The concept of platform economy stems from the theory of multi-sided markets when a platform serves various user groups and there is a dependency of the demands between the users. Rochet and Tirole ([24]) elaborated the framework showing that digital platforms incorporate indirect network externalities between different market participants. Thus, a marketplace platform connects sellers and buyers, a ride-sharing platform connects passengers and drivers, while a search engine connects users with advertisers. Unlike the traditional firms, the platform does not usually own the traded goods/services and provides only the digital infrastructure to ensure efficient interactions between the parties. Parker et al. ([23]) describe a platform as the business model that creates value through exchanges between independent producers and consumers. The platform provides technological infrastructure, defines governance rules and decreases search, coordination, and transaction costs. Similarly, Evans and Schmalensee ([11]) claim that digital platforms work as “matchmakers”, coordinating interactions between different parties and creating value by means of participation in the network rather than direct production.

One of the implications of this business model is the transition from pipeline competition to ecosystem competition. While traditional firms compete through production efficiency and differentiation of the produced goods/services, the platform competes by attracting a greater number of users, fostering complementary innovation, and creating a network of interconnected ecosystems of the products and services ([7]). Thus, the platform competition relies not only on increasing production capabilities, but also on the coordination of several market participants.

Digital platforms have developed into different organizational forms with various economic functions. Despite the existence of great overlap between the categories, the most typical digital platforms may be classified into six major groups.

Search platform: The search platform collects and provides digital information using advanced search algorithms. The most famous representatives are Google and Microsoft’s Bing that earn revenues mainly due to digital advertising based on the user data.

E-commerce: The e-commerce platform allows for trading between buyers and sellers providing them with the digital market, payment system, logistics infrastructure, and reputation mechanism. Amazon, Alibaba, eBay, and Flipkart are the typical examples of this category.

Social media: The social media platform provides the possibility to create, distribute, and consume digital content and maintain online social networks. The platforms like Facebook, Instagram, TikTok, LinkedIn, and X rely greatly on user-generated content and targeted advertising.

Application platform: The application platform connects the software developer and the consumer through the centralized digital marketplace. Thus, the examples of such platforms are Apple’s App Store and Google Play that provide application distribution, payment processing, software verification, and ecosystem governance.

Sharing economy: The sharing economy platform coordinates decentralized service provision by matching the independent provider with the consumer. Thus, Uber, Lyft, Airbnb, DoorDash, Ola, and Upwork are the examples of how digital platforms provide transportation, accommodations, food delivery, and freelancing services based on algorithmic matching.

Digital payment: The digital payment platform allows for secure electronic financial transactions between consumers, merchants, and financial institutions. Thus, PayPal, Apple Pay, Google Pay, Alipay, PhonePe, and Paytm are the examples of how platforms provide the electronic payment solutions. Moreover, many digital payment platforms have evolved to include the broader financial ecosystems incorporating lending, investment, and insurance services ([20]; [8]).

Nowadays, the large technology companies operate several types of platforms simultaneously. Amazon operates an ecosystem integrating e-commerce, cloud computing, digital advertising, logistics, streaming services, and artificial intelligence, while Google incorporates search, Android, YouTube, cloud computing, advertising, and AI services. The architecture of the digital platform ecosystem is depicted in Figure 1 below. This explains the way in which the digital platform owner provides the infrastructure that makes interaction possible between the consumers, firms, developers, advertisers, and other service providers. The interactions provide data, which are analyzed through the application of artificial intelligence to improve the platform services, and hence create a reinforcing feedback loop for the platform.

Figure 1. Conceptual architecture of a digital platform ecosystem showing interactions among platform participants, data generation, AI-enabled analytics, and continuous value creation ([23]; [11]; [15]).

It is necessary to differentiate among various types of organizational forms of digital platforms due to the fact that their approaches to generating value and exercising market power can be quite different. Transactional platforms, such as e-commerce platforms and sharing economy marketplaces, are mostly used for enabling transactions between buyers and sellers or consumers and service providers, so that their value creation process relies on indirect network effects, efficiency of matching and volume of transactions. App ecosystems, like Apple’s App Store or Google Play, serve as intermediaries between users and developers, controlling such processes as application distribution, payments processing and some technical requirements of applications. Their operation and market power rely on governance of ecosystems, switching costs and interoperability. Digital infrastructures provide such digital infrastructures as cloud computing to other companies and organizations, thus being influenced by economies of scale and technological dependence. The vertically integrated platform business model, which includes the combination of activities related to search, advertising, operating system, cloud computing, hardware, payments, or any other complementary activity within an established company, can leverage competitive advantages in multiple interdependent markets through economies of scope, data aggregation, product bundling, and ecosystem integration. Thus, the conclusions reached in this review do not suggest that market power in platform business models is created through the same factors for all platforms.

There are several economic principles that drive the development of digital platforms and distinguish platform markets from other traditional industries.

1)Network Effects

Network effects represent the primary principle of platform economies. Katz and Shapiro showed that the value of a good or service rises with an increasing number of people using it ([19]). In case of digital platforms, there may be both direct network effects where users benefit from interacting with bigger communities, and indirect ones where a higher level of activity on one platform side encourages people on the other side to join it. For example, a greater number of customers attracts new suppliers on the Amazon platform, but a higher number of suppliers leads to greater product variety and more buyers attracted ([24]). Such reinforcing loops usually result in “winner-takes-most” market structures accompanied by strong concentration.

2)Economies of Scale

There are high economies of scale in digital platforms because of large investment in the development of the software, infrastructure, cybersecurity, and AI, but the marginal cost of providing services for extra users remains relatively low. When the number of users grows, the average cost drops quickly, and allows successful platforms to scale globally with minimal additional costs ([23]). Such cost structure largely helps the emergence of dominant global platforms.

3)Economies of Scope

In addition to economies of scale, digital platforms benefit from economies of scope, since infrastructure, algorithms, cloud resources, customer relationships, and datasets can be used in several related markets. This makes it possible to combine in unified ecosystems such different services as search engines, cloud computing, digital advertising, AI, digital payments, and e-commerce, and thus to increase competitive advantage and customer loyalty ([7]).

4)Data as an Economic Resource

The data became some of the most valuable assets of the platform economies. Every interaction with a digital platform—searches, purchases, clicks, reviews, and location data—adds up to the proprietary database of the platform owner. Such data improve machine learning algorithms used in recommendation systems, advertising, demand forecasts, logistics, fraud detection, and personalization ([29]). The resulting data-AI feedback loop makes it possible to constantly improve the quality of services provided and attract new users, thus collecting new data and growing the platform even further.

5)Reduction of Transaction Costs

The reduction of transaction costs through better information provision, search reduction, easier contracts formation, electronic payments, and creation of reputation systems is another distinctive feature of digital platforms. It corresponds to the theory of the firm by Coase ([3]), according to which firms exist because they reduce transaction costs more efficiently than markets. According to transaction-cost economics by Williamson, the use of digital technologies improves coordination and reduces uncertainties and asymmetries of information ([31]). Therefore, platform economies often become more efficient than traditional market structures.

The combination of network effects, economies of scale, scope, data accumulation, and transaction cost reduction explains the fast development of platform-based business models in the global economy. At the same time, such economic characteristics create the structural conditions for market concentration and competitive advantages. In order to analyze the mechanisms by which digital platforms gain and maintain their market power, it is necessary to understand the economic foundations presented above.

The most essential source of market power in platform-based economies is network effects. They happen when value generated by a platform rises as more users become part of the network. The concept of network externalities was first described by Katz and Shapiro ([19]), who showed that the rising number of users makes the product or service more attractive for potential users as well. Later, Rochet and Tirole ([24]) demonstrated how the concept of network externalities can be extended to two-sided markets, where platforms create value for multiple groups of users at once by internalizing the indirect network effects. Network effects can be divided into two large categories. Direct network effects appear when users get more utility from interacting with a large group of users. Thus, social networks such as Facebook, LinkedIn, and WhatsApp become more useful when more people use them, as communication possibilities rise.

Indirect network effects, on the other hand, happen when the growing number of users on one side of the platform increases value for the users on another side. For example, the growing number of sellers on Amazon adds more variety of products, attracting more buyers, which, in its turn, creates incentives for more sellers to join the platform. The same feedback loop works for app stores, online marketplaces, and digital payment platforms ([24]). Strong network effects often result in self-reinforced growth, creating very concentrated markets characterized by “winner-takes-most” dynamics. After reaching the critical mass, competing platforms often fail to attract enough users, making incumbent’s position even stronger.

Data has become one of the most valuable strategic assets in the digital environment. Every user action including searching for information online, buying goods or services, writing reviews, using social networks, providing geo-coordinates, etc. creates data, which can be analyzed and used to improve the operation of the platform ([29]). Moreover, the competitive importance of data is augmented by artificial intelligence. The more data algorithms have, the better they perform. More accurate prediction means that more users will use a platform. In its turn, the growing number of users generates even more data, allowing platforms to improve their algorithms and performance further. Thus, data and AI create feedback loop resulting in cumulative competitive advantages that are almost impossible for the new entrants to replicate ([2]). Since access to the large-scale proprietary data cannot be simply bought or created by other platforms, established platforms gain substantial informational advantages over the new ones. Therefore, data accumulation has become one of the main sources of persistent market power in the digital economy ([5]).

However, network effects, data gathering, and artificial intelligence do not have to lead to market dominance. Their effects in competition may be diluted by multi-homing, which is the practice where users become part of more than one platform, as well as interoperability and portability of data, which makes switching less burdensome. Differentiation between the competing platforms may enable them to draw particular users, while innovation from new players may change the existing competitive advantages. In that sense, while network effects, data, and artificial intelligence may strengthen the dominant position of incumbent platforms, their role in creating market power depends on other aspects of competition, including switching costs and innovations.

Another important source of market power is switching costs, which make consumers less willing to move to a competing platform. Unlike traditional switching costs, which are mostly monetary, digital switching costs include the loss of accumulated digital assets, personalized recommendations, purchase history, cloud storage, social network, applications compatibility, and platform-specific knowledge ([26]). Consumer lock-in becomes especially relevant in the context of the integrated digital ecosystems. For example, users of the Apple ecosystem use such devices as iPhone, iCloud, Apple Pay, Apple Music, Apple Watch, and App Store. The replacement of any part of the ecosystem would mean the need to replace multiple other components and, thus, make switching costs higher ([27]).

The same situation happens with businesses that operate on Amazon Marketplace and Google Ads. Businesses usually spend a lot of time optimizing their work in the platform and using specific methods of promotion and attracting customers, which makes moving to the competing platform economically unprofitable. Such lock-in mechanisms decrease competitive pressure while giving incumbents the possibility to keep their customers for a long period of time.

Digital platforms exhibit exceptionally strong economies of scale, as the development of the software, cloud infrastructure, cybersecurity, and artificial intelligence requires large initial investments, while marginal cost of serving additional users is low. Growing in size, platforms’ average operating cost is falling rapidly, which allows companies to expand globally and stay profitable at the same time ([23]). Along with economies of scale, digital platforms get economies of scope, which consist of the use of technological infrastructure, algorithms, cloud resources, and customer relationships built in one business area in the context of multiple complementary products ([7]). Such strategy reduces the cost of operations and strengthens customer retention at the same time.

Ecosystem approach has become increasingly popular in modern digital platforms. The ecosystem strategy involves the integration of complementary products, such as operating system, app store, cloud infrastructure, digital payments, advertising networks, and artificial intelligence, into a unified technological environment ([27]). There are several competitive advantages associated with vertical integration. Companies get the preferential access to consumer data, coordinate user identity across all services, bundle complementary products, optimize advertising systems, and increase their market power in the adjacent industries. Moreover, such integration allows companies to practice self-preferencing, when they give preferential treatment to their own products in the context of search ranking or digital marketplace ([5]).

The competition authorities have raised concerns about the growing ecosystem expansion, claiming that it might substantially decrease market contestability by making consumers more dependent on the particular platform and decreasing the competitive opportunities for the independent players.

The interplay between network effects, data accumulation, switching costs, ecosystem integration, and economies of scale result in strong barriers to entry in digital markets. Unlike traditional industries, where barriers are often based on capital investment and access to the natural resources, barriers in digital markets are informational and technological. Although new firms often have innovative technologies, they cannot compete effectively due to the existence of the extensive user network, proprietary databases, developed developer community, and global technological infrastructure of the incumbent firms. Also, dominant companies often buy promising startups before they can become serious competitors ([6]). Therefore, the competition in the digital markets occurs more and more “for the market” rather than “within the market”.

Figure 2 identifies the key factors that contribute to the market power attained by digital platforms. The network effect leads to increasing numbers of users, which then results in increased data gathering, better artificial intelligence, increased service performance, user lock-in, expansion of the ecosystem, and high barriers to entry. This creates a reinforcing feedback loop, which helps the platform attain and maintain its dominance in the digital market sector ([5]; [27]).

Figure 2. Conceptual framework illustrating the major sources of market power in digital markets and their reinforcing interactions ([5]; [27]).

Table 1.Sources of market power in digital markets.

Table 1 describes the key factors that contribute to market power in digital platforms. Digital platforms do not work independently; rather, the key forces complement each other to create cumulative competitive advantages that give structure to digital markets that is different from industrial markets. The argument put forward in this section reveals that market power in digital platforms depends on the interaction of technological, economic, and organizational variables rather than any one of them. It is critical to understand these forces for developing appropriate measures to gauge competition in digital markets. This is why the next section covers conventional as well as new measures of market power in the digital economy.

In the past, competition regulators relied upon a series of indicators to estimate market concentration and the dominant position of companies. Market share of a company continues to be the most straightforward and commonly used indicator of market power. Companies with consistently high market shares exercise more control over the market processes and competitive environment. But the indicator becomes ambiguous in digital markets because many companies provide services free of charge, and thus it is impossible to make a valid comparison based on revenues ([1]).

Herfindhal-Hirschman index (HHI) represents another commonly used indicator of market concentration. It is defined as a sum of squares of market shares of all companies in the market,

where s i denotes the percentage market share of firm i . Higher HHI values indicate greater market concentration and potential competitive concerns ([16]). Competition authorities such as the U.S. Department of Justice (DOJ) and the Federal Trade Commission (FTC) routinely employ HHI thresholds during merger assessments.

Another classical indicator is the Lerner Index, which measures a firm’s pricing power by comparing price with marginal cost,

where P represents price and MC denotes marginal cost. Values approaching one indicate stronger monopoly power, whereas values close to zero suggest competitive markets ([21]). However, because many digital platforms offer free services to users, marginal-cost pricing provides limited information regarding actual competitive strength.

There are several features in the digital platform markets that reduce the validity of traditional competition measures. First, many of the services provided by platforms, such as search engines, social media platforms, mapping tools and communications applications are available for free. Platforms profit from users’ attention via advertisement, data analysis, subscriptions or complementary products. Therefore, price-based indicators significantly understate competitive power ([30]). Second, many digital platforms work in multi-sided markets where interplay between different sets of users creates indirect network effects. For example, Google search platform acts for the benefit of its users, advertisers, publishers and application developers at the same time. The market power in such cases cannot be assessed exclusively on the basis of pricing, it also requires coordination of interactions between multiple sets of participants ([24]). Third, digital markets are characterized by high dynamics, companies constantly innovate with the help of artificial intelligence, updates of software, cloud computing and development of ecosystems. The market power of a platform may have considerable future prospects despite its relatively small current market share due to the technology lead or innovation. Fourth, digital ecosystems make traditional boundaries of markets more blurred. Companies like Amazon, Google, Microsoft, Apple and Meta compete and collaborate across many industries simultaneously. It is increasingly hard to identify the relevant market by the means of antitrust practices ([5]).

Recognizing the limitations of the traditional methods, researchers have developed a set of market power indicators specific to digital platform markets.

1)User Base and User Engagement

The size and engagement level of a platform’s user base provide valuable information about its competitive status. Indicators like monthly active users (MAU), daily active users (DAU), average time spent per session, user retention and engagement rate tend to give better information about market power than traditional metrics based on revenues ([11]). A large, highly-engaged user base helps strengthen the network effects and makes the platform more attractive for advertisers and complementors.

2)Control of Data

Control over data has become the key indicator of digital market power. Platforms with rich datasets can utilize the data more efficiently than their competitors when developing recommendation systems, targeting ads, detecting fraud, optimizing logistics and building machine learning models ([29]). Since large datasets are not replicable, data control confers significant competitive advantage and erects entry barriers.

3)Artificial Intelligence Capability

The role of artificial intelligence has been growing in evaluating the competitiveness of a firm in digital markets. Better AI models mean higher quality of personalization, recommendations, prediction, automation of decision making and increased efficiency. Since algorithm’s performance usually gets better with access to larger datasets, AI capability is the embodiment of the accumulated competitive advantages of leading platforms ([29]).

4)Switching Costs

High switching costs reduce consumer mobility and increase market power. Indicators include limited interoperability, non-portability of digital assets, compatibility with complementary products, high migration cost and ecosystem dependency. High switching costs allow incumbents to maintain consumers despite existence of alternatives ([26]).

5)Ecosystem Control

Contemporary digital companies compete not with single products but with ecosystems. Therefore, the degree of vertical integration within operating systems, cloud services, digital payments, app stores, hardware and advertising networks gives important information about competitive position of firms ([7]). Companies with large ecosystems use their advantages in one market to gain advantages in adjacent markets.

6)Innovation Capability

Since digital competition becomes more technology-dependent, innovation capability is another key indicator of market power. R&D spend, patent portfolio, software development costs, acquisition of innovative startups and investments in AI all build up long-term competitive advantages ([25]). It is necessary to make a distinction between market concentration, competitive advantage, and market power because although these terms are related to each other, they are not identical. Market concentration involves the structure or market shares of the companies in question while competitive advantage means firm-specific strengths or resources which might help a company gain an advantage over its competitors. Market power implies that a firm is capable of acting largely independently of competitive constraints. It follows that having many users, lots of proprietary data, good AI technology or high investments into R&D might suggest competitive advantages of a platform and even market power but this does not mean that these characteristics ensure that market power exists.

Figure 3 depicts a multidimensional approach that enables the assessment of market power in digital markets through the combination of traditional competition metrics and digital metrics. This framework proves that structural concentration metrics need to be supported by user engagement metrics, data ownership metrics, AI capability metrics, switching costs metrics, ecosystem metrics, and innovation metrics to create a holistic view of platform dominance ([30]; [5]).

Figure 3. Integrated framework for measuring market power in digital markets using traditional and digital-specific indicators ([5]; [30]).

In sum, the measurement of market power in digital markets demands an expansion from the traditional price-oriented approach into a more general model which takes into account the relevance of data, artificial intelligence, engagement, integration, and innovation. While traditional measures of concentration can still be useful in pinpointing structural dominance, they offer a limited view of competition in digital markets. A more holistic analysis would thus call for a combination of the two types of measures. The multidimensional analysis mentioned above forms the basis of the discussion on market power in digital markets, which is presented in the next section. The analysis of market power in digital platforms markets needs both traditional concentration indices as well as specific digital indicators. Although the use of traditional indices like market share, HHI, Lerner index, concentration ratio and others provides valuable information about the structure of the market and its pricing power, it may not sufficiently reflect the specifics of the platform economy. Specific digital indices like user engagement, data possession, AI abilities, switching costs, integration into the platform’s ecosystem and innovation ability can provide valuable information about the position of digital platforms in the market. These indices are presented in Table 2 below.

Table 2. Comparison of traditional and digital-specific indicators used to assess market power in platform economies.

The European Union has become a world leader in regulating digital platforms using the Digital Markets Act (DMA) and the Digital Services Act (DSA). In the first act, gatekeeper platforms are obliged to give access to business users, provide data portability, avoid self-preferencing, and increase interoperability of competing services ([10]). The goal of the DMA is not destruction of large platforms but preservation of contestable digital markets by preventing activities that are unfair towards competitors. Together with DSA, the EU regulations strive to achieve transparency, consumer protection, and fair competition.

The policy framework in the United States continues to depend largely on Sherman Act, Clayton Act, and Federal Trade Commission Act. However, in recent years, there have been more investigations of large tech companies by the agencies involved in competition policy, including the investigations of Google, Meta, Amazon, Apple, and Microsoft.

Guidelines issued in recent years on mergers have emphasized the role of digital ecosystems, innovation competition, and potential competition more than current market shares. The emphasis is due to the realization that innovative start-up companies’ acquisitions and digital infrastructure can lead to reduced competition in the market irrespective of the lack of immediate impact on prices ([28]).

Several other countries have made their competition policies stronger for digital markets as well. In the United Kingdom, the Competition and Markets Authority formed Digital Markets Unit (DMU) to regulate companies that have Strategic Market Status (SMS). Additionally, China has formulated stricter laws for monopoly in the case of large internet platforms, while India has analyzed the app store market, digital payments market, online marketplaces, and search service under the Competition Act. Despite the differences in the policy frameworks, these initiatives have similar goals—improving market contestability, avoiding abuse of dominant position, transparency, and fair competition in the digital ecosystem ([30]). Figure 5 illustrates the main regulatory strategies taken by the leading nations for regulating digital platform markets. The figure shows the importance that is accorded to gatekeeper regulation, fair competition, transparency, interoperability, and consumer protection. There have been developments in the regulations for digital platforms in major economies due to emerging issues associated with market dominance, gatekeepers, data, and potentially anticompetitive behavior. Various regulatory frameworks have been implemented by different economies with similar intentions to foster fair competition and market contestability. Table 4 presents an overview of the main regulatory frameworks in the EU, USA, UK, China, and India along with the respective objectives and references.

Figure 5. Major regulatory approaches for digital platform competition ([30]; [10]; [28]).

Table 4. Comparison of competition policies for digital platforms.

In any case, competition policy seems to be developing into a new approach where more factors than price are being taken into account, such as data, networks effects, platform ecosystems, and digital gatekeepers. Despite different approaches to regulation in different countries, it seems evident that the concept of contestable digital markets can be maintained only through proactive control over the platforms involved.