems to be associated with predictive power. Baden-Fuller and Morgan [22, p.157] discuss “business models as descriptions of ‘kinds’ in a taxonomy, business models as either model organisms for investigation or business models as recipes”.

Baden-Fuller and Morgan [22] point to biology and economics when they discuss ‘model organism’. They point to research designs (experiments and different forms of manipulations) that enable scientists to build new knowledge, understanding and explanations of the world. Their suggestion is that more similarities are found between biological models and business models than economic and business models. Baden-Fuller and Morgan [22] build their argument on the idea that by learning ‘an example’ biologists also learn about ‘big picture’ of life. From that point of view, it could be argued that by learning, exploring, and understanding a specific, individual, exemplary business model we learn more about business models in general [22]. Even an ideal type of business model can be created. This ideal type of business model has at least two functions. It serves as a) benchmark in the field, and thus as exemplar case against which other business models are measured, and b) as a framework that could be filled with actual data and evaluated in the timeline. In both cases, a business model is a simplification and reduction of real life. It is variable because the constituencies can be changed and the relationships between constituencies vary. However, the accuracy of a model is less important than its capability to perform certain tasks [68]. As such, a de-contextualized business model has sufficient ‘generality’ to perform certain tasks, and to be considered theoretical, and it also has sufficient firmness in terms of framework and constituencies to be considered ‘practical’. It can be modelled; however, the sense of modelling is questionable due to unpredictable changes and evaluative judgements both well-presented above. Hence, its predictive power is questionable.

5 Conclusion

This paper focused on two research questions: a) Are business models scientific models? and b) What kind of value do they have as scientific models? The discussion indicates that the answer to the first question was sought within two relations. One is the relation reality-model, and the other one is the relation model-manager. If we discuss and theorize the first relation, then the business model is a scientific model that is however simplified, inaccurate, generalized, and incomplete. It does not incorporate all complexity of business processes and companies, or people and society. Hence, from that view, business models can have limited value as they represent a reduced picture of a company. If we theorize and focus on the relation model-manager, then we can explain business models as narratives about the company. Business models applied as scientific models but decontextualized and separated from the actors take the focus of discussions away from the basic definition of what business models are, namely constructs in time–space–actor relationship. Business models are then unstable, contextualized, re-narrated learning tools.

A generic model that represents a simplified reality has little value and no specific predictive or explanatory power as required by scientific models because the ends and the action and the actor(s) are not embraced in it. Even considering the minimalist inferential conception of scientific representation proposed by Suarez [71, p.776] who says that a representational model does not have “to shed light on the truth, empirical adequacy, or completeness” business models cannot be considered as scientific models without considering the ends towards which the business models are constructed. Only when the business model is contextualized does it become a strategic thinking tool that can be discussed and studied. The business model becomes a kind of a “case study” based on narrative. At the heart of it is a process of reflection that is expressed through narrative. This process gives the business model both analytical and practical value and, to some extent, predictive value. The latter is not expressed in numbers nor represented in maps and diagrams but through language. Predictive value is embedded in the process of “storytelling”, encompassing events of the past and present intertwined with the expected future. It is interesting that Magretta [72] said that business models are anything but arcane economical models; rather they are stories that explain how enterprises work. Later this definition was partially lost in some literature [22] with the consequence that the importance of the business model for the reflection process was minimized. The reflection process itself, based primarily on feelings and intuition, has the power to identify new relationships, new facts, and emerging processes within the company in light of past and present insights. The narrative is, therefore, not static as it is constantly revised with new insights. This represents the analytical value of the business model. The practical value of a business model comes from its conceptual representation of how a business operates. This allows the user to evaluate the underlying logic of the business quickly and efficiently, and thus serve as a managerial tool. The strategic nature of this tool is expressed in the ability to have a broad view, while also detailing the individual building blocks of the business. These characteristics allow us to understand the business model as a scientific model if we look at it from a social constructivist perspective. The latter is characterized by the fact that it does not envisage the generalization or conception of ’ideally typical’ solutions, but rather pragmatic solutions that are constructed within interactions of individuals. The business model also offers such solutions. On the one hand, in an uncertain and challenging business environment, it can provide an orientation for companies, and on the other, it offers a first-hand learning experience.

For managerial practice, this means that it is important to understand how narratives are narrated and re-narrated in a specific organizational temporal-spatial context. The basic narrative embodied in the business model is created by connecting many related partial narratives in a kind of sense-making web that reflects a coherent theme. Such a predominant narrative is self-reinforcing and can remain present in the organization despite a crisis or new even better narratives emerging within the organization. The business model as a narrative represents a stand-alone frame of reference that, when not properly understood, prevents an organization from questioning the principles on which its past success is based. In this way, it creates circumstances in the organization where the business model becomes a potential source of organizational inertia and path dependence and not a tool for transformation.

This paper is conceptual and as such does not have empirical data. The paper is not positioned as a thorough review of the extant literature nor does it seek to propose a new theory, but rather it links work across disciplines by providing a multi-level insight. Nevertheless, there is still great potential to theorize this perspective on business models and to develop future research on business models in this direction such as the phronetic view on business models.