Competition Law and Algorithmic Distribution Market Power

1. Introduction

Algorithmic distribution market power describes a situation in which a business uses algorithms, digital platforms, ranking systems, recommendation engines, app stores, search systems, or automated distribution infrastructure to exercise significant control over how products or services reach consumers.

Traditional distribution power often depended on ownership of physical infrastructure such as retail outlets, warehouses, transport systems, or dealer networks. In digital markets, distribution may instead depend on access to a platform, placement in search results, recommendation rankings, app-store visibility, default settings, or access to important data.

Competition law becomes relevant when a firm with substantial market power uses these algorithmic mechanisms to exclude competitors, favour its own products, restrict alternative distribution channels, impose unfair conditions, or strengthen barriers to entry.

The important point is that using an algorithm is not itself unlawful. Competition concerns normally arise from the conduct implemented through the algorithm and its effects on competitive conditions.

2. Meaning of Algorithmic Distribution

Algorithmic distribution occurs where software determines or significantly influences matters such as:

• which products consumers see;
• how products are ranked;
• which sellers receive prominent placement;
• which apps can reach users;
• which advertisements receive distribution;
• which suppliers are recommended;
• which products appear in search results;
• how sellers obtain access to customers;
• which distribution channels receive preferential treatment; and
• what contractual or technical conditions businesses must satisfy to obtain effective distribution.

For example, an online marketplace may contain thousands of sellers. Consumers cannot realistically examine every product. The platform's ranking algorithm therefore becomes an important distribution mechanism.

A seller appearing near the top of the results may receive substantial traffic, while another seller may effectively become invisible.

Consequently, control over the algorithm can sometimes produce economic power comparable to control over an important traditional distribution network.

3. Market Power in Algorithmic Distribution Markets

Competition authorities normally begin by defining the relevant market and determining whether the undertaking possesses substantial or dominant market power.

Important indicators can include market share, network effects, switching costs, economies of scale, control over data, consumer dependence, barriers to entry and control over important distribution infrastructure.

Digital platforms may obtain particularly significant distribution power where they operate as intermediaries between businesses and consumers.

This produces what is commonly called a multi-sided market.

A platform may simultaneously connect:

Consumers → Platform → Sellers

and sometimes:

Advertisers → Platform → Consumers

Algorithms determine how these different groups interact.

4. Algorithms as Digital Gatekeepers

A powerful platform may effectively become a gatekeeper between suppliers and customers.

Suppose competing businesses A, B and C sell products through Platform X. Platform X also introduces its own competing product.

If Platform X controls the ranking algorithm, it could theoretically give greater visibility to its own product.

The structure becomes:

Independent sellers → Ranking algorithm → Consumers

while the platform participates at two levels:

Platform = distributor + competitor

Competition concerns become especially important when independent businesses cannot realistically reach consumers through alternative channels.

5. Self-Preferencing

One major issue is algorithmic self-preferencing.

Self-preferencing occurs when a platform gives more favourable treatment to its own products or services than to competing products using the same platform.

Examples could include:

• higher search placement;
• preferred recommendation slots;
• special display formats;
• better access to consumer information;
• preferential default settings; or
• restrictions applying only to competitors.

Self-preferencing is not automatically unlawful in every jurisdiction or factual setting. Competition authorities normally examine market power, the nature of the preferential treatment, objective justification and actual or potential exclusionary effects.

A central authority here is Google Shopping.

The European Commission found that Google gave its own comparison-shopping service more favourable positioning and display in general search results while rival comparison-shopping services were subject to ranking mechanisms capable of lowering their visibility.

The General Court upheld the Commission's decision in 2021, and the Court of Justice dismissed Google's appeal in 2024. The litigation is especially important because it demonstrates how algorithmically controlled visibility can become an issue under abuse-of-dominance rules.

6. Algorithmic Ranking and Foreclosure

Foreclosure occurs where conduct makes it more difficult for competitors to obtain customers, inputs or effective access to the market.

Algorithmic foreclosure can occur without formally preventing competitors from operating.

For example, a platform could theoretically allow competitors onto its marketplace but systematically place them so far down its ranking that meaningful consumer access becomes difficult.

The distinction between formal access and effective access is therefore important.

A competitor might technically have access to a platform but still suffer competitive disadvantages because of:

• ranking discrimination;
• reduced visibility;
• default placement;
• data restrictions;
• interoperability restrictions; or
• discriminatory access conditions.

7. Default Distribution Arrangements

Defaults can also produce substantial distribution advantages.

Consumers frequently continue using pre-installed or default services rather than actively searching for alternatives.

This makes default placement commercially valuable.

The U.S. Google search litigation illustrates this issue. In the 2024 liability decision in United States v. Google LLC, the federal district court examined Google's agreements with browser developers, device manufacturers and wireless carriers concerning default search distribution.

The court found Google liable for unlawfully maintaining monopoly power in relevant search markets through its distribution arrangements. Subsequent proceedings addressed appropriate remedies.

This case demonstrates that competition law can examine contractual and technological distribution mechanisms together rather than considering algorithms in isolation.

Important Case Laws

8. Google Shopping — Google and Alphabet v European Commission

Case: Google and Alphabet v Commission (Google Shopping)

This is one of the most important authorities concerning algorithmically mediated distribution.

Google operated a general search service while also providing its own comparison-shopping service.

The European Commission concluded that Google gave more favourable positioning and display to its own comparison-shopping service than competing comparison-shopping services.

The General Court upheld the Commission's infringement finding in 2021. The Court of Justice subsequently dismissed Google's appeal in September 2024.

Importance

The case demonstrates that ranking and visibility can constitute important competitive parameters.

Where a dominant undertaking controls an important discovery mechanism, discriminatory treatment through that mechanism can potentially produce exclusionary effects.

The case therefore provides a major legal foundation for examining algorithmic distribution power under Article 102 TFEU.

9. United States v. Google LLC — Search Distribution

Case: United States v. Google LLC, U.S. District Court for the District of Columbia

The U.S. Department of Justice challenged Google's practices concerning distribution of its general search engine.

Google entered agreements relating to default search placement with important distribution partners, including browser developers and mobile-device participants.

In August 2024, the district court held that Google was a monopolist and had acted to maintain its monopoly in relevant general-search and search-advertising markets.

Importance

This case shows that distribution market power may arise from the interaction between:

• default settings;
• contractual arrangements;
• technological ecosystems;
• user behaviour; and
• scale advantages.

Algorithms become particularly significant because increased distribution can produce more queries and data, potentially reinforcing scale and competitive advantages.

10. European Commission v Google — Android

Case: Google Android

The European Commission's Android proceedings concerned restrictions associated with Google's Android ecosystem.

Among other matters, the Commission examined contractual requirements involving Google Search and Chrome and restrictions connected with Android device manufacturers.

The General Court largely confirmed the Commission's decision in 2022, while reducing the fine.

Importance

Android demonstrates that distribution power can arise through control over an ecosystem rather than simply ownership of a particular product.

Operating systems, app stores, browsers and default applications can operate together as interconnected distribution channels.

Competition law therefore examines whether arrangements involving these channels protect legitimate technological integration or unlawfully reinforce market power.

11. Apple App Store Practices — Music Streaming

Case: European Commission, Apple – App Store Practices (music streaming), AT.40437

This proceeding concerned Apple's rules affecting music-streaming applications distributed through Apple's App Store.

The Commission's 2024 decision focused on anti-steering provisions restricting music-streaming app developers from informing iOS users about alternative and potentially cheaper subscription possibilities outside Apple's ecosystem.

The Commission found that the conduct infringed Article 102 TFEU and Article 54 EEA.

Apple challenged the decision before the EU courts, so the procedural status of later litigation should be distinguished from the Commission's administrative finding.

Importance

The case illustrates how control over app distribution can generate competition concerns when platform rules restrict how competing businesses communicate with customers or direct them toward alternative purchasing channels.

Algorithmic distribution power is therefore not limited to rankings. It can include control over the architecture through which businesses reach customers.

12. Eturas UAB and Others v Lithuanian Competition Council

Case: C-74/14, Eturas UAB and Others v Lietuvos Respublikos konkurencijos taryba

Eturas involved an online travel-booking system used by multiple travel agencies.

A system administrator sent information concerning restrictions on discounts available through the platform.

The Court of Justice considered when participating travel agencies could be regarded as having participated in concerted conduct.

The Court stressed that participation could not simply be presumed merely because businesses used the same technological platform. Questions of knowledge and participation remained important.

Importance

Eturas is highly relevant to algorithmic markets because it demonstrates that traditional competition-law principles concerning agreements and concerted practices continue to apply when coordination occurs through digital infrastructure.

A platform or algorithm does not eliminate the need to establish the legal elements required for liability.

13. United States v. Topkins

Case: United States v. David Topkins

This U.S. prosecution concerned online sellers of posters sold through the Amazon Marketplace.

The participants agreed to coordinate prices and used automated pricing algorithms to implement the arrangement.

Topkins pleaded guilty to participating in the price-fixing conspiracy.

Importance

Although this case primarily concerns algorithmic pricing rather than distribution foreclosure, it establishes an important general principle:

Using software to implement anticompetitive conduct does not protect the underlying conduct from competition law.

Algorithms can therefore be treated as instruments through which traditional competition infringements are implemented.

The legal focus remains on the agreement and conduct of the market participants.

14. Trod Ltd / GB eye Ltd Online Poster Case

The UK's Competition and Markets Authority investigated competing online sellers of posters and frames.

The businesses agreed not to undercut one another on certain products sold through Amazon's UK marketplace.

Automated repricing software was used to help implement the arrangement.

Importance

Like Topkins, this case demonstrates how automated software can make anticompetitive arrangements easier to implement and monitor.

The underlying principle extends beyond pricing.

If businesses used automated systems to allocate customers, territories, sellers or distribution opportunities pursuant to an anticompetitive agreement, ordinary cartel principles could similarly become relevant.

15. Bronner v Mediaprint

Case: C-7/97, Oscar Bronner GmbH & Co KG v Mediaprint

Bronner predates modern algorithmic platforms but remains important when considering access to distribution infrastructure.

A newspaper publisher sought access to a rival's nationwide home-delivery system.

The Court of Justice established demanding conditions for treating refusal of access to infrastructure as an abuse of dominance.

Among other matters, indispensability and the elimination of effective competition became important elements of the analysis.

Importance for Algorithmic Distribution

Modern disputes may concern access to:

• platforms;
• APIs;
• app stores;
• ranking infrastructure;
• digital marketplaces;
• datasets; or
• technological ecosystems.

Bronner therefore provides an important conceptual starting point for determining when competition law can require access to infrastructure controlled by a dominant undertaking.

However, not every algorithmic discrimination case is treated as a Bronner-style refusal-to-supply case. The Google Shopping litigation is particularly important on this distinction.

16. Main Competition-Law Theories

Algorithmic distribution market power can potentially be examined through several competition-law doctrines.

Abuse of Dominance

Under Article 102 TFEU and comparable national rules, concerns can arise where a dominant platform uses distribution control to exclude competitors.

Potential theories include:

Self-preferencing — favouring the platform's products.

Discriminatory ranking — applying materially different visibility conditions to competing suppliers.

Tying — requiring one service to obtain another.

Exclusive dealing — limiting the ability of distributors or customers to deal with competitors.

Refusal to supply/access — denying access to infrastructure in circumstances satisfying the relevant legal test.

Predatory conduct — strategically sacrificing profits where the applicable legal requirements for predation are satisfied.

Margin squeeze — controlling an upstream input while competing downstream and leaving insufficient competitive margin in circumstances covered by the doctrine.

17. Algorithmic Distribution and Article 101 TFEU

Article 101 addresses agreements and concerted practices restricting competition.

Algorithms could facilitate arrangements concerning:

• price coordination;
• market sharing;
• customer allocation;
• output restrictions;
• territorial restrictions;
• coordinated distribution;
• resale price maintenance; or
• information exchange.

The important legal question is generally whether the required agreement or concerted practice can be established.

Algorithms do not themselves become legally responsible undertakings. Responsibility normally remains with the businesses controlling, adopting or using them.

18. Data as a Source of Distribution Power

Data can strengthen algorithmic distribution systems.

A platform with millions of users can observe consumer behaviour and use that information to improve:

• recommendations;
• rankings;
• advertising;
• product matching;
• personalization; and
• demand prediction.

Better algorithms may attract more users.

More users generate more data.

More data may improve the algorithm.

This can create a feedback mechanism:

More users → More data → Better algorithm → Better distribution → More users

Such feedback effects do not automatically establish unlawful monopoly or dominance. However, they can be relevant when authorities assess entry barriers and durability of market power.

19. Network Effects

Algorithmic distribution platforms frequently benefit from network effects.

A marketplace with many consumers attracts sellers.

More sellers increase product variety.

Greater variety can attract additional consumers.

This produces:

Consumers → Sellers → More Consumers → More Sellers

Strong network effects can make entry difficult for new platforms.

Competition authorities therefore consider whether platform conduct artificially strengthens these effects or prevents rivals from reaching sufficient scale.

20. Vertical Integration

Competition concerns can become particularly complex when the distributor also competes with businesses dependent on its distribution system.

Consider:

Platform

↓ distributes

Seller A — Seller B — Platform's Product

The platform controls distribution while simultaneously competing against A and B.

Potential concerns include whether the platform uses commercially sensitive information obtained from sellers, modifies rankings, restricts interoperability, imposes discriminatory conditions, or reserves valuable distribution opportunities for itself.

Vertical integration can also generate legitimate efficiencies, including lower transaction costs and improved service integration. Competition analysis therefore requires examination of both competitive harm and legitimate justifications.

21. Personalized Distribution

Algorithms increasingly personalize recommendations for individual consumers.

Two consumers searching for the same product may receive completely different results.

This creates difficulties for competition enforcement because discriminatory treatment may not be visible through a single standard search.

Authorities may therefore need to analyse:

• algorithmic outputs;
• historical ranking information;
• internal documents;
• platform experiments;
• data inputs;
• recommendation criteria; and
• actual traffic patterns.

Evidence of competitive effects may consequently become highly data-intensive.

22. Transparency Problems

Algorithmic distribution systems can operate as complex or partially opaque systems.

Competitors may know that their traffic has fallen without knowing whether the cause was:

• legitimate algorithmic improvement;
• changing consumer demand;
• poorer product quality;
• neutral ranking criteria; or
• exclusionary platform conduct.

Competition authorities therefore increasingly require sophisticated technical and economic evidence when investigating digital markets.

However, lack of transparency by itself does not automatically establish an antitrust infringement.

23. Barriers to Entry

Algorithmic distribution market power may create or reinforce entry barriers through:

Data advantages — entrants lack comparable historical information.

Network effects — consumers prefer established networks.

Switching costs — changing platforms can be inconvenient.

Default advantages — incumbent services receive automatic placement.

Ecosystem integration — multiple services reinforce each other.

Economies of scale — large platforms can spread technology costs across enormous user bases.

Brand recognition — consumers may trust established platforms.

Competition authorities examine whether these barriers result from legitimate competition or are reinforced through exclusionary conduct.

24. Efficiency Defences and Objective Justification

Not every algorithmic restriction is anticompetitive.

Businesses may argue that particular distribution rules improve:

• cybersecurity;
• consumer protection;
• privacy;
• platform integrity;
• fraud prevention;
• product quality;
• transaction efficiency; or
• interoperability.

Competition law therefore frequently requires examination of whether restrictions are genuinely necessary and proportionate to legitimate objectives.

A claimed efficiency cannot automatically justify exclusionary conduct. Authorities examine evidence connecting the restriction with the claimed benefit and whether less restrictive alternatives exist.

25. Remedies

Where competition authorities establish an infringement, possible remedies can include:

Cease-and-desist orders requiring termination of unlawful conduct.

Non-discrimination obligations requiring comparable treatment of competing services.

Contractual changes removing unlawful exclusivity or restrictions.

Choice mechanisms allowing consumers greater ability to select alternatives.

Interoperability requirements in appropriate regulatory or competition-law contexts.

Financial penalties where authorized by the relevant legal system.

Structural remedies may sometimes be considered, although competition authorities generally examine whether behavioural measures can adequately restore competitive conditions and must act within the governing legal framework.

26. Relationship with the Digital Markets Act

In the European Union, competition law now operates alongside the Digital Markets Act.

The DMA establishes specific obligations for designated gatekeepers. These can concern matters such as steering, interoperability, app distribution, data use and preferential treatment.

For example, in April 2025 the European Commission found Apple in breach of the DMA's anti-steering obligation and imposed a €500 million fine.

The DMA and traditional competition law should nevertheless be distinguished.

Traditional competition law normally requires investigation of matters such as relevant markets, dominance or agreements, competitive effects and the elements of a particular infringement.

The DMA establishes specified ex ante obligations applicable to designated gatekeepers.

27. Economic Test for Algorithmic Distribution Market Power

A useful analytical framework is:

Step 1 — Define the relevant market

Identify the relevant product/service and geographic market.

Step 2 — Establish market power

Consider market shares, entry barriers, switching costs, data advantages and network effects.

Step 3 — Identify the distribution mechanism

Determine whether distribution depends on rankings, recommendations, defaults, app stores, operating systems or another technological gateway.

Step 4 — Identify the conduct

Determine whether the conduct involves discrimination, exclusivity, tying, self-preferencing, refusal of access or coordination.

Step 5 — Examine foreclosure

Assess whether competitors' effective ability to reach consumers is materially reduced.

Step 6 — Establish causation

Distinguish competitive harm caused by the challenged practice from changes resulting from legitimate competition or consumer preferences.

Step 7 — Consider justification

Examine security, privacy, quality, efficiency and other legitimate explanations.

Step 8 — Determine appropriate remedy

Any intervention should address the identified infringement and comply with proportionality and the applicable statutory framework.

28. Key Case-Law Summary

29. Conclusion

Algorithmic distribution has transformed the concept of market power. A company no longer needs to own physical stores or delivery networks to influence market access. Control over search rankings, recommendation systems, defaults, operating systems, app stores, marketplaces and data-driven discovery mechanisms can provide substantial influence over how businesses reach consumers.

Competition law does not prohibit algorithmic distribution or market power itself. The central question is whether firms obtain or maintain power through conduct prohibited by the applicable competition rules.

The leading authorities—including Google Shopping, United States v Google, Google Android, Apple App Store Practices, Eturas, Topkins, Trod/GB eye and Bronner—show that existing competition-law concepts can be applied to technologically sophisticated markets. At the same time, algorithmic systems create difficult evidentiary questions because rankings are dynamic, personalization differs between users, network effects can reinforce incumbency, and the boundary between legitimate product design and exclusionary conduct can be highly fact-specific.

Accordingly, modern competition analysis focuses not simply on who owns the algorithm, but on who controls access to customers, how that control is exercised, whether competitors are foreclosed, and whether the challenged conduct represents competition on the merits or an unlawful restriction of competition.

I’ve kept this focused on algorithmic distribution market power and included more than six relevant authorities, while distinguishing direct algorithmic cases from older distribution precedents that supply the governing legal principles.