Abstract
This article presents a common typology of multi-sided platforms and virtual communities. The analysed entities comprise 69 of Poland’s most prominent websites, representing one of two business models. Based on three dimensions: collaboration among users, reputation-based user competition, and user multi-sidedness, we identified four business models. These are problem community, object community, object market, and reputation market. The differences in their logics of value creation go beyond the traditional division between multi-sided platforms and virtual communities. The identified business models were depicted in qualitative system dynamics diagrams. The typology was verified using the statistical clustering technique, which yielded corresponding results.
Similar content being viewed by others
Explore related subjects
Find the latest articles, discoveries, and news in related topics.Avoid common mistakes on your manuscript.
1 Introduction
A typical online discussion forum differs from an average marketplace. The former can be classified as a community, while the latter is a multi-sided platform. However, many entities evade easy classification. Are Couchsurfing, Airbnb, and Booking.com virtual communities or multi-sided platforms? These companies likely do not share a single business model, although they belong to a particular continuum of companies that facilitates lodging booking. Couchsurfing connects people looking for a place to stay while also being interested in hosting guests, Booking.com is an intermediary between tourists and hotels, and Airbnb combines features of both services.
As the examples above demonstrate, companies often blur the lines between the two business models, and it can be difficult to classify them as one or the other. By examining virtual communities and multi-sided platforms together, researchers can better understand the continuum of companies that combine the characteristics of both models. This can lead to a more detailed understanding of how these models operate and create value for their users and stakeholders. Furthermore, as the boundaries between virtual communities and multi-stakeholder platforms are blurred, it is essential to have a comprehensive framework for analysing these business models. This can also be of great importance to policymakers as it helps them regulate these entities, ensuring a healthy and competitive market.
As we demonstrate later, despite the extensive classificatory scholarship, multi-sided platforms and virtual communities are classified separately in different research streams. They tend to be confused with each other, and the resulting classes are often excessively heterogeneous and sometimes include other business models. There is a conspicuous gap in the research on classifications encompassing both business models. Hence, the goal of the research was to identify the most common types of both business models and their logics of value creation.
The vital contribution of the research is the system of four identified business models. It may be treated as a middle-range theory explaining how virtual communities and multi-sided platforms create value. We show that the differences in value creation logics go beyond the traditional division between multi-sided platforms and virtual communities. We explain the value creation logics in identified business models with system dynamics diagrams. The study introduces new typology dimensions, i.e., collaboration among users, user competition, and user multi-sidedness. It advances thinking, especially about the last one, as user multi-sidedness has not been operationalised yet. The research also highlights objects as the key elements of platform value propositions that attract users to the website and satisfy their needs. Finally, this article contributes to the debate on the role of a business model in market competition.
This article proceeds as follows. We begin by discussing the existing classifications of both business models. Next, we present a detailed methodology for creating the typology. The subsequent section describes the resulting four business models. Afterwards, we discuss their logics of value creation and implications for research on classifications of business models. The article ends with a description of the main conclusions, further research, and limitations. The appendices present the assignment of analysed sites to typology dimensions and the quantitative verification of the typology.
2 Literature review
Typologies of the business models applied by Internet companies rose in popularity at the turn of the millennium, tracking the growing popularity of Internet companies, whose operations followed a different logic than those of traditional companies. That period saw the emergence of the business model classifications that remain in use today. They include the typologies proposed by Timmers [72], Weill and Vitale [81], [58], and Canzer [11]. Despite their different areas of analysis (e.g. Internet companies, e-business companies) and the varying number of types or taxons they distinguish, most classifications designed around 2000 and more recently [20] differentiate between multi-sided platforms and virtual communities.
A community is a voluntary association of actors, typically lacking in a priori common organisational affiliation but united by a shared instrumental goal (West, Lakhani, 2008). A multi-sided platform is an intermediary enabling direct interactions between two or more distinct sides, each affiliated with the platform [33]. Both entities are described as wielding significant influence over consumer behaviour and companies’ operations. Prahalad and Ramaswamy argue that thematic consumer communities (…) are revolutionizing emerging markets and transforming established ones [57, p. 3]. The platforms’ effect on traditional vertically integrated markets is described straightforwardly by van Alstyne et al. when a platform enters a pipeline firm’s market, the platform almost always wins [75]. Even if these claims seem hyperbolic, the blurred boundaries separating the two business models and their significant roles in transforming modern economies are reason enough to research their relationship.
Virtual communities and multi-sided platforms share common features. User interaction plays a crucial role in virtual communities and multi-sided platforms. Consequently, the network effect occurs in both types of entities. In the case of communities, the direct network effect is predominant, while the indirect one is more pronounced on multi-sided platform (Belleflamme & Peitz, 2021, [68]. Managing the value for users on multi-sided platforms and in virtual communities involves development of a product, just like in the case of one-sided companies, and networks of users [66]. Examples of the former category comprise creating a user-friendly website and providing assistance. Developing a network of users, on the other hand, is often based on the so-called curation, i.e. filtering and controlling users’ access to the platform, the activities in which they participate, and the connections they form with other users [52].
Despite essential similarities, both types of companies have been classified separately. Tables 1 and 2 present identified typologies and taxonomies of both business models along with the analysis criteria on which they are based.
Examining the above summary in quantitative terms, it is evident that classifications of multi-sided platforms are more popular than these of virtual communities. Both typologies and taxonomies are characterised by a similar number of resulting classes, their median is four. The difference is outlined in the number of criteria used—in the case of typologies the median is 2, while with taxonomies it amounts to 3,5.
As is apparent from the research cited above, multi-sided platforms and virtual communities have been classified separately according to various criteria. This is both a consequence of their multidimensional nature and a result of the diverse academic disciplines in which the studies were conducted. Multi-sided platforms are typically analysed in information sciences, management, and economics from the perspective of their architecture, business models, network effects, and market functions. On the other hand, virtual communities are more often studied in psychology, sociology, and marketing. Common perspectives include member engagement, conflicts, and brands. Anchoring the classification in different academic disciplines results in disparate variables, leading to other classes.
Despite a significant number of classifications, few studies combine both types. Muñoz and Cohen (2017) analysed sharing economy business models, while Perren and Kozinets [54] researched lateral exchange markets. However, even these studies do not exhaust the combined categories of multi-sided platforms and virtual communities. Due to the research subject, they overlook non-transactional communities and marketplaces that connect buyers and institutional sellers.
In the case of articles presenting typology, there is a tendency to describe the research method only briefly. In particular, many articles do not include the operationalisation of the dimensions, description of the research group, or empirical verification. Typologies are often considered conceptually derived, resulting from a reflection following the accumulation of the relevant knowledge, thus without a formalised research process. Additionally, the resulting type system is at times a supplementation rather than the main subject of a study. On the other hand, articles presenting taxonomies usually provide a more in-depth description of the research method, which probably arises from the use of quantitative methods.
As mentioned above, multi-sided platforms and virtual communities are regarded as separate entities in general typologies of Internet companies. Nevertheless, they tend to be confused with each other in detailed typologies. Täuscher and Laudien treat a digital product community as a marketplace (2018), while others categorise multi-sided platforms as communities [64]. The relationship between the two business models is also ambiguous in non-classificatory studies. In Choudary's view, community building is one of the platform's key activities (2015). Wirtz et al. [82] point out that in studies on platforms in the peer-to-peer sharing economies, the focus has been on transaction entities and community-based online services. These approaches treat the community as an element of the platform’s business model. In his discussion of reputation systems, Dellarocas [18] presents a dissenting view, counting multi-sided platforms such as eBay as communities. Kaplan and Haenlei [38] similarly count YouTube as a content community, while the site is typically considered a multi-sided platform.
Furthermore, the classification subject itself is frequently diverse, thus hindering comparisons while illustrating the abundance and ambiguity of concepts. This is particularly apparent in the case of multi-sided platforms, which tend to be given related labels, such as platforms, multi-sided markets, and marketplaces. Sometimes, the boundary between different types of business models is shifted arbitrarily. Platforms may be understood as the technological foundation on which other firms develop complementary innovations [17] and hence cloud services or even microprocessors fall into this category. In another research stream, a network effect is considered a sufficient condition for being a platform. Therefore, content providers which connect advertisers with users [25], or even Netflix, which uses recommendation systems, are treated as platforms (Belleflamme & Peitz, 2021. The first approach comes from technical sciences,the other one, relating to the network effect, is used in economics.
The heterogeneity of the companies belonging to the resulting classes is a common shortcoming. This may be perceived as contradicting the principle of typology, which is transforming the complexity of apparently eclectic congeries into well-ordered sets of a few rather homogeneous types [6]. In the research by Cusumano et al. [17], the category of transaction platforms contains heterogeneous entities, such as marketplaces (Amazon Marketplace, Airbnb), dating sites (Match.com), instant messaging (Snapchat), and social media (Twitter, Instagram, LinkedIn). Cennamo treats search engines, along with social media and dating services, as examples of information markets (2021).
Despite the extensive classificatory scholarship, multi-sided platforms and virtual communities are classified separately in different research streams anchored in various disciplines. The classification subjects are diverse, often going beyond one business model, but in no case analysed exhausting both virtual communities and multi-sided platforms. Often companies belonging to one business model are assigned to another. Many typologies are conceptually derived and thus lack the formal classification process. As a result, the resulting classes themselves happen to be excessively heterogeneous including dissimilar entities, sometimes representing even other business models such as content providers.
In light of the analysed typologies and taxonomies of multi-sided platforms and virtual communities, the need for a common classification of both business models becomes apparent. To address the deficiencies or even contradictions sometimes occurring in the cited studies, a potential classification should be characterised by an accurately described research process and clear results of the classification process. Hence, the goal of this research was to identify the most common types of both business models and their logics of value creation. The accompanying research question concerns the differences in the logics of value creation between various types of virtual communities and multi-sided platforms.
3 Research methodology
A typology can be understood as an organised system of types [16]. Types are simplified theoretical entities that combine the characteristics of actual objects. Typologies represent the variety of objects occurring in a particular category. The analysed objects are assigned to categories that should be internally homogeneous and externally heterogeneous. According to this perspective, typologies are classification tools comparable to taxonomies [6]. An alternative approach posits that typologies are not the result of classification because typologies aim to create types rather than categorise analysed objects [22]. Comparing typologies to taxonomies, the typology is more inductive, relies on fewer dimensions, and is more deeply anchored in theory. Taxonomy typically employs many variables, and grouping uses quantitative methods such as cluster analysis. Taxonomies are more generally used in biology [6], while typologies are well-established tools in the social sciences [16].
The typology in question was developed following the four stages of analysis in the process of type construction distinguished by Kluge [41]. This approach is widely acknowledged and has been utilised in numerous articles on typology. It is worth noting that when describing the stages of analysis, the phrase "ideal type" was used, while when describing the resulting four types, the phrase "business models" was utilised.
3.1 Development of relevant analysing dimensions
The first stage is the development of relevant analysing dimensions. Based on the analysis of the literature and the websites in question, three dimensions have been selected and operationalised into ordinal variables. These are collaboration among users, reputation-based user competition, and user multi-sidedness.
Barret et al. [7] state that online communities enable members with common interests or goals to collaborate and interact with one another virtually. While collaboration among users is an essential feature of communities, it plays a lesser role on multi-sided platforms, where it may occur, for instance, when users review the activities of other users.
Competition among users is essential on multi-sided platforms, where users from at least one group compete to interact with members of the other participant group. The highly competitive nature of platforms stems from their transparency. This, in turn, is driven partly by user rating systems, seller reputations developed over long periods, and filtering mechanisms. The competition occurs also in communities. Tim O’Reilly, who popularised the term “Web 2.0”, states that rival ideas and solutions compete with one another in a free market for ideas [48]. Community members compete also for status and distinction [43]. However, it may be argued that competition among community members, unlike among users of multi-sided platforms, is peripheral to collaboration towards a shared goal.
There is a greater diversity of user roles on platforms than in communities. Multi-sided platforms are sometimes described as intermediaries connecting distinct users [28] or at least users belonging to distinct sides [71]. Communities, by contrast, are often said to connect individuals with similar interests, identities, etc. [48], who can serve different functions within those communities [43, 64]. Therefore, a diversity of roles is common to both models although it is more prevalent on multi-sided platforms.
These three phenomena are essential to value creation in virtual communities and multi-sided platforms. They also appear to vary in intensity in both types of entities. Hence, they were chosen as dimensions of the common typology of virtual communities and multi-sided platforms. These variables have not been used in any of the above-cited classifications.
They were consulted then with seven experts, including academic researchers and practitioners specialising in the area in question. The respondents were presented with the research agenda and the typology dimensions. Then they were encouraged to make and express appropriate reflections [23]. The interviews took place without a standardised scenario and usually lasted 20–45 min. These consultations confirmed the general correctness of the dimensions and prompted minor amendments. The number of grades was reduced in one of the dimensions, and terminology was modified in two others.
To operationalise the analysing dimensions, we need to introduce the concept of the object. This is a unit of content submitted to a community or multi-sided platform by its users. It typically constitutes the first post in a given thread, thereby enabling other individuals to respond with their statements. Examples of objects include an image, video, or audio file; a description and photograph of a product along with its price; a classified ad for a second-hand item; a recipe; a job ad, and the first post in a thread on a discussion forum (e.g., posing a question or describing a problem). The posting of objects is, on the one hand, an element of collaboration because users strive to achieve a common goal by publishing particular types of content (e.g., images) on a given site; on the other hand, it has features of competition, because users desire to draw other users’ attention to their objects. Users of all virtual communities and multi-sided platforms publish objects; therefore, this function cannot serve as a distinguishing feature and has not been included in the dimensions of the typology.
3.1.1 Collaboration among users
The first dimension of the typology is collaboration among users. Collaboration in online communities involves offering knowledge to others as well as adding to, recombining, modifying, and integrating knowledge that others have contributed [26]. Quantitative rating of objects and posting a comment that does not relate to other comments but the object itself are activities that create the content available to all but do not require adapting the message to comments by other users, hence they are ranked the lowest (Table 3.). The publication of a related comment requires greater alignment between the actions of the individual and the group. Finally, collaborative creation of objects demands synchronisation, cooperative decision making, and negotiating the rules of collaboration [65]. An example of this type of action is creating an entry on a multifaceted topic by a team of Wikipedians.
3.1.2 Reputation-based user competition
In virtual communities and on multi-sided platforms, users compete with one another by posting objects to promote their ideas or to interact with other users. These actions are common to both entities and cannot serve as a distinguishing feature. On some sites, users can pay for a more favorable placement of an object or an advertisement thereof. This competition occurs primarily in marketplaces and is largely absent from virtual communities and non-transactional multi-sided platforms.
Therefore, user competition has been reduced in this study to reputation-based user competition. As Dellarocas notes [18], reputation is here one’s past actions within the context of a specific community, presented in a manner that can help other community members make decisions with respect to whether and how to relate to that individual. A central reputation system is almost always essential in large communities where interactions among individual members rarely occur [18] and where there are hierarchical inequalities among users [64]. Reputation systems may rely on raw activity statistics, user feedback, and synthetic metrics [18, 85]. The proposed dimensions of reputation-based competition correlate with the indications above (cf. Table. 4).
Aside from the above levels, there exists also a stage of reputation-based competition in which users reap tangible benefits from their reputations. These can include more frequent displays of a user’s objects to other people, promoting their objects in popular areas of the site, or granting particular distinctions to users or their objects. These may be assigned by algorithmic detection or administrating operators [85]. However, this is extraordinarily difficult to identify without information sourced from inside the organisation in question. Therefore, this level has not been accounted for in the typology.
3.1.3 User multi-sidedness
An analysis of the literature devoted to multi-sided platforms and virtual communities reveals a multitude of approaches to user multi-sidedness. The term reflects the complementarity of functions that members of the participant groups of a given platform can play.
Many studies of communities and C2C platforms emphasize the equal status of users without mentioning their multi-sidedness, for example, equivalently positioned economic actors [54], peer-to-peer transactions [46], people who transact directly with one another [10], individuals who form special ties with other users and share information, ideas, and interests (Park 2018). Some authors accentuate the multitude of functions available to users, who may perform one or more roles in the systems [15, 34], participate in both the supply and demand sides [71], or both obtain and provide resources or services [24]. Other subject discussions simply mention two or more groups of users [31, 13, 29, 60]. The multi-sidedness of user groups is sometimes explained as the property of the entities they are, e.g., consumers and companies [49], buyers and sellers (Crittenden et al., 2017), external producers and consumers [52], autonomous complementors and consumers [34], demand- and supply-side participants [8, 76], independent contractors or professionals and clients [33], and suppliers and customers [1].
As is apparent from these lists, various streams of research emphasise either the homogeneity or heterogeneity of a group. Furthermore, users can be assigned to a particular group persistently (based on what a given entity is) or temporarily (depending on the function it currently performs). Table 5 presents the degrees of user multi-sidedness as a dimension of the typology. They have been distinguished based on various criteria, such as the entity’s characteristics, the function it performs, and whether the site is open or closed in nature.
3.2 Grouping the cases and analysing the empirical regularities.
In this stage, an analysis was conducted of 69 of the most popular Polish virtual communities and multi-sided platforms. These entities were selected based on web traffic rankings compiled in a study by Gemius/PBI, which quantified online traffic (site-centric measurements) and user behaviour (user-centric measurements). The study is the standard method of measuring online audiences in Poland, commonly used in planning advertisement campaigns (PBI, 2022). The analysed entities were selected from a list of approximately 1800 various websites with the highest number of Polish users in May 2019. Therefore, the entities in question are the most popular virtual communities or multi-sided platforms among Polish Internet users. The majority are domestic companies that operate in the consumer market. Companies were determined to serve a large number of users as they have developed stable functions that were the subject of assessment and classification in this study. This, however, is not without complications. Companies at this stage of development often comprise several business models. In such cases, we attempted to select and rate the dominant functions (e.g., in the case of the e-commerce platform Allegro, we recognised B2C rather than C2C sales as the primary function). If this was not possible, the company was excluded from the studied group (such was the case with Facebook). The analysed entities also excluded communities and platforms belonging to popular web portals. In Gemius/PBI studies, these sites are listed collectively without providing detailed data on individual entities belonging to a particular group.
Next, the selected entities were rated according to the three dimensions. The subject of coding was the presence of certain website functionalities. An analogical assessment was conducted after less than a year. The two sets of classifications were essentially congruous. The only difference was the assignment of one social media platform, which pertained to selecting its primary function, which was the object of assessment. A table with coding values is shown in Appendix 1.
3.3 Analysis of meaningful relationships and type construction
In theory, three dimensions of five grades each could result in 125 types. However, the analysis of empirical regularities and meaningful relationships (i.e. substantial adequacy of attribution of a website to an empirically founded group) led to distinguishing only four ideal types. This remains in line with Fiss’s approach to typology [27], according to which the complexity of the world should be pragmatically reduced to a limited set of ideal types. Each entity was assigned to one of the types, making the assignments mutually exclusive and exhaustive [6]. The distinguished types were then verified empirically, bearing in mind the suggestions made by Kluge [41] as well as Doty and Glick [22]. The empirical verification yielded corresponding results and is presented in appendix 2.
Four distinct types were identified using the three variables among the 69 entities analysed. They are problem communities, object communities, object markets, and reputation markets. Due to visual constraints, the location of the types in a three-dimensional space is presented in two tables (cf. Tables 6 and 7). In both, four distinct types emerge. One interpretation is that the typology could have been developed in two ways with fewer variables, and the resulting types would have been the same. The first way would be to consider two variables: collaboration among users and reputation-based user competition. The second way would be a typology based on one dimension: user multi-sidedness. This situation can be interpreted as confirming the correctness of the resulting types. It demonstrates that they could be distinguished using different sets of variables derived from the literature on the subject. Table 8 presents profiles of the identified business models.
3.4 Characterisation of constructed types
The distinguished types were described following the modified Amit and Zott’s [3] concept of the business model, i.e., a system (originally: a bundle) of specific activities that are conducted to satisfy the perceived needs of the market, including the specification of the parties that conduct these activities and how these activities are linked to each other. This understanding of a business model holistically depicts how a company does business, as opposed to what exactly it does or when or where it does it. This perspective emphasises the process of value creation over value capture (e.g., revenue model, pricing strategy).
The system dynamics methodology was used to create detailed descriptions of the business models. This approach enables the illustration of feedback loops in complex systems. System dynamics analyses are flexible and able to capture various interactions and their characteristics, and—as Morecroft states [47]—are particularly good for highlighting feedback loops that contribute to the dynamics. This methodology is commonly employed in studies of business models, strategy, or competition (Sterman 2010; Morecraft 2015; [12, 30], including multi-sided platforms [13, 62, 70, 80], and virtual communities [44]. As is typical of most articles about business models, we used the feedback system without specifying the power of individual effects in quantitative terms. Consequently, the method employed in this research aligns with the categories of qualitative system dynamics and feedback loop diagrams. Specifically, it allows us to present generic business models in the form of system archetypes, but precludes us from conducting simulations to ascertain the systems’ behaviours over time [83]. The resulting system dynamics models have successfully undergone the validation procedure for boundary adequacy and structure assessment (Sterman, 2010).
4 Description of identified business models
The following describes identified business models. They serve as ideal types based on median values and contain the typical characteristics of the companies belonging to a given category [45, 6]. For this reason, the description uses the singular form. The qualitative system dynamics method was used to describe the logic of value creation in each of the types. A brief description of examples accompanies each business models. The exemplary websites are located precisely as the identified business models on the three analysing dimensions.
4.1 Problem community
In simple terms, a problem community is a discussion forum or a set of thematically associated discussion threads. One user initiates a thread, and subsequent posts form a discussion. Users collaborate to a large degree by responding to previous posts by other users.
Content is typically displayed in chronological order, beginning with the most recent post. The community moderators block content that does not meet specific standards. However, they do not select content to be displayed in key parts of the site (except for community rules, frequently asked questions, etc.). The user reputation mechanisms rely on quantitative summaries of an individual’s activity within a given community (for example, the number of posts). Active users earn badges or status messages indicting their increased involvement. However, a more excellent reputation does not result in any tangible benefits. The problem community has the lowest degree of diversity in the functions performed by users. Starting a new thread or commenting on an existing one does not require access to different site functions or the fulfillment of other requirements (e.g., having a product for sale).
The simplified business model of the problem community can be described as comprising three main elements: users, content, and a value for users (cf. Fig. 1). The more users in a given community, the more content there is. This, in turn, reinforces the value for users. Consequently, new people join the community, raising the number of users and leading to further development.
Examples of a problem community are the websites: dojrzewamy.pl and stackoverflow.com. The first one is a community for teenagers (literally wearematuring.pl) in which they ask questions or describe their problems, which are later answered by their peers. The stackoverflow.com is a popular programmers’ community that operates similarly.
4.2 Object community
The next type identified in this research is the object community. In this community, users post objects and comment on them. Objects can include various types of content: memes, recipes, or homework assignments, for example.
In contrast to problem communities, comments are associated with particular objects rather than other comments. As a result, the comments together do not form a linear whole but are instead a set of independent statements about a particular object. There is less synchronisation and a lower degree of collaboration than in the problem community. Users build their reputations with systems presenting their aggregated activity in numerical values, status messages, or ranks.
The multi-sidedness of this community is somewhat higher than that of the problem-based community. Nevertheless, it is still a relatively homogeneous group of people who share an interest in a particular subject, post objects, and comment on them. This model displays greater user multi-sidedness, as creating and publishing an object (e.g., a movie) requires more effort and sometimes more excellent technical skill than simply commenting underneath it. Some less involved users or those with fewer skills perform the roles of commenters only.
The simplified business model of the object community resembles that of the problem community (cf. Fig. 2). The more users it has, the more content (objects and comments) it amasses, which in turn increases the value for users. However, this model also presents new factors and their attendant dependencies. In comparison to the previous community type, the role of the moderator also involves selecting objects and displaying them in essential parts of the community, such as the main page or the list of categories. Content curation thus resembles editor’s picks in the printed press and similarly increases value for users. Another new feedback loop is the widespread use of mechanisms that facilitate users' sharing of objects on social media. In this manner, informal communication encourages new people to visit the object-based community.
An example of an object community is the website kwejk.pl. This is a typical meme website. Users post their content, often of funny or shocking nature. The website community comments on them and rates them. The administrators pick up the content to be displayed in the major sections of the website (e.g., the homepage). The website gotujmy.pl (letscook.pl) is based on the akin logic. The users post here recipes which are later available to the website community.
4.3 Object market
The third identified type is the object market. It is a multi-sided platform on which one group of users publishes objects intended for another group. This is how classified ad, job, and dating sites operate.
The lowest level of collaboration among users occurs in this model: users do not rate objects. These services serve as classified ad boards that help connect both sides, rarely offering any additional functions. What is particularly noteworthy in this case is the third criterion, namely the user multi-sidedness. Interactions among users are marked by a contextual two-sidedness. Each entity performs a specific role in a particular transaction but may perform another function in a different transaction. For example, a user can buy one item on an auction platform and sell another.
In a simplified model of the object market, an increase in the number of sellers (an assumption made for simplicity; users do not always have to be sellers) leads to a larger number of objects (cf. Fig. 3). The more objects there are on a site, the more appealing it is and the more traffic it receives from buyers. This results in an increased number of transactions. On this site, transactions reduce the number of available objects as posts are removed. The company must place a greater emphasis on acquiring new objects when the sellers remove existing ones. This relationship is exclusive to this type. In the remaining three types, objects are persistent. An increase in the number of buyers also leads to growth in the number of sellers as they join the site. This ideal type accurately depicts the operations of a classified ad site but provides a less accurate description of job and dating sites. Nevertheless, it remains consistent with the principles of the typology. Individual real types can share the characteristics of the ideal type in varying degrees.
The website gratka.pl (bargain.pl) is an example of the object market. Users post classified advertisements here. The website community may access them; however, it cannot comment on them or rate them. The seller profiles predominantly present the list of their offers with scarce information about the seller. The other classified website is sprzedajemy.pl (wesell.pl), which follows the same pattern.
4.4 Reputation market
The final identified type is the reputation market. It is a multi-sided platform that typically connects buyers and sellers; however, unlike in object markets, the users’ roles are fixed and do not change across transactions, and the products offered are repeatable. This type of site features elaborate user feedback-based reputation systems. Examples of reputation markets include price comparison sites and aggregators for companies offering various services (e.g. tourism, insurance, travel, online-to-offline services).
Buyers collaborate by reviewing sellers with whom they have completed transactions or interacted in other ways. Out of all the business models distinguished in this research, reputation-based competition is most vital in this type. Sellers and their products are rated across many dimensions. This type also displays the highest level of user multi-sidedness: one group comprises sellers (or service providers), while the other comprises buyers. Notably, these roles are typically asymmetrical: buyers collaborate by rating sellers, while the latter compete based on their reputation.
The simplified business model of the reputation market resembles that of the object market, with two differences (cf. Fig. 4). An increase in the number of transactions does not result in a decreased number of objects, as these are being continuously offered (e.g., hotel rooms). There is a reputation system that allows users to rate sellers. This is a crucial element of the business model. The elaborate reputation system facilitates the purchase of products by reducing the risk involved. This increases the number of transactions made by existing users and encourages new buyers to join the platform. The elaborate reputation system also positively affects the loyalty of sellers. A company with a favourable rating based on reviews from thousands of buyers may find its reputation a barrier to exiting a given marketplace.
The ceneo.pl is the major Polish price comparison website and an example of the reputation market. It enables to find an online store offering a product at a low price and possessing an adequate reputation level. The website community heavily reviews both products and sellers. Similarly, the website niania.pl (nanny.pl) associates parents with babysitters. Its central section presents profiles of the babysitters who are reviewed in four dimensions, i.e.,—the quality of care, honesty, responsibility, and contact with children. Interestingly, niania.pl also utilises the business model of the object market. Parents may post an ad about hiring a babysitter, which is not ranked or reviewed. This section is, however, less popular.
5 Discussion
5.1 Diversity of value creation logic in distinguished business models
This article presents the common typology of virtual communities and multi-sided platforms. It resulted in four business models which present their main activities and specifications of these activities are linked to each other [3]. The four business models were depicted with qualitative system dynamics diagrams, therefore. Both typology and system dynamics models meet the criteria of being a theory as constructs are identified, relationships among these constructs are specified, and these relationships are falsifiable [22, 63]. In the case of typologies, constructs are the analysed dimensions and the ideal types. In the case of system dynamics, in turn, constructs are their elements. The relationships between the constructs were identified, discussed, and—in the case of the typology dimensions and ideal types—also verified quantitatively. Hence the presented typology as a system and the resulting business models may be treated as middle-range theories [22, 42], which attempt to generalise beyond a particular case but not beyond one set of circumstances [84].
The presented typology highlights the continuum of companies belonging to multi-sided platforms and virtual communities. This contribution to the stream of research on platform architecture, in conjunction with mechanism design theory [36], provides novel insights into how these companies operate and create value. Researchers can delve deeper into the intricacies of platform architecture by understanding the properties of resulted business models. This understanding is also beneficial for practitioners in choosing the suitable business model for their ventures and for regulators in ensuring a balanced and competitive market.
The four business models identified in the typology are problem community, object community, object market, and reputation market. There are similarities in value creation logics in all four types, as well as differences between them. Interestingly, the differences do not coincide with the division into virtual communities and multi-sided platforms.
Their common feature is the centrality of objects to their business models. Objects are crucial to their value propositions because they satisfy users’ needs and encourage them to return to the website. Objects also contribute to acquiring new users as they appear in search results or are shared on social media. Among the companies analysed in this research, however, almost no entities displayed high user engagement in collaboration within the scope of a single object (the exceptions being Wikipedia and its economics-oriented clone, mfiles.pl). This suggests that virtual communities and multi-sided platforms operating at large scales publish objects created by individual users rather than relying on intense collaboration among many people on one object. This observation may suggest that a collaborative society [37] consists more of cooperating individuals than of complex teams requiring coordination of activities.
In the three distinguished types, objects are long-term. After acquiring an object, one can earn money on it in the long term (by selling, displaying ads, etc.). The same cannot be said of object markets such as classified ad sites. An object, such as a used car ad, disappears after the transaction is finalised. Therefore, managing objects in an object market involves more than just accumulating objects and extracting permanent income from them, as is the case with an object community featuring recipes, for example. Managing objects in an object market necessitates the continuous acquisition of new objects not just for growth but also to compensate for lost objects.
Regardless of the objects they acquire, virtual communities and multi-sided platforms also compete based on user feedback in the form of object ratings or reviews. This is particularly apparent in the object community and reputation market, where it forms the basis of reputation systems. Systems of this type typically display quantitative scores and written reviews. It helps control the quality of its objects, for instance, by providing better exposure to ones with higher ratings. This conclusion contradicts Choudary's [15] observation that platforms, unlike pipes, do not control core value units' quality and quantity. The reviews can be rated by the marketplace's users and sorted in order of helpfulness. This shows that they become a valuable resource, especially when other companies offer similar objects, as is the case of e-commerce platforms.
The differentiation of the availability of product reviews between the object and reputation market makes sellers operate under different circumstances. In the object markets, sellers often lack established reputations, the loss of which could constitute switching costs, and therefore find it simpler to leave. In contrast, in reputation markets, sellers cherish their reputation based on consumer reviews, as it allows them to sell products, sometimes at higher prices. Reputation developed over a long period of time may, however, become an exit barrier for the seller, as it is not transferable to other places. Based on this reasoning, one could posit the hypothesis that the business model of the object market is less enduring than that of the reputation market, mainly when the latter has a base of sellers and products thoroughly reviewed by buyers.
Consequently, there are differences in the methods of object curation across identified business models. It is moderation of discussion in problem communities. Object curation in an object community relies on the selection and prominent display of a few objects, which may be conducted by an administrator (i.e., editor's pick) or by the algorithm. It is also the case in the object markets, where additionally the platform needs to source new objects to replace removed ones. In the reputation markets, objects are curated hugely based on user feedback.
Interesting differences are visible in terms of customer multi-sidedness and, more broadly, the network effect. The problem community and the reputation market are the extreme types. The first one constitutes a place for discussion and sometimes other forms of collaboration; it is dominated by the direct network effect. The latter is a highly specialized mechanism of matchmaking of demand and supply with an indirect network effect. Object community and the object market are examples of neighbouring types that share common properties, e.g., the user roles are complementary but may be subject to contextual changes. It is reasonable to ask whether the third type of network effect should not be introduced, which is an intermediate stage between the two extreme types—there are two separate functions of users here, however, users can assume a selected role depending on the context. Again, this division contradicts the traditional division into multisided platforms and virtual communities.
As shown above, there are significant differences between types within a category (e.g., in the aforementioned issue of object durability between two types of multisided platforms) and at the same time similarities between business models belonging to different categories (similar levels of multi-sidedness in the case of object community and object markets). This contradicts the principles of classification, according to which entities within one category should be similar to each other and as different as possible from entities in another category. In other words, the used types of business models should be internally homogeneous and externally heterogeneous. Multi-sided platforms and virtual communities are not such. Hence the need to look at these two business models from a different perspective. This is partly accomplished by using the term platforms with the indication that there are platforms with various levels of multi-sidedness or different types of network effect (Belleflamme & Peitz, 2021). This approach is not common in the literature, though, as the discourse on platforms is dominated by multi-sided ones.
5.2 Implications for research on business models classifications
The companies analysed in this study tend to cluster in particular places rather than being evenly spread throughout the typological space. This correlates with the approach adopted by Amit and Zott [3], which states that a business model depicts how a company does business, as opposed to what exactly it does or when or where it does it. This approach better reflects the business models of the companies analysed here than that of Casadesus-Masanell and Ricart [12], in which a business model is a system of choices and consequences. When viewed from the perspective of this study, business models appear to display a kind of determinism that is the opposite of individual decisions made by managers. If a company belongs to a particular type, it is characterized by specific functions (e.g., user rating systems in reputation markets). This comes as no surprise because a similar situation occurs in traditional business models: a grocery store, for example, has several consistent features, such as product displays, cash registers, etc. In this sense, the ideal types identified in this study can hardly be considered innovative. If the analysed companies and business models are innovative, it must therefore be the result of factors other than the general characteristics of the types of companies they are. Täuscher and Laudien [71] reach a similar conclusion in their taxonomy of marketplace businesses. They attribute the innovativeness of certain entities to a composition of several archetypal business models.
Similar observations can be made about competitiveness. If one assumes that the companies in a given sector employ the same generic business models, then competition can occur based on the individual elements of the business model (e.g., number of users), interactions between them, or emergent phenomena (e.g. marker leadership). Companies can thereby compete based on a selected element or, more broadly, based on their business model, in the systemic sense of the term. As we analysed companies during our research, we also observed business model-based competition in the typological meaning of the term. Some object communities featuring recipes or essays for students competed with providers who offered similar content but had created or purchased it themselves. Some reputation markets competed in this manner with online retailers. This situation resembles the competitive relationship between a local open-air market and nearby grocery stores.
The business models of companies attributed to the same types are similar, even if they operate in different markets. One possible way of explaining the similarities between business models can be found in the concept of a dominant design, which posits that variety decreases in the later stages of a product’s life cycle, and the product itself becomes standardised [74]. In this case, we apply the concept of dominant design to a business model rather than a product. The early years of the commercial Internet can be regarded as an era of ferment, with many alternative designs. Some business models discussed at the time no longer exist or have declined in popularity. One example of such a model is the e-mall, which aggregates online retailers belonging to a particular market [72]. In the following stage, the era of incremental change, a dominant design is a single architecture that establishes dominance in a category [4] and defines the design attributes that are widely accepted as meeting a users’ needs [73]. The business model that superseded thematic e-malls was the price comparison site, which allowed users to quickly search for products offered at low prices by reputable retailers. Price comparison sites are structured around their users’ current consumption needs, while e-malls listed offers from various stores in a given category.
The models identified in this study are generic. Internet companies operating at large scales often combine several business models. Facebook includes object community (users of the community post and comment on objects), problem community (discussion groups), object market (Facebook Marketplace), and to some degree, reputation market (fanpages vs. readers). This conclusion correlates with the approach adopted by Aversa et al. [5], who identified six business models within the Amazon.com ecosystem. The diversification of business models, meaning the inclusion of their different types, helps achieve demand-side complementarities. Approaching the issue of the diversification of business models from the system dynamics perspective, one could mention another goal, namely the creation and reinforcement of feedback loops that are beneficial to the company [12].
It is also worth paying attention to the companies mentioned in the introduction. Couchsurfing is an object community of people sharing a place to stay overnight. The object here is the presentation of the person and their home. Given that the user can act in two roles as a guest and host, he or she is rated by the people who have hosted or are hosting the person. Booking.com is a reputation market showcasing hotel listings available continuously; these are heavily reviewed by guests, influencing how they are displayed on the site. Airbnb works in much the same way as Booking.com. There are differences, however, in the short-term listings associated with subletting a flat for the duration of a holiday [32]. These listings are not offered continuously, disappear after booking, and often have no reviews. In this respect, Airbnb operates as an object market. This example shows that the real types in practice can combine the characteristics of the ideal types distinguished in the typology.
6 Conclusions, further research, and limitations
A vital contribution of this study is the finding that the traditional division into virtual communities and multi-sided platforms is inadequate due to significant differences within categories and similarities across them, which contradicts the principles of classification. The presented typology required the introduction of new concepts like the object as a unit of content and its three dimensions. The four distinguished business models, their system dynamics models and the whole typology can be considered middle-range theories, aiming to generalise beyond specific cases within a set of circumstances.
The resulting business model archetypes might constitute a basis for further research. They are new research problems stemming from the identified business model. The first one is the revenue model, especially how the methods of generating income differ among the four types. Another potential study area could explore strategies formulated concerning customers and competitors, such as competing based on the quality or quantity of objects [69]. The question of the extent to which the analysed types contribute to category monopolisation and disruption of adjacent markets is undoubtedly of great importance. Finally, do some of these types occur together, i.e. are they often used simultaneously by one company? Their coexistence has been observed in multi-sided platforms encompassing problem community functions.
Another avenue of research would be to describe the identified types using a method other than system dynamics. Examples of such methods could include the Business Model Canvas [50] and their platform-related modifications [21], as well as Amit and Zott’s approach involving four sources of value creation in e-business [2]. Given how different these approaches are, using them to identify the characteristics of the business models distinguished in this study could produce novel results.
A discussion of the findings must also mention their limitations. The main assumptions, particularly the typology dimensions, were chosen based on a subjective view of the current knowledge of virtual communities and multi-sided platforms. For this reason, the dimensions were consulted with academic researchers and practitioners. The selection of the major function to be assessed in analysed entities might have been, in some cases, marked by researcher bias.
The entities analysed here were among the sites most frequently visited by Polish Internet users. Aside from a few global firms, most are domestic companies. The Polish Internet market is a competitive one, as witnessed by the fact that many global players have attempted to enter it, only to make a hasty retreat. However, we did not perceive the significant impact the unique characteristics of the Polish market have on the simplified business models of the identified types, at least when compared to companies operating in the EU and the United States.
References
Alt, R., & Zimmermann, H. D. (2014). Editorial 24/4: Electronic markets and business models.
Amit, R., & Zott, C. (2001). Value creation in e-business. Strategic Management Journal, 22(6–7), 493–520.
Amit, R., & Zott, C. (2010). Business model innovation: Creating value in times of change (No. D/870). IESE Business School.
Anderson, P., & Tushman, M. L. (1990). Technological discontinuities and dominant designs: A cyclical model of technological change. Administrative science quarterly, 604–633.
Aversa, P., Haefliger, S., Hueller, F., & Reza, D. (2020). Customer complementarity in the digital space: Exploring Amazon’s business model diversification. Long Range Planning, 101985.
Bailey, K. D. (1994). Typologies and taxonomies: An introduction to classification techniques (No. 102). Sage.
Barrett, M., Oborn, E., & Orlikowski, W. (2016). Creating value in online communities: The sociomaterial configuring of strategy, platform, and stakeholder engagement. Information Systems Research, 27(4), 704–723.
Bazarhanova, A., Yli-Huumo, J., & Smolander, K. (2020). From platform dominance to weakened ownership: How external regulation changed Finnish e-identification. Electronic Markets, 30(3), 525–538.
Blaschke, M., Haki, K., Aier, S., & Winter, R. (2019). Taxonomy of digital platforms: A platform architecture perspective. 14th international conference on Wirtschaftsinformatik, 572–586.
Caldieraro, F., Zhang, J. Z., Cunha, M., Jr., & Shulman, J. D. (2018). Strategic information transmission in peer-to-peer lending markets. Journal of Marketing, 82(2), 42–63.
Canzer, B. (2006). E-business: Strategic thinking and practice. Houghton Mifflin.
Casadesus-Masanell, R., & Ricart, J. E. (2011). How to design a winning business model. Harvard Business Review 89(1/2), 100–107.
Casey T., & Töyli J. (2012) Dynamics of two-sided platform success and failure: An analysis of public wireless local area access. Technovation, 31(12), 703–716.
Cennamo, C. (2021). Competing in digital markets: A platform-based perspective. Academy of Management Perspectives, 35(2), 265–291.
Choudary, S. (2015). Platform scale: how an emerging business model helps startups build large empires with minimum investment. [Place of publication not identified] : Platform Thinking Labs.
Collier, D., Laporte, J., & Seawright, J. (2012). Putting typologies to work. Political Research Quarterly, 65(1), 217–232. https://doi.org/10.1177/1065912912437162
Cusumano, M., Gawer, A., & Yoffie, D. (2019). The business of platforms : Strategy in the age of digital competition, innovation, and power. HarperCollins Publishers.
Dellarocas, C. (2010). Online reputation systems: How to design one that does what you need. MIT Sloan Management Review, 51(3), 33.
Derave, T., Prince Sales, T., Gailly, F., & Poels, G. (2021, June). Comparing digital platform types in the platform economy. In International Conference on advanced information systems engineering (pp. 417–431). Springer, Cham.
Doligalski, T. (2018). Internet Business Models in the Consumer Market—A Typological Approach. Marketing i Rynek, 12, 5–13. Retrieved March 29, 2023 from https://www.doligalski.net/internet-business-models/
Doligalski, T. (2021). Platform canvas: Does the platform business model imply disruption and monopolisation? In T. Doligalski, M. Goliński, & K. Kozłowski (Eds.), Disruptive platforms (pp. 1–17). Routledge.
Doty, D. H., & Glick, W. H. (1994). Typologies as a unique form of theory building: Toward improved understanding and modeling. The Academy of Management Review, 19(2), 230. https://doi.org/10.2307/258704
Eisend, M., & Kuss, A. (2019). Research methodology in marketing: Theory development. Springer.
Ertz, M., Durif, F., & Arcand, M. (2016). Collaborative consumption: Conceptual snapshot at a buzzword. Journal of Entrepreneurship Education, 19(2), 1–23.
Evans, D. (2003). Some empirical aspects of multi-sided platform industries. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.447981
Faraj, S., Jarvenpaa, S. L., & Majchrzak, A. (2011). Knowledge collaboration in online communities. Organization Science, 22(5), 1224–1239. https://doi.org/10.1287/orsc.1100.0614
Fiss, P. C. (2011). Building better causal theories: A fuzzy set approach to typologies in organization research. Academy of management journal, 54(2), 393–420.
Filistrucchi, L., Geradin, D., & Damme, E. (2012). Identifying two-sided markets. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.2008661
Fürstenau, D., Auschra, C., Klein, S., & Gersch, M. (2019). A process perspective on platform design and management: Evidence from a digital platform in health care. Electronic Markets, 29(4), 581–596.
Gary, M. S., Kunc, M., Morecroft, J. D., & Rockart, S. F. (2008). System dynamics and strategy. System Dynamics Review: The Journal of the System Dynamics Society, 24(4), 407–429.
Gawer, A. (2014). Bridging differing perspectives on technological platforms: Toward an integrative framework. Research Policy, 43(7), 1239–1249.
Gyódi, K. (2019). Airbnb in European cities: Business as usual or true sharing economy? Journal of Cleaner Production, 221, 536–551.
Hagiu, A., & Wright, J. (2015). Multi-sided platforms. International Journal of Industrial Organization, 43, 162–174.
Hein, A., Schreieck, M., Riasanow, T., Setzke, D., Wiesche, M., Böhm, M., & Krcmar, H. (2019). Digital platform ecosystems. Electronic Markets, 30(1), 87–98.
Holland, C., & Gutiérrez-Leefmans, M. (2018). A Taxonomy of SME E-Commerce Platforms Derived from a Market-Level Analysis. International Journal of Electronic Commerce, 22(2), 161–201.
Hurwicz, L., & Reiter, S. (2006). Designing economic mechanisms. Cambridge University Press.
Jemielniak, D., & Przegalińska, A. (2020). Collaborative society. MIT Press.
Kaplan, A., & Haenlein, M. (2010). Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons, 53(1), 59–68.
Kaplan, S., & Sawhney, M. (2000). E- hubs: The new B2B marketplaces. Harvard Business Review, 78(3), 97–103.
Kim, J., & Min, J. (2019). Supplier, tailor, and facilitator: Typology of platform business models. Journal of Open Innovation: Technology, Market, and Complexity, 5(3), 57.
Kluge, S. (2000). Empirically grounded construction of types and typologies in qualitative social research. Qualitative Social Research, 1(1).
Kopainsky, B., & Luna-Reyes, L. F. (2008). Closing the loop: Promoting synergies with other theory building approaches to improve system dynamics practice. Systems Research and Behavioral Science: The Official Journal of the International Federation for Systems Research, 25(4), 471–486.
Levina, N., & Arriaga, M. (2014). Distinction and status production on user-generated content platforms: Using Bourdieu’s theory of cultural production to understand social dynamics in online fields. Information Systems Research, 25(3), 468–488.
Mao, Y., Vassileva, J., & Grassmann, W. (2007, January). A system dynamics approach to study virtual communities. In: 2007 40th Annual Hawaii International Conference on System Sciences (HICSS'07) (pp. 178a-178a). IEEE.
McKinney, J. C. (1969). Typification, typologies, and sociological theory. Social forces, 48(1), 1–12.
Mittendorf, C. (2018). Collaborative consumption: The role of familiarity and trust among millennials. Journal of Consumer Marketing.
Morecroft, J. D. (2015). Strategic modelling and business dynamics: A feedback systems approach. Wiley.
Muegge, S. (2013). Platforms, communities, and business ecosystems: lessons learned about technology entrepreneurship in an interconnected world. Technology Innovation Management Review, 3(2), 5–15.
Oliveira, D., & Cortimiglia, M. (2017). Value co-creation in web-based multi-sided platforms: A conceptual framework and implications for business model design. Business Horizons, 60(6), 747–758.
Osterwalder, A., & Pigneur, Y. (2010). Business model generation: A handbook for visionaries, game changers, and challengers. Wiley
Panek, T. (2009). Statystyczne metody wielowymiarowej analizy porównawczej. Szkoła Główna Handlowa-Oficyna Wydawnicza.
Parker, G., Alstyne, M., & Choudary, S. (2016). Platform revolution. W. W. Norton & Company.
Perscheid, G., Ostern, N. K., & Moormann, J. (2020). Determining platform governance: Framework for classifying governance types. In ICIS.
Perren, R., & Kozinets, R. (2018). Lateral exchange markets: How social platforms operate in a networked economy. Journal of Marketing, 82(1), 20–36.
Porter, C. (2006). A typology of virtual communities: a multi-disciplinary foundation for future research. Journal of Computer-Mediated Communication, 10(1), doi: https://doi.org/10.1111/j.1083-6101.2004.tb00228.x
Pousttchi, K., & Gleiss, A. (2019). Surrounded by middlemen-how multi-sided platforms change the insurance industry. Electronic Markets, 29(4), 609–629.
Prahalad, C. K., & Ramaswamy, V. (2004). The future of competition: Co-creating unique value with customers. Harvard Business Review Press.
Rappa, P. (n.a.). Business models on the web: Professor Michael Rappa. Retrieved November 02, 2020, from http://digitalenterprise.org/models/models.html
Reimers, K., Guo, X., & Li, M. (2018). Beyond markets, hierarchies, and hybrids: An institutional perspective on IT-enabled two-sided markets. Electronic Markets, 29(2), 287–305.
Rochet, J. C., & Tirole, J. (2006). Two-sided markets: A progress report. The RAND journal of economics, 37(3), 645–667.
Rodríguez-López, N., & Diz-Comesaña, M. (2016). Toward a taxonomy of virtual communities from a value cocreation perspective. Psychology & Marketing, 33(12), 1166–1173.
Ruutu, S., Casey, T., & Kotovirta, V. (2017). Development and competition of digital service platforms: A system dynamics approach. Technological Forecasting and Social Change, 117, 119–130.
Schwaninger, M., & Grösser, S. (2008). System dynamics as model-based theory building. Systems Research and Behavioral Science: The Official Journal of the International Federation for Systems Research, 25(4), 447–465.
Sibai, O., Valck, K., Farrell, A., & Rudd, J. (2015). Social control in online communities of consumption: A framework for community management. Psychology & Marketing, 32(3), 250–264.
Spagnoletti, P., Resca, A., & Lee, G. (2015). A design theory for digital platforms supporting online communities: A multiple case study. Journal of Information Technology, 30(4), 364–380.
Srinivasan, R., Lilien, G., & Rangaswamy, A. (2004). First in, first out? The effects of network externalities on pioneer survival. Journal of Marketing, 68(1), 41–58.
Staub, N., Haki, K., Aier, S., & Winter, R. (2021). Taxonomy of digital platforms: A business model perspective. In Proceedings of the 54th Hawaii International Conference on System Sciences (p. 6163).
Trabucchi, D., & Buganza, T. (2021). Landlords with no lands: a systematic literature review on hybrid multi-sided platforms and platform thinking. European Journal of Innovation Management.
Trabucchi, D., Buganza, T., & Verganti, R. (2021). Quantity or quality? Value creation in two-sided platforms. Technology Analysis & Strategic Management, 33(2), 162–175.
Täuscher, K., & Abdelkafi, N. (2018). Scalability and robustness of business models for sustainability: A simulation experiment. Journal of Cleaner Production, 170, 654–664.
Täuscher, K., & Laudien, S. (2018). Understanding platform business models: A mixed methods study of marketplaces. European Management Journal, 36(3), 319–329.
Timmers, P. (1998). Business models for electronic markets. Electronic Markets, 8(2), 3–8. https://doi.org/10.1080/10196789800000016
Tiwana, A. (2013). Platform ecosystems: Aligning architecture, governance, and strategy. Newnes.
Utterback, J. M., & Abernathy, W. J. (1975). A dynamic model of process and product innovation. 1975, 3(6), 639–656.
Van Alstyne, M. W., Parker, G. G., & Choudary, S. P. (2016). Pipelines, platforms, and the new rules of strategy. Harvard business review, 94(4), 54–62.
Veisdal, J. (2020). The dynamics of entry for digital platforms in two-sided markets: A multi-case study. Electronic Markets, 30(3), 539–556.
Walesiak, M., & Dudek, A. (2010). Finding groups in ordinal data: an examination of some clustering procedures. In Classification as a Tool for Research (pp. 185–192). Springer, Berlin, Heidelberg.
Walesiak, M., & Gatnar, E. (Eds.). (2009). Statystyczna analiza danych z wykorzystaniem programu R. Wydawnictwo Naukowe PWN.
Walesiak, M. (1999). Distance Measure For Ordinal Data. Argumenta Oeconomica, 8(2), 167–173.
Wang, J., & Lai, J. Y. (2020). Exploring innovation diffusion of two-sided mobile payment platforms: A system dynamics approach. Technological Forecasting and Social Change, 157, 120088.
Weill, P., & Vitale, R. (2001). Place to space: Migrating to ebusiness models. Harvard Business School Press.
Wirtz, J., So, K., Mody, M., Liu, S., & Chun, H. (2019). Platforms in the peer-to-peer sharing economy. Journal of Service Management, 30(4), 452–483.
Wolstenholme, E. F. (1999). Qualitative vs quantitative modelling: The evolving balance. Journal of the Operational Research Society, 50(4), 422–428.
Woodside, A. G. (2003). Middle‐range theory construction of the dynamics of organizational marketing‐buying behavior. Journal of Business & Industrial Marketing.
Ziaie, P., & Krcmar, H. (2012). Introducing a design framework for reputation systems in multi-tier production communities. Digital Creativity, 23(2), 144–157.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix 1. Assignment of analysed sites to typology dimensions
Collaboration among users | Reputation-based user competition | User multi-sidedness | Ideal types | |
---|---|---|---|---|
wykop.pl | 4 | 3 | 3 | 1 |
dojrzewamy.pl | 4 | 3 | 2 | 1 |
stackoverflow.com | 4 | 3 | 2 | 1 |
vinted.pl | 2 | 4 | 4 | 3 |
sadistic.pl | 3 | 3 | 3 | 2 |
salon24.pl | 4 | 3 | 2 | 2 |
cda.pl | 3 | 3 | 3 | 2 |
tekstowo.pl | 2 | 3 | 3 | 2 |
sciaga.pl | 3 | 3 | 3 | 2 |
smaker.pl | 3 | 3 | 3 | 2 |
znanylekarz.pl | 3 | 5 | 5 | 4 |
przyslijprzepis.pl | 3 | 3 | 3 | 2 |
chomikuj.pl | 3 | 3 | 3 | 2 |
lubimyczytac.pl | 3 | 3 | 3 | 2 |
bryk.pl | 3 | 3 | 4 | 2 |
demotywatory.pl | 3 | 3 | 3 | 2 |
pyszne.pl | 3 | 4 | 5 | 4 |
ulub.pl | 1 | 3 | 3 | 2 |
pomagam.pl | 3 | 1 | 4 | 3 |
kwejk.pl | 3 | 3 | 3 | 2 |
gotujmy.pl | 3 | 3 | 3 | 2 |
opineo.pl | 3 | 5 | 5 | 4 |
wiocha.pl | 3 | 3 | 3 | 2 |
zaliczaj.pl | 3 | 3 | 3 | 2 |
matematyka.pl | 4 | 3 | 3 | 1 |
anonimowe.pl | 3 | 3 | 3 | 2 |
chamsko.pl | 3 | 3 | 3 | 2 |
pomocedomowe.pl | 3 | 5 | 5 | 4 |
piekielni.pl | 4 | 3 | 3 | 2 |
przepisy.net | 3 | 3 | 3 | 2 |
brainly.pl | 3 | 3 | 3 | 2 |
zrzutka.pl | 3 | 3 | 4 | 3 |
goldenline.pl | 4 | 2 | 1 | 1 |
nk.pl | 4 | 2 | 1 | 1 |
forumprawne.pl | 4 | 1 | 3 | 1 |
forumogrodnicze.pl | 4 | 3 | 3 | 1 |
mfiles.pl | 5 | 2 | 3 | 2 |
forumsamochodowe.pl | 4 | 3 | 3 | 1 |
patronite.pl | 1 | 3 | 4 | 3 |
zszywka.pl | 2 | 3 | 3 | 2 |
biblionetka.pl | 3 | 3 | 3 | 2 |
temysli.pl | 3 | 3 | 3 | 2 |
rankinglekarzy.pl | 3 | 5 | 5 | 4 |
mojabudowa.pl | 3 | 3 | 3 | 2 |
filmweb.pl | 4 | 3 | 3 | 1 |
fotka.pl | 3 | 4 | 3 | 2 |
allegro.pl | 3 | 5 | 5 | 4 |
ceneo.pl | 3 | 5 | 5 | 4 |
otomoto.pl | 1 | 3 | 5 | 3 |
skapiec.pl | 3 | 5 | 5 | 4 |
nocowanie.pl | 3 | 5 | 5 | 4 |
e-turysta.pl | 3 | 5 | 5 | 4 |
niania.pl | 3 | 5 | 5 | 4 |
olx.pl | 1 | 3 | 4 | 3 |
sprzedajemy.pl | 1 | 3 | 4 | 3 |
otodom.pl | 1 | 3 | 5 | 3 |
pracuj.pl | 1 | 3 | 5 | 3 |
gratka.pl | 1 | 3 | 4 | 3 |
praca.pl | 1 | 3 | 5 | 3 |
oglaszamy24.pl | 1 | 1 | 4 | 3 |
szafa.pl | 1 | 4 | 4 | 3 |
oferia.pl | 3 | 5 | 5 | 4 |
tabelaofert.pl | 1 | 3 | 5 | 3 |
youtube.com | 3 | 3 | 3 | 2 |
wikipedia.org | 5 | 3 | 3 | 2 |
instagram.com | 3 | 3 | 3 | 2 |
alieexpres.com | 3 | 5 | 5 | 4 |
twitter.com | 4 | 2 | 2 | 2 |
linked-in.com | 4 | 2 | 2 | 1 |
Appendix 2. Quantitative verification of theoretical attributions
The correctness of the attribution of the analysed entities to the types distinguished in the typology (from now on referred to as theoretical attribution) was verified by comparing it to a classification created using a quantitative method. This method involved identifying groups characterised by the greatest degree of internal homogeneity and the greatest degree of external heterogeneity.
The quantitative classification was conducted with the help of the general distance measure, which is used to group objects based on ordinal variables [79]. Four groups were identified using hierarchical agglomerative clustering with the average linkage method. The analysis was performed with the help of the clusterSim package available in the R programming environment.
In this classification, the three dimensions of the typology were used as the axes of a multidimensional space. Virtual communities and multi-sided platforms were graded on a scale of 1–5, with the higher values representing a greater intensity of a particular feature. Due to using a weak and uniform scale of measurement, the variables were not normalised. The ratings assigned to the features are represented by points in the three-dimensional space. By comparing the distance between points using the clustering method, we can identify proximate objects [78].
The agglomerative method was used to determine the groups. This method produces a dendrogram that depicts a hierarchy and a method of linking the objects into groups with the greatest possible degree of homogeneity [51]. The four identified groups form separate branches on the dendrogram. The average linkage method was used to group the objects, achieving a high validation score measured with the adjusted Rand index [77].
The clustering results are depicted in the form of a dendrogram (cf. Fig.
5) and basic structure measures (Table 9). The dendrogram illustrates the connections and hierarchies between the analysed entities. Cutting the tree at the dotted line produces four groups of entities of the following sizes: n1 = 16, n2 = 27, n3 = 13, and n4 = 13.
Structure analysis of individual rating dimensions enables the characterisation of the distinguished groups (Table 9). The first group is marked by a high degree of collaboration among users and low mean values of reputation-based competition and user multi-sidedness. Notably, the collaboration value displays a small range, indicating a high concentration level for this feature. Entities with low collaboration values do not occur in this group. The opposite is true of reputation and multi-sidedness, which occur at varying levels but are concentrated at the bottom end of the range of grades (3 and lower). This group corresponds to the type defined as a problem community. Except for the dimension depicting user multi-sidedness, the median values of individual features assigned to the group using the method correspond to the ideal type. Five websites were classified differently. These entities were assigned to the second type in theoretical attribution, defined as an object community. The greatest differences between the quantitative and theoretical attribution in the first group were observed.
The second group is the most numerous and is characterised by an average rating (approximately 3) in each dimension. A high concentration can be observed around the mean value, with collaboration receiving the lowest ratings, some of which fall below 3. The median values for individual features correspond perfectly to the ideal type of object community. One entity, a crowdfunding platform, was classified differently in the theoretical attribution, where it was assigned to the object market rather than the object community.
The third group is marked by a very low collaboration value, average reputation-based user competition levels, and large average multi-sidedness. The range indicates a large divergence of grades in the case of competition, indicating a lack of full homogeneity in this group in this regard. As in the second group, complete correspondence is observed between the median values of individual features and the conceptually depicted ideal type. All observations assigned to the third group also belonged to the type defined as the object market.
The fourth group encompasses entities with average collaboration values but very high levels of reputation-based competition and multi-sidedness, with a bias towards the latter, where all entities scored 5 (with a range of 0). Complete concentration is also observed in customer collaboration, indicating high conformity of scores in this dimension. Furthermore, the fourth group displayed complete correspondence between quantitative and conceptual clustering. The median value of individual features produced by clustering is identical to that of the ideal type defined as the reputation market.
The data presented in Table 9 reveals a high level of conformity between the median values of features assigned to groups using quantitative clustering and the characteristics of ideal types. One exception is user multi-sidedness in problem communities. The last two groups display high conformity with the size of the theoretically assigned groups. Greater differences occur in the first group, which was assigned more entities in the statistical clustering than in the theoretical attribution. Subsequent analysis reveals the validity of the latter.
In general, clustering using either technique produced convergent results. Among the 69 entities classified with the help of the quantitative method, 63 were assigned to the same group as in the theoretical attribution, indicating 91% conformity between the two analysis methods.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Doligalski, T. Common typology of multi-sided platforms and virtual communities: analysis of business models using qualitative system dynamics. Electron Commer Res (2023). https://doi.org/10.1007/s10660-023-09700-w
Accepted:
Published:
DOI: https://doi.org/10.1007/s10660-023-09700-w