Competition Market Study of Online Marketplaces in Poland, Latvia and Lithuania

8. Concerns vis-à-vis business users

Abstract

The interaction between online marketplaces and the sellers using their platforms is shaped by three key features: the potential conflict of interest stemming from the hybrid nature of online marketplaces, seller dependency and informational and logistical asymmetries. These characteristics define how marketplaces operate, influence sellers’ strategic choices and create conditions that may reinforce power imbalances, with implications for competition. In particular, these structural features can give rise to exclusionary or exploitative behaviours, such as self-preferencing, the imposition of unfair trading conditions and tying. This chapter examines in detail the structural relationship between online marketplaces and sellers in Poland, Lithuania and Latvia.

This chapter turns to the structural relationship between online marketplaces and sellers across Poland, Lithuania and Latvia. Building on the market analysis in previous chapters, it examines how the structural features arising from configuration of these platforms and their growing role in digital commerce has shaped the commercial environment for sellers.

This section examines the structural relationship between online marketplaces and third‑party sellers, a critical factor in understanding market dynamics and competitive pressures. The interaction between platforms and sellers is shaped by three key features: the potential conflict of interest stemming from the hybrid nature of online marketplaces, seller dependency, and informational and logistical asymmetries. These characteristics define how marketplaces operate, influence sellers’ strategic choices, and create conditions that may reinforce power imbalances, with implications for competition.

The remainder of this chapter will explore how these structural features can give rise to exclusionary or exploitative behaviours and, in this light, identify specific practices that may emerge, such as, inter alia, self-preferencing, the imposition of unfair trading conditions and tying.

8.1. Structural features of the relationship between online marketplaces and sellers

Three interrelated structural features are key to understanding the dynamics of this relationship and, in turn, the risks it may pose to effective competition.

First, online marketplaces operate as hybrid entities: they serve as intermediaries connecting third-party sellers with consumers, while at the same time offering their own retail service in direct competition with those sellers. This dual role creates an inherent potential conflict of interest, as platforms may have both the incentive and the ability to favour their own offerings over those of third-party sellers.

Second, and as shown in previous chapters, sellers increasingly depend on online marketplaces as critical gateways to reach consumers. For many sellers, especially small and medium‑sized ones, platforms offer access to a broad customer base that would be difficult or prohibitively costly to replicate through other channels. This dependency is often reinforced by indirect network effects, where seller presence drives buyer engagement and vice versa.

Third and because of the previous points, the relationship between platforms and sellers is marked by a significant informational and logistical asymmetry, including disparities in bargaining position, control over contractual terms, and access to and control over commercially sensitive data. Platforms are able to observe, process and monetise granular seller and consumer data, while individual sellers often have only limited visibility and recourse.

These three features – potential conflict of interest, dependency, and informational and logistical asymmetry – are deeply interconnected and structural in nature and while analytically distinct, these features often reinforce one another and are present to varying degrees across the three national markets. Together, they shape the competitive landscape in which sellers operate and can give rise to behaviours that restrict competition or distort market outcomes.

8.1.1. Potential conflict of interest arising from the dual role of online marketplaces

The first aspect to be assessed in the structural relationship between online marketplaces and sellers is the potential conflict of interest arising from the hybrid nature of the platform’s business model. From a competition law perspective, concerns about platform neutrality are particularly relevant in situations where an intermediary holds significant market power, since only in such cases can discriminatory conduct materially affect competition and potentially trigger the application of abuse of dominance rules. As previously explained, in Poland, Lithuania and Latvia, the leading platforms operate under a hybrid model. This dual role creates an inherent and structural potential conflict of interest, as the platform operator simultaneously hosts and competes with independent sellers.

This is a common concern for online marketplaces, as recognised by numerous competition authorities. For instance, the European Commission and several other national competition authorities have investigated Amazon, focussing on how the company may have leveraged its dual role to favour its own retail operations.1 Similarly, even though the business model differs, Facebook Marketplace, an OCAS platform, has also attracted scrutiny regarding its incentives and ability to privilege its own services when acting as both platform and competitor.2 These cases highlight how the hybrid model embeds a potential for discriminatory conduct, independent of whether such conduct is actually observed.

At the core of this issue lies the tension between platform neutrality and commercial self‑interest. When the same entity also operates as a seller, it acquires a direct financial interest in securing sales on its own platform. This creates an incentive to shape the platform environment in ways that favour its own retail division, whether through design choices, data use or governance practices. From a competition law perspective, such risks are particularly relevant where the intermediary holds a dominant position, since in that context it bears a “special responsibility” not to allow its conduct to impair genuine, undistorted competition within the internal market.3 This concern has been confirmed in the case law, where the European Court of Justice has clarified that dominant platforms carry indeed a special responsibility not to allow their conduct to impair genuine, undistorted competition within the internal market.4

Importantly, the problem is structural: even if platforms may not always have a clear incentive to harm third-party sellers, since doing so could reduce the platform’s overall attractiveness or undermine indirect network effects, the alignment of platform governance and retail strategy within a single entity can still erode trust in the platform’s neutrality. The potential for self-preferencing is embedded in the business model and exists independently of whether discriminatory behaviour is actually observed. From a competition law perspective, this raises concerns about exclusionary conduct and the risk of discrimination, particularly where the platform holds significant market power or acts as a critical gateway for market access.

The risks stemming from this structural and potential conflict of interest are not theoretical. They are observable across the major platforms in the three countries. Evidence of these risks, alongside those stemming from seller dependency and informational and logistical asymmetries, will be illustrated in the final part of this chapter, which describes the anticompetitive practices that these structural features may incentivise.

In Poland, Lithuania and Latvia, the leading online marketplaces, Allegro, Pigu, and 220.lv, all operate under this hybrid model, combining intermediation services with their own direct retail operations.

When a platform captures a significant share of consumer traffic and seller transactions, the inherent potential conflict of interest can have tangible effects for both sellers and consumers. Independent sellers may face reduced opportunities to compete fairly, which can weaken their incentives to invest, innovate or offer better terms. While platforms have a clear interest in maintaining user trust, concerns remain that the opacity of platform operations and the lack of effective alternatives can nonetheless erode confidence in the platform’s neutrality. The specific ways in which this conflict of interest may give rise to anticompetitive conduct will be examined in the following sections.

Moreover, these risks arise in large part because of the high level of dependency that sellers often have on these platforms as essential gateways to reach consumers – a dynamic explored in the next section. As such, dependency is not merely an amplifying factor, but often a necessary condition for the conflict of interest to result in anticompetitive effects.

8.1.2. Seller’s structural reliance on online marketplaces

A second key feature of the relationship between online marketplaces and sellers is the significant and growing reliance of sellers on these platforms to access consumers. This dependency is a general structural characteristic of online marketplaces: as the segment consolidates around a single leading player in each national market covered by this report, sellers, especially small and medium-sized enterprises (SMEs), increasingly view these platforms as essential distribution channel to reach customers at scale. For many, participation is no longer optional but a commercial necessity.

This makes competition for sellers a key concern: reduced choice among platforms can weaken sellers’ bargaining position, limit their ability to switch and expose them to potentially exploitative conditions, ultimately reducing their ability to compete, invest or innovate.

This structural reliance, which can entail a degree of dependency, is reinforced by factors such as the platforms’ extensive reach, technical and logistical infrastructure, and indirect network effects that entrench their central role in digital commerce. The resulting asymmetry in dependence between sellers and platforms can have important implications for bargaining power and competitive dynamics.

In the context of this report, the next sections explain why such structural reliance is particularly relevant in Poland, Lithuania and Latvia, drawing on the evidence of market concentration and platform dominance presented in Chapter 7.

As detailed in Chapter 7, the largest online marketplaces function as unavoidable commercial partners for a broad range of sellers and as already mentioned, particularly for SMEs. Sellers depend on these leading platforms not merely as optional sales channels, but as the main means to access consumer demand at scale. This is because being listed on such platforms often ensures a baseline level of visibility and credibility for the product offering, which individual sellers would struggle to achieve independently. While consumer awareness of the seller’s identity may remain limited, the platform’s reach and interface still allow sellers to access a large customer base they might not reach through other channels.

This visibility stems from the platform’s aggregation function. As described in Chapter 4, online marketplaces also shape how offers are presented and discovered, through algorithms that rank listings, recommendation tools and increasingly, sponsored advertising slots that sellers can purchase to enhance their placement. This reinforces sellers’ structural reliance on the platform, as visibility is neither automatic nor guaranteed, but determined by the platform’s internal logic, which can also be monetised through advertising services. Even so, sellers gain substantially more exposure simply by being present on the platform than they would through independent sales distribution channels.

This situation may be analysed by analogy through the lens of the essential facility doctrine,5 albeit in a qualified sense. While online marketplaces cannot formally be characterised as public utilities, they may nonetheless perform a functionally similar role when they become the primary point of intermediation between supply and demand in digital commerce. In such circumstances, access to a dominant online marketplace can amount to a de facto precondition for effective participation in downstream retail markets. Where a platform concentrates the majority of consumer traffic and serves as the main interface through which consumers engage in online commerce, sellers that are excluded – or that face unfavourable access conditions – may be effectively foreclosed from the market. The platform’s role thus extends beyond intermediation to a form of structural control over market access.

This reasoning finds support in the General Court’s judgment in Google Shopping. The Court recognised that Google’s general search results page “has characteristics akin to those of an essential facility,” given the absence of viable substitutes and the structural dependence of competing specialised search services on access to that interface.6 In other words, the Court acknowledged the quasi‑essential function performed by Google’s interface as an unavoidable channel to users.

A similar reasoning applies in the context of dominant online marketplaces. While these platforms may not meet the strict conditions of the essential facility doctrine, their function as the primary commercial distribution channel in concentrated national markets may render them quasi‑essential facilities. Sellers’ dependency on such platforms resembles the dependence of specialised search services on Google’s traffic recognised in Google Shopping. In both instances, the issue is not only one of binary access (inclusion versus exclusion), but also of the conditions under which access is granted.

In the marketplace context, this extends to the neutrality and quality of intermediation: the design of search ranking algorithms, the allocation of visibility and promotional opportunities or the uneven enforcement of platform rules may all materially affect sellers’ ability to reach consumers. In hybrid models, these risks are compounded by the incentive to favour the platform’s own retail operations. In this sense, the analogy with Google Shopping underlines that foreclosure may result not from outright denial of access, but from discriminatory or self-preferencing conduct that degrades rivals’ effective visibility.

The relevance of this quasi‑essential facility framework can be illustrated by reference to the practices of Allegro. The investigation initiated by the Polish competition authority (UOKiK) into Allegro concerned allegations that the platform afforded preferential treatment to its own retail operations and to selected business partners, to the detriment of independent sellers. Such conduct directly echoes the concerns articulated in Google Shopping, where discriminatory self-preferencing was found to distort visibility and access to demand. In Allegro’s case, the concern was not the outright denial of access to the marketplace, but the degradation of competitive conditions through opaque and potentially biased intermediation mechanisms. The investigation highlights how control over visibility, rankings and promotional mechanisms enables the platform to influence competitive outcomes, even without denying access outright. By shaping which offers are more visible or attractive to consumers, Allegro could effectively privilege its own retail operations or selected partners, illustrating the core risk identified by the quasi‑essential facility framework: that control over a critical, non-replicable platform can allow a dominant operator to distort competition.

While no comparable enforcement action has been undertaken in relation to Pigu, its structural position in Lithuania and Latvia raises similar considerations. Pigu operates as a hybrid platform, competing directly with the sellers that rely on its marketplace to reach consumers. Its significant share of online commerce, combined with its role in controlling fulfilment and logistics services, enhances its ability to shape the terms of access to consumers.

Moreover and as already mentioned, online marketplaces’ dual role as both rule‑setters and commercial counterparties allows them to determine the architecture of search and transaction flows, while at the same time offering its own competing products or services (in hybrid models). In this situation, sellers have little control over the design of the marketplace and cannot opt out of changes in ranking algorithms, commission structures or mandatory use of ancillary services.

This imbalance creates a structural relationship of dependence, wherein sellers must accept the platform’s evolving terms in order to access consumers. Over time, this dependency may deepen: the longer sellers operate within a given ecosystem, the more they adapt their logistics, marketing and business processes to that environment. This may increase the costs of sellers using alterative sales platforms, including multihoming, especially when alternative platforms do not provide comparable visibility or consumer reach. Even where alternatives exist in theory, they often fail to offer a meaningful substitute in practice.

This structural dependency creates different risks: dominant platforms may use their position to impose higher commissions, introduce mandatory ancillary services or condition access to visibility on additional payments. Such practices enable platforms to extract rents or impose terms that would likely not be sustainable in a more competitive environment.

It is the online marketplace provider, who exercises overall control and management of the marketplace, that holds decision making power. In most cases, individual sellers lack the leverage to influence the platform’s policies. Small and medium-sized sellers especially possess significantly less economic strength and scale compared to the platform and the disparity remains substantial. In certain markets, sellers have no viable option but to rely on the platform.

In Poland, as also supported by previous analysis, sellers heavily depend on Allegro to reach consumers. The platform offers the most visible and trafficked interface for online marketplaces in the country. Its search and ranking system shapes buyer choices, while tools such as Allegro Smart, Allegro Pay and fulfilment services are often essential for maintaining competitiveness on the platform. Sellers who do not participate in these integrated services may find themselves at a disadvantage, both in terms of visibility and consumer trust.

Once embedded in Allegro’s ecosystem, many sellers have limited capacity to diversify their sales channels. Operating a separate online store involves acquiring new traffic, managing logistics independently, setting up a payment system and building trust from scratch, costly undertakings with uncertain returns. The ability to multi-home is limited in practice. Even when technically possible, sellers often lack the resources, expertise or incentives to do so, given Allegro’s centrality in Polish e‑commerce.

This seller lock-in is not an isolated outcome but a direct manifestation of Allegro’s broader ecosystem strategy. As outlined in Section 7.3, Allegro’s vertical/conglomerate integration across logistics, payments, related and supporting services, combined with its data advantages and targeted acquisitions, creates a self-reinforcing environment where participation in the platform becomes increasingly indispensable.

Chapter 7 details how this expansion reinforces both seller and consumer dependency. Allegro accounts for around 81% of the Polish online marketplace segment from the seller perspective and 71% of Polish consumers who accessed any of the top five marketplaces in the year 2025 visited Allegro, 47.7% of whom visited no other platform. Its Fulfilment by Allegro programme ensures fast, nationwide delivery, while around 85% of Allegro’s traffic comes from direct or organic sources, underscoring its role as the default e‑commerce channel in Poland (see Chapter 7).

In Lithuania, Pigu.lt plays a similarly central role in connecting sellers to consumers. The platform offers not only listings but, as already described through the previous chapters, it also offers an integrated suite of services including logistics, marketing tools and consumer-facing services such as price guarantees and loyalty schemes. These features increase conversion rates for participating sellers, but also tie them more closely to the platform’s infrastructure.

Pigu exhibits similar dynamics to Allegro, as documented in Chapter 7. For instance, its Fulfilment by Pigu (FBP) programme increases sales by around 37% compared to non-FBP listings, and leverages both its own lockers in Vilnius and partnerships with the majority of locker providers in the country. Moreover, 89.73% of Pigu.lt’s traffic comes from users based in Lithuania, signalling a comparable platform centrality (see Chapter 7).

Sellers on Pigu.lt would have difficulties in reaching a comparable audience outside the platform. Setting up a proprietary website or using a general classifieds platform typically results in lower visibility, less trust and higher customer acquisition costs. While alternatives exist in theory, few offer the same combination of reach, convenience and service integration. This results in a practical lock-in effect, limiting sellers’ ability to operate independently or in parallel on other platforms. As Chapter 7 explains, the structural integration of fulfilment systems, user loyalty and localised services in Pigu.lt’s offer creates material barriers to effective multi-homing for sellers in Lithuania.

Finally, sellers in Latvia exhibit similar patterns of dependence on 220.lv. The platform offers a comprehensive and locally tailored e‑commerce environment, including integrated delivery services, promotional tools and buyer protection mechanisms. As with its Lithuanian counterpart, participating in this environment significantly enhances seller visibility and transaction volume. As highlighted in Chapter 7, Pigu’s operations in Latvia also benefit from its logistics infrastructure and partnerships, reinforcing seller reliance on 220.lv. The FBP programme’s effectiveness and local adaptation, including integration with leading locker networks, contribute to a similarly high degree of seller dependence. In both cases, the interplay between fulfilment capabilities, platform visibility and consumer trust reduces the practical feasibility of multi-homing for sellers.

However, this dependence on the platform’s centralised infrastructure and consumer base comes at the cost of commercial autonomy. Sellers often lack the bargaining power to influence the platform’s rules or terms, and the combination of behavioural inertia, algorithmic opacity and high switching costs deters many from establishing a parallel presence elsewhere. In this environment, a seller’s commercial viability increasingly hinges on continued access to the dominant platform – exclusion or even unfavourable conditions of access, can significantly threaten their ability to operate at all. This structural dependency reinforces the platform’s role, magnifying its capacity to set terms unilaterally and potentially exploit or exclude trading partners.

Such dynamics not only affect individual sellers but can also distort competitive conditions in downstream retail markets. These effects are compounded when viewed alongside the first factor discussed, the potential conflict of interests naturally stemming from the hybrid role of marketplaces, and they intersect with a third key structural feature: the informational and logistical asymmetry, which translated in imbalance of data access between the platform and its sellers. This final aspect will be explored in the next section.

8.1.3. Informational and logistical imbalances in platform governance

A third structural feature shaping the relationship between online marketplaces and sellers is the asymmetry of information and logistical control. This asymmetry also arises from the hybrid nature of most leading online marketplaces, which act both as intermediaries and as direct competitors to the sellers operating on their platforms.

As explained above, this dual role creates a potential conflict of interest and places sellers in a position of structural reliance and dependency to access consumers and remain competitive, which is deepened by the platform’s unilateral control over key informational and operational levers – including access to commercially sensitive data, algorithmic ranking and recommendation systems, contractual conditions and the provision of analytics tools. In parallel, many market-leading marketplaces also control crucial logistical infrastructure, such as integrated fulfilment, warehousing and delivery services. Sellers who do not or cannot rely on these services may face reduced visibility or slower delivery standards, placing them at a competitive disadvantage. These forms of asymmetry – informational and logistical – limit sellers’ ability to negotiate or compete on equal terms.

The informational and logistical asymmetries described above manifest concretely in the standardised, non-negotiable terms imposed by online marketplaces. These asymmetries provide the incumbent platforms with superior knowledge of market dynamics (through access to transaction, consumer and pricing data) and control over essential logistics and ancillary services (such as delivery networks, payment systems or visibility tools). This combination places the platform in a structurally superior position vis-à-vis individual sellers, who lack equivalent information and cannot replicate the logistical infrastructure on which access to consumers depends. Sellers are typically required to accept these “take‑it-or-leave‑it” conditions in order to access the platform and remain competitive. This reflects the structural dependency outlined above: sellers rely on the incumbent platform to reach consumers, while the platform has little incentive to tailor conditions to individual business users.

A further manifestation of these asymmetries is the platform’s privileged access to commercially valuable data. As the central node in the transaction ecosystem, the platform collects granular data on consumer behaviour, product performance and market trends. This creates a significant information asymmetry: while the platform with strong market power enjoys a comprehensive overview of market dynamics, individual sellers typically receive only high-level or selectively curated insights, often at the platform’s discretion. As a result, sellers must make strategic decisions based on partial or outdated information, reinforcing their structural disadvantage.

Although some leading platforms provide dashboards or insights to sellers, including through advertising and promotional services, these tools are developed and controlled by the platform operator. The scope, granularity, and timeliness of the information remain at the platform’s discretion, and sellers generally have no means of independently verifying the completeness or neutrality of the data presented.7

In addition to data-related asymmetries, sellers also face significant informational disadvantages stemming from the opacity of platform algorithms. These systems determine core aspects of marketplace functioning, including product rankings, search visibility and pricing suggestions. Pursuant to the P2B Regulation,8 online intermediation services are required to set out the main parameters determining ranking and the reasons for their relative importance. The Commission’s guidelines9 further clarify that significant changes to those main ranking parameters must also be disclosed. However, these obligations do not extend to a detailed disclosure of algorithmic design, specific weightings or technical triggers for visibility adjustments. In practice, there is a high risk that sellers remain unable to ascertain why particular listings are promoted or demoted, how updates affect their performance, or whether certain categories of sellers or products are systematically advantaged. While the P2B framework mitigates opacity with respect to high-level ranking criteria, substantial informational asymmetries persist in relation to the operation and impact of algorithmic systems. This residual opacity may still mask conduct capable of distorting competition.

Finally, analytics tools provided by incumbent platforms may further exacerbate the structural dependency of sellers. These tools are often bundled with advertising or visibility-enhancing services and may include performance dashboards, engagement metrics or keyword optimisation suggestions. However, sellers remain entirely reliant on the platform for the accuracy, reliability and interpretability of such tools. There is no independent oversight or guarantee that these systems reflect objective data or that they are not designed to steer sellers toward greater reliance on paid services. Unlike ordinary commercial relationships where contractual imbalances are mitigated by the possibility of switching, here the dependency is reinforced by the platform’s control over essential data and visibility. Sellers cannot replicate these tools independently, nor can they easily substitute them by moving to another channel, given the platform’s control over access to consumers. This creates further reliance and dependency, as sellers may feel compelled to invest in advertising or promotional services without a clear understanding of their cost-effectiveness or comparative benefits. Moreover, the lack of transparency around how advertising algorithms interact with product visibility adds another layer of complexity and opacity.

Such concerns are not purely hypothetical.10 Allegro’s operations demonstrate a significant asymmetry in the acquisition, use and sharing of data generated on its platform, which raise concerns regarding the potential for competitive distortions vis-à-vis third-party sellers.11 As prescribed in its Terms and Conditions and privacy policies, Allegro is authorised to collect and process extensive datasets, including transactional details, buyer identifiers and offer content, and to use such data for analytical, statistical, and marketing purposes, including profiling buyers to tailor recommendations and predict future behaviours (Allegro, 2026[1]).

The platform employs a sophisticated recommendation system that evaluates multiple variables – such as views, co-views, purchases, delivery estimates and seller ratings – to prioritise offers, while also conducting promotion testing to refine services and maximise profitability. Sellers, by contrast, are only granted access to limited data pertaining to their own sales and may purchase broader statistical insights through tiered Allegro Analytics subscriptions (Allegro, 2026[2]); however, these do not bridge the information gap concerning individual user behaviour, detailed transaction data or the proprietary algorithms governing offer ranking and pricing.12

This structural data asymmetry, coupled with the opacity surrounding Allegro’s algorithms, may restrict sellers’ capacity to compete effectively on the platform, underscoring the potential for conflicts of interest inherent in Allegro’s dual role as both marketplace operator and retail competitor.

At the same time, Pigu’s operations illustrate a pronounced asymmetry in the acquisition, processing and sharing of platform-generated data, which may result in significant imbalances to the detriment of third-party sellers (Pigu, 2026[3]). Pigu systematically collects extensive procurement and browsing data from buyers, including details on ordered goods, quantities, prices, payment methods, delivery information, pages visited, time spent, items added to the cart and banner clicks. This comprehensive data acquisition, combined with the platform’s use of profiling and personalisation techniques, enables Pigu to create highly granular, structured datasets that are critically important for tailoring recommendations, predicting consumer behaviour and optimising business strategies, thereby generating value far beyond raw or unprofiled information.

While such data-driven insights are central to Pigu’s ability to refine its offer placement and promotional strategies, the extent to which sellers can access comparable data is markedly limited. Sellers are generally confined to information relating to their own transactions and basic statistical summaries accessible through the seller portal or available for purchase, with no access to the detailed, individual-level data or algorithmic mechanisms that underpin product ranking, pricing and promotional testing.13

Furthermore, Pigu retains exclusive control over its proprietary search and pricing algorithms and does not disclose either the specific criteria or their weightings used to determine product placement and offer visibility, despite the fact that these factors – such as price, delivery conditions and seller ratings – are central to sales outcomes on the platform. This informational imbalance can restrict sellers’ capacity to develop competitive strategies on equal footing, potentially entrenching Pigu’s competitive advantage in its dual role as both platform operator and retailer. Consequently, the persistent opacity of algorithmic processes and the absence of effective data-sharing measures serve to reinforce the structural asymmetry between Pigu and third-party sellers, increasing the risk of competitive distortions within the marketplace.

The analysis above shows that informational and logistical asymmetries are structural features of platform governance in online marketplaces. These asymmetries stem from the platform’s central position in the transaction ecosystem and are reinforced by its control over data flows, algorithmic systems and fulfilment infrastructure. The limited transparency around ranking criteria, data access, and performance analytics restricts sellers’ ability to operate and compete on an equal footing. While platforms may provide certain tools and insights, access remains largely at their discretion, and the underlying mechanisms often remain opaque. The dynamics observed in marketplaces such as Allegro and Pigu illustrate how these structural imbalances may shape the competitive environment and the nature of platform-seller relationships more broadly.

8.1.4. Conclusion

Together, the hybrid role of platforms, seller’s structural reliance, and informational and logistical asymmetries generate concrete risks for competition.

In particular, this structural configuration can increase barriers to entry and expansion in the online intermediation market itself, as well as in related services, such as logistics and advertising, by enabling the platform to leverage its privileged position and superior information and logistical advantages to favour its own offerings (self-preferencing) or restrict opportunities for rivals. At the same time, the dependence of sellers on the platform, coupled with their limited negotiating power, facilitates the exercise of market power through exploitative practices, such as the imposition of unfair contractual terms, restrictive pricing obligations or most-favoured-nation clauses (MFN clauses).

These dynamics can also create incentives for foreclosure effects, even if such incentives are not absolute. While hybrid intermediation platforms profit from third-party sales through commission revenues, they may nevertheless find it profitable to disadvantage certain sellers when this strategy reinforces their own position – for instance, by directing traffic towards their private‑label products, promoting sellers who rely on their advertising or logistics services or raising rivals’ costs in ways that increase dependence on the platform. In this sense, the risk is not necessarily of complete exclusion of third-party sellers, but of selective discrimination that entrenches the platform’s market power.

Overall, this combination of factors can distort competition, limit innovation and ultimately harm consumers through reduced choice, higher prices or lower quality. Recognising how these features interact is therefore crucial for effective competition enforcement.

The next section explores concrete ways in which these risks may materialise in practice, including self-preferencing (see also Box 8.1 below for an example from the European Commission), the imposition of unfair trading conditions, such as by imposing so-called Most-Favoured-Nation (MFNs) clauses and other conduct, such as tying, that can exploit or exclude business users.

8.2. Potential competitive risks to sellers

Having outlined the main structural features that shape the relationship between online marketplaces and independent sellers, this section examines the potential anticompetitive conducts that may arise from these dynamics. It focusses on the risks associated with certain types of behaviour – in particular, self-preferencing by hybrid platforms, imposition of unfair trading conditions, such as Most-Favoured-Nation (MFNs) clauses and tying – in light of the potential conflict of interest, asymmetry and dependencies identified above. Where relevant, the discussion draws on available evidence to illustrate how such conducts may manifest in practice.

In doing so, this report’s analysis combines a conceptual framework with concrete examples from the online marketplaces under review, highlighting how structural features create incentives for certain behaviours and shape their likely competitive impact. The objective is to show not only the types of risks that may arise, but also the mechanisms through which they operate, the channels through which they may affect sellers and competition, and the conditions that make them more likely to materialise in each national context.

8.2.1. Self-preferencing

One of the key risks that emerges from the interaction of the structural features identified above is the risk of self-preferencing. Although the number of concluded investigations remains limited, the practice has attracted increasing attention from EU competition authorities as a potential source of harm.

The concept of self-preferencing has been recognised as a potential abuse of dominance where a dominant platform uses its intermediation power to systematically favour its own services in rankings to the detriment of rivals. This theory of harm was central to the European Commission’s findings in the Google Shopping decision,14 which concluded that such conduct could distort competition and harm consumers by limiting access to better or more relevant alternatives.

In the specific context of online marketplaces, self-preferencing can take multiple forms, all of which flow directly from this embedded and potential conflict of interest. For instance, a marketplace that controls the design and functioning of search and ranking algorithms can deliberately place its own retail products in more prominent positions on search result pages, regardless of whether those products are actually more competitive in terms of price, quality or delivery conditions. This can effectively steer consumer attention and purchasing decisions towards the platform’s own offerings, while relegating better or cheaper third-party alternatives to positions of lower visibility where they are unlikely to be discovered, thus distorting the outcome of competition.

The problem is further compounded by the platform’s privileged access to vast volumes of granular, non-public transactional data generated by the activities of third-party sellers, including sales volumes, conversion rates, customer preferences and trends in popular products. By leveraging this sensitive business information, the platform’s own retail division can selectively target best-selling product lines for replication under private labels or make informed decisions about pricing, stock levels and marketing strategies with significantly reduced risk, since the initial market testing costs are borne by independent sellers. Such conduct not only weakens the competitive position of sellers who lack comparable insights, but also discourages innovation and investment by third parties who fear being undercut by the very intermediary on which they depend.

Moreover, the risk of self-preferencing is amplified by the inherent opacity of algorithmic ranking, recommendation systems and eligibility conditions for prominent features such as “Buy Box” placement, shipping guarantees or loyalty schemes. This opacity is linked to the informational and logistical imbalance described above and exacerbates the different bargaining powers that the platform and its dependent sellers have, reinforcing the potential conflict of interest that makes self-preferencing so harmful in practice.

In addition to competition enforcement, recent regulatory frameworks have also sought to address self-preferencing through clear ex ante prohibitions. These rules aim to prevent conduct that could distort competition by leveraging the platform’s dual role to favour its own products or services over those of third-party sellers. Notably, Article 6(5) of the EU’s DMA15 explicitly prohibits designated gatekeepers from treating their own products or services more favourably in ranking or related conditions than those of third parties, effectively codifying the principle that the neutrality of intermediation services must be safeguarded to preserve fair competition. The DMA’s approach is significant because it reflects a broader understanding that the structural conflict of interest embedded in hybrid marketplace models is not easily eliminated by transparency obligations or behavioural commitments alone; rather, clear and enforceable obligations are necessary to prevent gatekeepers from using their control over critical parameters of competition, such as access to data, ranking criteria or promotional tools, to disadvantage rival sellers. Other jurisdictions have begun to adopt similar approaches, tailored to their domestic legal and market contexts.16

Ultimately, the risk of self-preferencing is emblematic of how the dual role of online marketplaces can entrench the imbalance of power between dominant platforms and dependent sellers, reducing sellers’ ability to compete on the merits, limiting consumer choice and distorting incentives for innovation.

As a way of example, in Italy, the national competition authority has pursued a case against Amazon for using its platform to give preferential treatment to sellers using its own logistics services, illustrating how conflicts of interest on online marketplaces can raise competition concerns.17

A clear and relevant example illustrating how such structural conflicts of interest can materialise at the national level is the Polish competition authority’s investigation into Allegro. As explained in Chapter 3, in that case, UOKiK found that Allegro had granted its own retail arm privileged access to promotional tools and marketing functions that were either not equally accessible or were restricted for competing third-party sellers. In simple terms, Allegro’s dual role enabled it to boost the visibility of its own products through special promotional placements and discounts while limiting the reach of rival sellers who depended on the platform to access Polish consumers. This unequal treatment allegedly reinforced Allegro’s competitive advantage by leveraging its position as an indispensable intermediary to direct more traffic and sales to its own retail offerings, reflecting precisely the risks that arise from the conflict of interest inherent in the hybrid marketplace model.18

It is also important to note that in the case of Allegro, the risk of self-preferencing can manifest not only through practices internal to the online marketplace itself but also via external platforms that are nonetheless part of the broader ecosystem controlled by the same undertaking. As discussed in Chapter 4, Allegro operates a CSS which functions as an additional channel for directing consumer traffic. By controlling both the online marketplace and the CSS, Allegro is in a position to potentially steer users towards its own retail products or preferred listings through favourable placement or ranking on the comparison site. In simple terms, this means that Allegro can reinforce the visibility and attractiveness of its own offerings beyond the confines of its marketplace, effectively amplifying the self-preferencing effect.

The CSS thus operates as an extension of the structural and potential conflict of interest inherent in the hybrid platform model, as the platform both provides essential intermediation services and competes with the sellers that depend on them. This dual role creates incentives to favour its own downstream offerings – including through affiliated CSSs – and raises concerns that the mechanisms determining visibility, ranking, or access to promotional tools may not be applied in a neutral or transparent manner.

While there have not yet been any formal competition enforcement actions addressing self-preferencing by leading national players in Lithuania or Latvia, the structural conditions that create such risks are readily observable in these markets as well, most notably in the case of Pigu. Like Allegro and as already discussed, Pigu combines intermediation services for independent sellers with its own retail activities, enjoying privileged access to non-public data and control over search rankings, placement algorithms and promotional tools. In markets with relatively high platform dependency and limited alternative routes to reach consumers at scale, the incentive and ability to engage in subtle or overt forms of self-preferencing are significant.

Indeed, while some of these practices could, in principle, fall under competition enforcement scrutiny, the structural features of these marketplaces can generate competitive concerns even in the absence of a formal finding of abuse. The combination of intermediation services with in-house retail activities, control over platform data and influence on visibility and promotional tools creates conditions where sellers may face subtle disadvantages or constrained opportunities and consumers may encounter reduced choice or less transparent offerings. These structural dynamics illustrate that the potential for harm extends beyond instances that would necessarily trigger enforcement action.

Where this occurs, the potential consequences for competition can be far-reaching. In the core intermediation market, self-preferencing may undermine the visibility and competitiveness of third-party sellers, making it harder for them to grow and sustain viable businesses, which in turn reduces consumer choice and limits price competition. Given the often opaque nature of platform algorithms and promotional mechanisms, these effects can materialise before they come to light, potentially causing irreparable harm to sellers. This highlights the rationale for considering interim measures or mechanisms for rapid intervention, as waiting for full enforcement proceedings could allow competitive disadvantages to become entrenched and, in some cases, lead to the exit of affected businesses.

Over time, this can reinforce the platform’s market power as sellers become increasingly reliant on its ecosystem and less able to multi-home or switch. In related markets, such as logistics, advertising, or payment services, self-preferencing practices may further entrench the platform’s position by steering demand towards its integrated or affiliated services, foreclosing independent providers and raising barriers to entry. As discussed in Section 7.3, this dynamic can contribute to a self-reinforcing cycle of expansion and entrenchment, deepening structural dependencies and diminishing competitive constraints across the broader digital ecosystem.

Box 8.1. The Amazon case – Asymmetry of power and hybrid platform risks

While Amazon does not appear to hold a strong market position in the jurisdictions covered by this study at this stage, the European Commission’s assessment in other national markets offers a useful illustration of the potential competition risks arising from hybrid platform models and self-preferencing behaviours.

On 20 December 2022, the European Commission adopted a formal decision concluding two antitrust investigations into Amazon’s commercial practices in the European Economic Area (EEA). The investigations were conducted under Article 102 of the Treaty on the Functioning of the European Union (TFEU) and Article 54 of the EEA Agreement. These provisions prohibit the abuse of a dominant market position. The cases in question, AT.40 462 – Amazon Marketplace and AT.40 703 – Amazon Buy Box, examined Amazon’s dual role as both a marketplace provider and a direct retailer, and whether this structure gave rise to unfair competitive advantages.

The first investigation, Amazon Marketplace, focussed on the company’s use of non-public commercial data. When third-party sellers operate on Amazon’s platform, Amazon gains access to data such as transaction volumes, product performance, pricing and consumer preferences. The Commission found that Amazon used this sensitive data to adjust its own retail strategies – such as pricing, stock planning and product development – thereby competing directly with those same sellers on its own platform. This self-preferential use of data was seen as incompatible with fair platform governance, particularly in national markets where Amazon was determined to hold a dominant position, namely in Germany and France.

The second investigation, Amazon Buy Box, analysed how Amazon selected which product offers were prominently displayed in the “Buy Box” on product pages. The Buy Box is the default purchasing interface for consumers and accounts for a significant majority of sales – according to Amazon’s own data, over 80% of purchases were made through the Featured Offer shown in the Buy Box. The Commission concluded that Amazon’s algorithms systematically favoured its own retail offers, as well as those from sellers using Amazon’s own logistics network, Fulfilment by Amazon (FBA). In parallel, the investigation also reviewed access to the Amazon Prime label. Offers fulfilled outside of Amazon’s network (the so-called Merchant Fulfilled Network) faced greater difficulty in being eligible for Prime benefits, which in turn affected their visibility and competitiveness.

In both cases, the Commission preliminarily concluded that Amazon had abused its dominant position in several national markets. In the Buy Box case, the concerns extended beyond Germany and France to include Spain. While Amazon did not accept the findings of the Commission’s preliminary assessment, it nonetheless proposed a set of commitments under Article 9(1) of Regulation (EC) No 1/2003. These commitments were intended to address the Commission’s competition concerns without the need for a formal finding of infringement.

The commitments, made binding by the Commission’s decision, included Amazon’s obligation to refrain from using non-public data collected from third-party sellers to inform its own retail operations. Amazon also agreed to apply equal treatment to all sellers in the selection criteria for the Buy Box and Prime eligibility, regardless of whether they use Amazon’s fulfilment services. Furthermore, Amazon committed to displaying a second competing offer in the Buy Box in certain conditions, to increase visibility of alternative options.

8.2.2. Imposition of unfair trading conditions

Unfair trading conditions (UTCs) represent a well-established form of exploitative conduct under EU competition law. Article 102(a) TFEU explicitly prohibits dominant undertakings from directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions.19 These conditions encompass contractual or commercial terms that are unjustifiably onerous, disproportionate, or otherwise distort the balance of rights and obligations to the detriment of a weaker counterparty. Such practices are especially prevalent in vertical relationships characterised by significant bargaining power imbalances, where one party has limited or no ability to reject disadvantageous terms.

In online marketplaces, UTCs manifest through diverse mechanisms, often involving a combination of pricing rules, contractual restrictions and discretionary enforcement. Common forms include excessive or non-transparent commission fees that erode sellers’ margins beyond what would be sustainable in a competitive environment, or mandatory payments for ancillary services that sellers must accept to secure sufficient visibility or favourable ranking. Other restrictive practices include limitations on sellers’ ability to price or promote goods on other channels.

The imposition of UTCs can give rise to both exploitative and exclusionary consequences in the market. From a competition perspective, sellers may face disproportionately high costs, limited operational flexibility or constrained margins, which can indirectly translate into higher prices, reduced innovation, or decreased consumer choice. From an exclusionary perspective, such conditions may foreclose alternative sales channels, deter new entrants, or weaken the competitive constraints smaller or emerging platforms can exert. Collectively, these effects tend to reinforce the market power of leading platforms and diminish overall market dynamism.

The systemic nature of UTCs in digital markets has prompted regulatory responses at both EU and national levels. The DMA, for example, explicitly prohibits designated gatekeepers from imposing certain unfair contractual restraints detrimental to business users and consumers. Notably, Article 5(3) of the DMA bans broad MFNs clauses that restrict sellers from offering more favourable terms on competing platforms.20 More generally, the DMA obliges gatekeepers to ensure fair, reasonable and non-discriminatory conditions for business users, mirroring the principles underlying Article 102 TFEU’s prohibition of UTCs.

Overall, the imposition of UTCs reflects the structural dependence many sellers have on large online marketplaces. This pattern of conduct exacerbates power imbalances, restricts competitive dynamics and risks entrenching exploitative outcomes. The following section will focus specifically on MFN clauses as a concrete example of how UTCs can operate and contribute to these broader competitive concerns and will assess how this materialise in practice in the three jurisdictions at stake. Indeed, if a platform holds a dominant position and its MFN clause materially limits sellers’ pricing freedom or forecloses rival platforms, it can be characterised as the imposition of unfair trading conditions under Article 102(a) TFEU.

8.2.3. Most Favoured Nations (MFN) clauses

Certain structural features of online marketplaces may facilitate the use of contractual or commercial practices that can negatively affect competition. One such practice is the use of MFN clauses. MFN clauses, sometimes referred to as parity clauses, can be defined as contractual provisions whereby sellers agree not to offer better terms – such as lower prices or a wider selection – on other sales channels, including competing platforms or their own direct stores. Put simply, an MFNs clause enables the platform to guarantee to consumers that its offer will be at least as favourable as those available elsewhere, reducing the likelihood that sellers will undercut the platform’s prices or product assortment.

MFNs can deter price competition and reinforce seller dependency on the platform. At the same time, some argue that MFNs serve a pro‑competitive function. The overall effect of such clauses depends on the specific market context, including the platform’s market position and the degree of seller multi-homing.

Indeed, while MFNs clauses may raise concerns about restricting competition, they can also generate potential efficiencies in certain contexts. For example, MFNs are often justified on the grounds that they prevent so-called free‑riding21 by sellers who might otherwise benefit from the platform’s investments in marketing, reputation or customer acquisition, only to divert sales to cheaper channels. By guaranteeing that the platform offers the most competitive price or terms, MFNs can incentivise the platform to invest in service quality, innovation and consumer trust. These efficiencies are more likely to materialise in markets where platforms do not hold a very strong position, where seller multi-homing is common, or where the platform’s role in stimulating demand is particularly significant. In such cases, MFNs may contribute to greater inter-platform competition, improved consumer experience and reduced search costs. However, and as already mentioned, the relevance of these benefits must be assessed in light of the platform’s market power and the broader market dynamics.

MFN clauses can generally be divided into two categories: wide MFNs and narrow MFNs. Under a wide MFN, a seller is prohibited from offering more favourable conditions on any other intermediation platform or sales channel, including its own website or stores. In contrast, narrow MFNs only prevent a seller from offering better terms on its own direct channels, while allowing potentially better offers on other third-party platforms.

The potential anticompetitive risks of MFN clauses have been recognised in multiple jurisdictions and in wider contexts than just online marketplaces. For example, in its press release concluding an investigation against Amazon Japan, the Japan Fair Trade Commission identified three principal competitive harms: (i) restricting sellers’ commercial freedom to adjust prices or assortments on alternative channels; (ii) distorting competition among platform operators by enabling the platform to secure best prices and widest selection without competing vigorously on commissions or service quality; and (iii) reducing incentives for innovation and market entry because lower platform fees would not necessarily translate into consumer benefits under MFN constraints (Japan Fair Trade Commission, 2017[4]).

The EU has also addressed these risks both in the Vertical Restraints Guidelines and through the DMA.22 Notably, the EU Guidelines on Vertical Restraints highlight several factors that should be considered when assessing the competitive impact of MFNs, such as the relative market power of the platform, the extent of coverage of such clauses, the degree of multi-homing by sellers and consumers, barriers to entry and the role and viability of direct sales channels for sellers.23

As to the DMA, Article 5(3) of the DMA prohibits designated gatekeepers from imposing wide MFNs clauses that restrict business users from offering their goods or services under more favourable conditions on competing intermediation services. However, the DMA does not expressly address narrow MFNs, which remain subject to case‑by-case assessment under general competition law.24

The case of Allegro Prices Program: Eligibility and exclusion of offers

The Allegro Prices Program25 offers an interesting example of how certain platform pricing practices may interact with the dynamics discussed above. Under this voluntary support programme, Allegro incentivises sellers to maintain competitive pricing on its platform by automatically reducing prices on selected listings,26 compensating sellers for the difference and boosting their product visibility in search rankings. Sellers that meet quality and turnover conditions can benefit from these adjustments without upfront financial risk.27

As noted above in relation to the potential benefits of MFN clauses – such as preventing free‑riding – Allegro itself frames the purpose of its Allegro Prices Program in similar terms. According to the platform, “When you sell on Allegro, you can take advantage of many free forms of sales support that help drive your business growth and ensure buyer satisfaction. Within the Allegro Prices program, we cover a part of the order price – so that buyers pay less while your profit stays the same. When we advertise your offers outside our platform, we increase their traffic, and you do not bear any additional costs.”28

In other words, the Allegro Prices Program allows the platform to automatically reduce the price shown to consumers for certain products, while ensuring that sellers still receive the full amount they initially set. Allegro covers the difference between the seller’s original price and the discounted price offered to buyers. This means that sellers can appear more competitive – especially in price‑sensitive categories – without having to sacrifice their margins. In addition to the financial compensation, products included in the programme also benefit from greater visibility in search results, increasing the likelihood of sales. The programme is only available to sellers that meet specific performance criteria (such as a minimum turnover or quality rating) and participation is voluntary.

A notable feature of this programme is that sellers whose products are offered at lower prices through other sales channels (including not only their own online stores but potentially other marketplaces) can be excluded from the programme’s benefits, under certain circumstances. While participation is technically optional, sellers relying heavily on Allegro for sales and consumer reach may face commercial pressure to align their pricing strategies accordingly. Indeed, an updated version of the Allegro Prices Programme, effective 1 July 2025, introduces stricter conditions for maintaining participation. Allegro announced on 23 June 2025 that sellers whose external prices differ from their Allegro listings beyond a narrow threshold may have all of their offers removed from both the Allegro Prices Programme and Allegro-sponsored external advertising (e.g. on Google, Ceneo, Facebook or Instagram).29 This revision underscores the platform’s heightened emphasis on price alignment across sales channels.

In practical terms, this mechanism may produce effects similar to those generally associated with wide MFNs provisions Wide MFNs clauses typically require sellers not to offer better terms (especially lower prices) on any other sales channel, including their own website or competing marketplaces. While the Allegro Prices Program does not impose a contractual obligation to maintain price parity, it creates strong economic incentives that may have equivalent effects. Specifically, sellers offering lower prices elsewhere may be excluded from benefits such as price subsidisation and higher visibility in search results – both critical drivers of sales on the platform. This dynamic may discourage sellers from pricing more competitively on other channels, thereby reducing inter-platform price competition and reinforcing Allegro’s position, especially in a context where sellers are highly reliant on the platform.

Given the importance of price competitiveness and search result visibility for online sales, access to such promotional tools may be critical for a seller’s viability on the platform. In a market where Allegro is found to hold significant market power, and where alternative online sales channels are limited, such conditions could dampen incentives for competing marketplaces to attract sellers by offering lower commission fees or better terms and could ultimately reduce competitive pressure across the ecosystem.

Even though Allegro covers the cost of the price reduction, the programme can influence seller behaviour. Sellers may avoid offering lower prices elsewhere if doing so risks losing the promotional benefits and price compensation on Allegro. In practice, this can reduce incentives for sellers to multi-home, i.e. to sell at competitive prices on other platforms or their own stores, because prioritising Allegro becomes more attractive.

At the same time, the programme’s stated objective is to prevent free‑riding, encouraging sellers to pass on the benefits of Allegro’s investments in platform promotion and consumer trust. While such mechanisms may, in some cases, support the platform’s broader ecosystem, their potential effects on competition – particularly where they resemble price parity conditions – are complex and context-dependent. A full assessment of their competitive impact would fall within the scope of enforcement proceedings rather than a market study. Nonetheless, the presence of such provisions raises questions about the balance between platform co‑ordination and seller autonomy that merit closer consideration in specific factual and legal contexts.

In sum, while the Allegro Prices Program does not constitute a contractual MFN clause in the strict sense, it illustrates how certain price parity conditions, embedded in incentive schemes, may raise similar competition policy questions. Any final assessment would require careful examination of its scale, sellers’ dependency on the programme and the broader market context.

The case of Pigu: Price ceiling clause in seller agreement

The contractual framework governing seller participation on Pigu’s platforms provides a clear illustration of how platform rules can affect seller pricing. The seller agreement gives Pigu significant discretion over pricing, including the unilateral ability to set maximum price limits on product listings.

In particular, the seller agreement concluded between Pigu and the sellers30 hosted within its platforms, stipulates that Pigu retains the right to establish a maximum price for any product listed on its platform. The seller agreement31 grants Pigu sole discretion to determine and to modify such maximum prices at any time and without prior notice to the sellers concerned.

Under the terms of the agreement, when Pigu sets or amends a maximum price for a specific product, the relevant price or the data used in its calculation is made available to the seller through the system or by other means. However, if such information remains inaccessible to the seller, Pigu may not exercise its right to reject or delist offers exceeding the maximum price, as the seller must be informed of the applicable ceiling in order to comply.

If a seller considers that the designated maximum price is set at an unreasonably low level and therefore renders the listing of the product commercially unviable or likely to incur additional costs, the seller may notify Pigu of such concerns. Nevertheless, such notification does not oblige Pigu to reconsider or revise its decision and Pigu retains ultimate discretion in setting or modifying the price ceiling.

Should a seller’s offer price exceed the maximum price established by Pigu, Pigu is entitled to refuse to activate the listing or to deactivate the product. If the non-compliance is identified prior to the product being listed, Pigu must provide prior notice to the seller. Conversely, where a product has already been listed, Pigu reserves the right to deactivate the listing without prior notice, but remains obliged to inform the seller of such deactivation ex-post. Notably, the seller agreement does not specify the precise criteria, data sources or algorithms employed by Pigu in determining maximum prices, creating a risk of unpredictability for sellers.

The practical implications of this pricing mechanism warrant further reflection. In effect, Pigu’s right to impose a maximum price can work like a MFN clause, even though no formal MFN obligation appears in the contract. An MFN clause typically requires a seller to offer a platform a price at least as low as the price offered on any other channel. By unilaterally setting a price ceiling, Pigu can produce the same result: if the ceiling is set at or below the prices the seller charges elsewhere, the seller may be required to either lower those outside prices to stay aligned, accept lower margins on Pigu or stop selling the product there. This mechanism can therefore standardise prices across sales channels and reduce the scope for sellers to compete on price. Such de facto MFN effects can weaken inter-platform competition because rival marketplaces find it harder to attract sellers with lower prices, dampening incentives to cut commissions or improve services and ultimately keeping consumer prices higher.

In summary, Pigu holds wide discretionary power to set and modify maximum prices for all products offered on its platform. While sellers are granted access to the relevant pricing data and may raise objections, these objections do not bind Pigu to any reconsideration. Moreover, the right to delist products without prior notice once they have been made available for sale exacerbates the risk that sellers, particularly those with a significant reliance on Pigu, may be deprived of meaningful pricing autonomy and face unexpected commercial outcomes.

Conclusion

The mechanisms outlined in this section reflect the growing complexity of commercial relationships between online marketplaces and sellers, particularly in environments characterised by strong platform intermediation and limited alternative sales channels. Whether through formal MFN clauses, incentive‑based pricing schemes or contractual price ceilings, these practices illustrate how platforms may influence or constrain sellers’ pricing decisions across channels. While the precise effects of such mechanisms depend on a range of market-specific factors, their design and implementation raise broader questions about the evolving distribution of commercial power in digital markets and the structural dependencies that may arise (or be reinforced) as a result.

8.2.4. Tying

Tying is a practice that merits close examination in the context of online marketplaces, as it directly reflects the structural relationship between platforms and their business users. While tying can occur in many sectors, it raises particular concerns in platform markets with dominant incumbents, due to the dual role of online marketplaces as both intermediaries and providers of ancillary services. In such settings, tying should not be seen merely as a discrete commercial strategy but rather as a manifestation of the structural dependencies that shape market behaviour. Under EU competition law, tying typically involves making the use or access to one product or service (the tying product) conditional upon the use of another, separate one (the tied product). When implemented by a dominant firm, such conduct may infringe Article 102 TFEU, which prohibits abuses of dominance that distort competition and affect trade between Member States.

In Lithuania, Latvia and Poland, the conditions for tying are present due to the way national marketplaces structure their ecosystems. The leading platforms in those national markets, i.e. Allegro, Pigu and 220.lv, indeed bundle core intermediation functions with services such as logistics, fulfilment, payments, advertising, and customer support. These additional services, while nominally optional, are increasingly functionally indispensable for sellers to compete effectively. This reflects a broader ecosystem strategy, discussed in Section 7.3, whereby dominant platforms combine marketplace intermediation with vertically integrated or tightly affiliated service offerings that lock in sellers and raise the cost of using independent alternatives.

In Poland, Allegro’s role exemplifies this model. Its extensive logistics infrastructure – built through acquisitions like X-Press Couriers and partnerships with InPost – enables fast, nationwide delivery that smaller platforms struggle to replicate. Sellers benefit from Allegro’s FBA programme, which enhances visibility and sales performance, but in practice also limits sellers’ ability to switch or multi-home. As highlighted in Chapter 5, Allegro accounts for the overwhelming majority of marketplace traffic, and seller multi-homing appears constrained, with many merchants relying exclusively on the platform. Similarly, Pigu and its Latvian counterpart 220.lv operate hybrid models that blend first-party retail with third-party intermediation, while offering integrated fulfilment and payment services on which participating sellers heavily depend.

In these highly concentrated and relatively small markets, the integration of ancillary services into the leading marketplace offer creates significant barriers to switching. Sellers who attempt to operate across platforms must often maintain separate inventories, co‑ordinate with incompatible fulfilment systems and bear higher operational costs. Even in the absence of formal contractual tying, this form of functional dependency creates commercial pressure that can limit sellers’ ability to choose independent providers.

This dynamic distorts competition not only in the intermediation market but also in the tied service markets. Independent logistics or payment providers, for example, may be foreclosed from access to a substantial segment of e‑commerce activity if sellers feel compelled to use platform-controlled services to remain competitive. These concerns are amplified by the lack of viable alternatives in national markets and the high switching costs faced by sellers, as also discussed in Chapter 7.

For the competition authorities of Lithuania, Latvia and Poland, several aspects warrant closer factual inquiry. First, evidence on how search ranking, algorithmic visibility or consumer-facing labels (such as “fast delivery” or “recommended seller”) are linked to the use of platform-controlled fulfilment or payment solutions would help determine whether sellers who decline the tied service are placed at a competitive disadvantage. Second, data on relative pricing and take‑up rates of the integrated services, compared with independent alternatives, could clarify whether the platform’s terms amount to a de facto obligation. Third, information on the profitability and market shares of independent logistics and payment providers would shed light on possible foreclosure effects and barriers to entry. Such evidence would be relevant not only for assessing exclusionary abuse under Article 102 TFEU but also for evaluating whether self-preferencing or discrimination provisions of the DMA could apply, in the event the same groups would be designated under such Act.

While direct evidence of explicit tying clauses remains limited in these three countries, the observed operational practices and the strong economic incentives they create point to a structural strategy capable of reinforcing the market power of such platforms. The potential effects include reduced choice and higher costs for sellers, weakened innovation in ancillary markets and diminished competitive pressure on the core intermediation service. A detailed investigation of these mechanisms – particularly the link between visibility advantages and the use of integrated logistics or payment solutions – could therefore provide a sound basis for future enforcement action, whether under EU competition law or the aligned national provisions that empower each authority to address abuses of dominance within its territory.

8.3. Conclusion

This chapter has examined the profound challenges arising from the structural characteristics of online marketplaces and their far-reaching implications for business users. The analysis has highlighted three critical, interrelated features that define these digital ecosystems: the potential conflict of interest inherent in platforms’ dual role as both indispensable intermediaries and direct competitors; the escalating dependency of sellers on these platforms to access essential consumer demand; and the significant informational and logistical asymmetry that shapes platform‑seller relations. These features are not theoretical abstractions but have been concretely illustrated through the platform strategies and competitive dynamics analysed in Chapter 7.

While this report aims to provide a comprehensive view of the competitive landscape of online marketplaces in Poland, Lithuania and Latvia, this chapter has specifically focussed on articulating the structural vulnerabilities embedded in the design and operation of these markets. Drawing on the platform strategies documented in Chapter 4 and analysed in Chapter 7 – such as Allegro’s vertical integration into logistics and payments, its acquisitions of Ceneo, eBilet and Mall Group, and the platform’s use of Allegro Pay and Smart loyalty schemes – Chapter 8 sought to show how these ecosystem-building efforts reinforce seller dependence and deepen power asymmetries. Similarly, Pigu’s fulfilment programmes, ownership of key logistics infrastructure, and expansion across the Baltics via 220.lv, as presented in Chapters 4 and 7, reflect a comparable logic of entrenchment. These strategic choices, though often framed as service enhancements, have cumulative effects that reduce sellers’ commercial autonomy and limit the contestability of both intermediation and related service markets.

Importantly, the relevant risks do not emerge in isolation but are deeply embedded in the logic of platform ecosystems described in Section 7.3, where integrated service offerings become functionally indispensable for sellers.

This structural interdependence – whether through fulfilment, payments or advertising services – translates into de facto exclusivity, particularly in smaller, highly concentrated national markets where platform alternatives are limited and multi-homing, if at all possible, entails significant costs. In this light, the practices analysed in this chapter are not isolated instances of potentially harmful business conduct but rather manifestations of a broader competitive dynamic already taking shape across the three national markets under study.

It is therefore essential to underscore that the anticompetitive risks discussed in this chapter are illustrative rather than exhaustive. They serve to demonstrate how structural conditions create fertile ground for harmful behaviours, even in the absence of formal violations. The objective here has not been to offer a definitive assessment of platform conduct – aside from specific, publicly acknowledged cases such as the UOKiK’s decision against Allegro – but to reveal how the design of these markets enables systemic vulnerabilities. As demonstrated across both Chapters 7 and 8, these concerns are not episodic but structural and persistent. Their recurrence or evolution in new forms poses an ongoing risk to fair competition and consumer welfare.

Key findings

This chapter examined the structural relationship between online marketplaces and sellers in Poland, Lithuania and Latvia. It identified three interrelated features which, although analytically distinct, may create persistent imbalances in these markets and, under certain conditions, may give rise to competition concerns. These features – potential conflicts of interest, dependency and asymmetry – are not isolated phenomena. Rather, they frequently reinforce one another, amplifying their impact on the platform‑seller relationship.

• Hybrid business models may create conflicts of interest. The leading marketplaces (Allegro, Pigu, 220.lv) combine intermediation services for third-party sellers with their own retail operations. This dual role embeds an inherent potential conflict of interest, as platforms may have both the incentive and the ability to privilege their own offers over those of independent sellers. Even where such conduct is not observed, the mere coexistence of intermediation and retail activities within the same entity may undermine confidence in platform neutrality and may facilitate discriminatory outcomes.

• Seller dependency on dominant platforms may reduce commercial autonomy. For many sellers, especially SMEs, access to large online marketplaces has become indispensable to reach consumers at scale. High concentration levels, strong indirect network effects and the integration of loyalty schemes, fulfilment and payment services mean that sellers often view participation as unavoidable, and this is unlikely to change in the short-medium term. This dependency may limit sellers’ capacity to multi-home or to develop independent sales channels, raising switching costs and reinforcing lock-in effects. Over time, such dependency may reduce sellers’ bargaining power and may expose them to less favourable contractual or commercial conditions.

• Informational and logistical asymmetries may reinforce platform control. Platforms enjoy privileged access to detailed consumer and transaction data, control over algorithms that determine visibility and sales outcomes, and discretion over integrated fulfilment systems. By contrast, sellers typically have access only to partial or selectively curated data and must adapt to standardised, non-negotiable contractual terms. This informational and logistical asymmetry may constrain sellers’ ability to compete on equal terms and may entrench the platform’s privileged position.

• Potential implications for competition. This structural configuration does not necessarily lead to harmful outcomes in every case, but it may facilitate both exclusionary and exploitative practices. Exclusionary risks may arise where platforms use their hybrid position, data advantages or logistics control to steer demand towards their own offers or services, potentially foreclosing rivals. Exploitative risks may arise where dependency and asymmetry allow platforms to impose unfair trading conditions, including parity clauses, restrictive pricing obligations or mandatory use of ancillary services. These risks may be particularly pronounced in smaller and more concentrated national markets, where alternatives for sellers are limited.

Overall, the analysis in this chapter suggests that the structural relationship between marketplaces and sellers may create a market environment prone to imbalances of power, with consequences for competition that might not necessarily be fully addressed by antitrust enforcement. Further, while not all risks materialise in practice, the combination of hybrid business models, seller dependency, and informational and logistical asymmetries may set the conditions under which anticompetitive conduct can emerge, and may reduce sellers’ capacity to compete and innovate on equal footing.

References

[1] Allegro (2026), Allegro Terms & Conditions, https://allegro.pl/regulamin/en?srsltid=AfmBOorjXAYPzIoRNTHkWMhloNsMrUe889rVhaft-g32e5eTNQOKqDp4 (accessed on 1 April 2026).

[2] Allegro (2026), What Allegro Analytics is and how it works, https://help.allegro.com/en/sell/a/what-allegro-analytics-is-and-how-it-works-ZMl7xwEwaSW (accessed on 1 April 2026).

[4] Japan Fair Trade Commission (2017), Closing the Investigation on the Suspected Violation of the Antimonopoly Act by Amazon Japan G. K., https://www.jftc.go.jp/en/pressreleases/yearly-2017/June/170601.html.

[3] Pigu (2026), Privacy Policy, https://pigu.lt/lt/t/privatumo-politika (accessed on 1 April 2026).

Notes

← 1.  See European Commission, AT.40 462 – Amazon Marketplace and AT.40 703 – Amazon Buy Box, Decision of 20 December 2022, https://ec.europa.eu/competition/antitrust/cases1/202310/AT_40703_8990760_1533_5.pdf.

← 2.  See European Commission, Case AT.40 684 – Facebook Marketplace, Decision of 14 November 2024, https://ec.europa.eu/competition/antitrust/cases1/202513/AT_40684_10582539_13405_4.pdf.

← 3.  This expectation is reflected in the Amazon case just mentioned and also in Regulation (EU) 2022/1925 (Digital Markets Act), Article 6(5), which establishes a legal obligation of neutrality in ranking and related practices for designated gatekeepers.

← 4.  See, e.g. Case C‑322/81, NV Nederlandsche Banden-Industrie Michelin v Commission (Michelin I), ECLI: EU: C:1983:313, para. 57; and Case C‑497/99 P, Irish Sugar plc v Commission, ECLI: EU: C:2001:454, para. 112, https://curia.europa.eu/jcms/jcms/Jo1_6308/.

← 5.  See Judgment of the Court of 26 November 1998, Oscar Bronner GmbH & Co. KG v Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co. KG and Others, Case C‑7/97, ECLI: EU: C:1998:569, paras. 41‑46 (setting out the cumulative conditions under which access to a facility may be mandated, including indispensability, elimination of competition, and absence of objective justification).

← 6.  Case T‑612/17, Google and Alphabet v Commission, ECLI: EU: T:2021:763, paras. 224‑225.

← 7.  In recognition of these concerns, certain jurisdictions have adopted ex-ante data-sharing obligations targeting powerful platforms. For example, Article 6(10) of the Digital Markets Act (DMA) requires designated gatekeepers to provide business users with access to high-quality, effective, continuous and real-time data generated in the context of the use of the platform’s services, including personal data provided or generated by end users in their interactions with sellers. Similarly, under Article 5(2)(i)(d) of Japan’s Act on Improving Transparency and Fairness of Digital Platforms, designated platforms must disclose the categories of data they collect and the conditions under which such data may be used.

← 8.  “Providers of online intermediation services shall set out in their terms and conditions the main parameters determining ranking and the reasons for the relative importance of those main parameters as opposed to other parameters.” – Article 5(1), Regulation (EU) 2019/1 150 (P2B Regulation), https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32019R1150.

← 9.  Commission Notice – Guidelines on ranking transparency pursuant to Regulation (EU) 2019/1 150 (OJ C 424, 8.12.2020) provides guidance on the ranking transparency requirements in Article 5 of the P2B Regulation, including how and when the description of main ranking parameters should be updated, and clarifies that detailed algorithmic workings need not be disclosed.

← 10.  In the FTC v. Amazon, the US Federal Trade Commission alleged that Amazon developed a pricing algorithm known as “Project Nessie,” which tested whether competing retailers would follow its price increases Where competitors were expected to match higher prices, Amazon implemented price hikes, thereby potentially softening price competition across the market. While these allegations remain subject to judicial review, they illustrate how a dominant platform may use algorithmic tools to influence market outcomes beyond its own transactions. Complaint, Federal Trade Commission v. Amazon.com, Inc., Case 2:23‑cv‑01 495 (W.D. Wash. 2023).

← 11.  Article 8.11 of Allegro’s Terms and Conditions: “The User does not have access to all data which are available to the Company. The User has free of charge access to data relating to their actions as part of Allegro. On Allegro Lokalnie only data on individual completed Transactions is available. The Company enables paid access to certain statistical data relating to the sale as part of the Allegro platform. Statistical data relating to the sale as part of Allegro are stored by the Company also after the termination of the Agreement with the User.”, https://allegro.pl/regulamin/en?srsltid=AfmBOorjXAYPzIoRNTHkWMhloNsMrUe889rVhaft-g32e5eTNQOKqDp4.

← 12.  Allegro Subscription: what it offers and why you should activate it – About Subscriptions – Help for sellers – Allegro, https://help.allegro.com/sell/en/a/allegro-subscription-what-it-offers-and-why-you-should-activate-it-Xx7vbV7WdcD#basic-subscription.

← 13.  https://pigu.lt/lt/marketplace/en#:~:text=Track%20order%20statuses%20and%20analyse,trading%20results.

← 14.  European Commission, Case AT.39 740 – Google Search (Shopping), Commission Decision of 27 June 2017.

← 15.  Article 6(5) of Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector (Digital Markets Act) [2022] OJ L 265/1.

← 16.  For example, Japan’s Act on Improving Transparency and Fairness of Digital Platforms does not ban self-preferencing outright but imposes robust disclosure requirements on key ranking parameters and the use of third-party seller data, thereby seeking to mitigate the risks that flow from the same structural conflict of interest.

← 17.  AGCM, Case A528 – Amazon Logistics, Decision of 30 November 2021.

← 18.  Polish Office of Competition and Consumer Protection (UOKiK), Decision DOK‑3/2022 of November 2022 in the Allegro case; see also https://www.concurrences.com/en/bulletin/news-issues/december-2022-4460/the-polish-competition-authority-fines-the-nation-s-largest-online-trading.

← 19.  Article 102(a) TFEU literally prohibits, as an abuse of a dominant position, “directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions.”

← 20.  See Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector (Digital Markets Act) [2022] OJ L 265/1, Article 5(3).

← 21.  In this context, “free riding” means that a seller takes advantage of the investments made by a platform (such as advertising or bringing in customers) but tries to avoid paying for them by offering the same product at a lower price on another sales channel, like their own website, to attract those customers away from the platform.

← 22.  See Commission Notice – Guidelines on Vertical Restraints (2022/C 248/01) [2022] OJ C 248/1, para 377.

← 23.  See Commission Notice – Guidelines on Vertical Restraints (2022/C 248/01) [2022] OJ C 248/1, para 78. The Commission Notice states the following: “(a) the market position of the provider of online intermediation services that imposes the obligation and of its competitors; (b) the share of buyers of the relevant online intermediation services that are covered by the obligations; (c) the homing behaviour of the buyers of the online intermediation services and of end users (how many competing online intermediation services they use); (d) the existence of barriers to entry to the relevant market for the supply of online intermediation services; (e) the significance of the direct sales channels of buyers of the online intermediation services and the extent to which those buyers are able to remove their products from the platforms of the providers of online intermediation services (de-listing).”

← 24.  See Regulation (EU) 2022/1925 of the European Parliament and of the Council of 14 September 2022 on contestable and fair markets in the digital sector (Digital Markets Act) [2022] OJ L 265/1, Article 5(3).

← 25.  https://help.allegro.com/en/sell/a/allegro-prices-the-support-program-for-the-sellers-ZMl7xw8Y6Uw#what-you-can-gain-with-allegro-prices.

← 26.  Section 4(1) of Allegro Price Program states that through the programme mechanism, the Company will reduce the Partner’s Proposed Price by at least a percentage equal to the Minimum Reduction by Allegro, but no more than 60% of the Partner’s Proposed Price.

← 27.  https://salescenter.allegro.com/terms/regulamin-programu-alleobnizka-zl95B570rsM.

← 28.  https://help.allegro.com/en/sell/a/on-july-1-we-will-change-the-qualification-requirements-for-the-allegro-prices-program-and-our-ads-D7R4ZmxEGcG#how-do-you-compare-the-prices-of-products-i-sell-in-a-set-.

← 29.  Exclusion applies if price differences occur for more than 10% of all products listed under the same NIP number and involve over 50 distinct products. An offer is deemed “more expensive” if the price difference equals or exceeds: 15 grosz or 1% (whichever is greater) for items priced up to PLN 50.49; 0.5% of the product value for items priced between PLN 50.50 and 100.49; or 0.25% of the product value for items priced above PLN 100.50, https://help.allegro.com/en/sell/a/on-july-1-we-will-change-the-qualification-requirements-for-the-allegro-prices-program-and-our-ads-D7R4ZmxEGcG#how-do-you-compare-the-prices-of-products-i-sell-in-a-set-.

← 30. https://lt2.pigugroup.eu/uploaded/files/Seller%20Agreement_General%20Conditions%20(with%203%20annexes)_2025_12_15%20EN.pdf.

← 31.  Articles 5.6‑5.6.4 of the Seller Agreement between Pigu.lt/220.lv and sellers establish that, to ensure the Website remains attractive to consumers through competitive pricing, the Service Provider may set a maximum price for any Product sold on the Website and may revise such prices at its sole discretion and without prior notice. Under Article 5.6.1, any maximum price set for a specific Product applies uniformly to all Sellers offering that Product on the Website but does not limit the Seller’s pricing freedom outside the Website. If a Seller considers the maximum price unreasonably low and believes it prevents listing the Product profitably, the Seller may inform the Service Provider, which will decide at its discretion whether to adjust the price. Pursuant to Article 5.6.2, the Service Provider may refuse to activate a Product listing if the Seller’s price exceeds the maximum price and must notify the Seller through the System or other means. Under Article 5.6.3, once a Product is listed, the Service Provider may monitor prices periodically and deactivate any listing that exceeds the maximum price without prior notice, while remaining obliged to notify the Seller afterwards. Finally, Article 5.6.4 requires that the maximum price or the information needed to calculate it be made available to the Seller; if this is not done, the Service Provider cannot reject or deactivate listings under the preceding clauses.