Competition Law and Preferential Treatment in Digital Ecosystems

Introduction

Preferential treatment in digital ecosystems, often called self-preferencing, occurs when a digital platform gives its own products, services, subsidiaries, affiliated sellers, or ecosystem participants more favourable treatment than competing third parties.

Examples include:

• ranking a platform's own products above rival products;
• giving preferential placement to its own services in search results;
• using non-public seller data to compete against marketplace sellers;
• giving preferential access to a platform's data, APIs, app stores, or payment systems;
• favouring affiliated logistics, advertising, cloud, or payment services;
• manipulating algorithms so that affiliated businesses receive greater visibility;
• imposing discriminatory conditions on competing businesses while favouring affiliated businesses.

The competition-law concern is particularly significant where the platform is simultaneously the operator of an essential digital gateway and a competitor within that gateway.

1. Meaning of Preferential Treatment

Preferential treatment is not automatically unlawful.

A platform may legitimately favour its own service where the preference is based on objective criteria such as:

• quality;
• security;
• technical compatibility;
• consumer protection;
• fraud prevention;
• reliability;
• genuine innovation;
• objectively superior performance.

The competition concern arises when the platform uses market power over the gateway to distort competition in a related market.

A typical structure is:

Dominant platform → controls gateway → competes downstream → favours own downstream product → rivals lose visibility/access → competition is weakened.

For example:

Search engine → controls search rankings → operates shopping service → places own shopping service above competing comparison services.

2. Why Digital Ecosystems Create Special Competition Problems

Digital ecosystems differ from traditional markets because a single firm may control several interconnected layers.

Typical ecosystem

Operating system

↓

App store

↓

Payment system

↓

Search/discovery

↓

Advertising

↓

Cloud/data infrastructure

↓

Consumer-facing services

The same company may participate at almost every level.

This creates a potential vertical conflict of interest.

The platform operator has access to:

• competitors' data;
• search queries;
• transaction information;
• ranking information;
• consumer behaviour;
• seller performance;
• advertising information;
• application usage;
• pricing information.

It may therefore have both the ability and incentive to favour its affiliated services.

3. Principal Forms of Preferential Treatment

A. Search-ranking preference

A platform gives its own services:

• higher rankings;
• greater screen space;
• prominent display;
• special search boxes;
• preferential recommendation;
• richer presentation.

This was central to the Google Shopping and Naver cases.

B. Marketplace preference

A marketplace operator may favour products sold by itself or its affiliates.

Possible mechanisms include:

• Buy Box allocation;
• search rankings;
• recommendation systems;
• customer reviews;
• promotional placement;
• Prime-type programmes;
• logistics integration.

The Amazon Marketplace investigation illustrates these concerns.

C. Data-based preference

A platform can use competitors' non-public data to improve its own competing products.

For example:

Marketplace collects seller-level sales data → platform identifies high-demand product → platform launches/expands its own competing product → platform uses superior information to compete.

This creates a data asymmetry that may be relevant under abuse-of-dominance rules.

D. App-store preference

An app-store operator may favour:

• its own applications;
• its own payment system;
• affiliated subscriptions;
• affiliated streaming services;
• its own advertising services.

Competition concerns can involve both self-preferencing and exclusionary access restrictions.

E. Default and pre-installation preference

A platform can give its own service an advantage by making it:

• the default search engine;
• pre-installed;
• difficult to uninstall;
• prominently positioned;
• technically integrated into the operating system.

This can reinforce network effects and make entry by competitors more difficult.

4. Legal Framework

A. Abuse of Dominance

Traditional competition law generally asks:

1. Is the undertaking dominant?
2. What is the relevant market?
3. Is the undertaking using that dominance to favour itself or disadvantage rivals?
4. Does the conduct produce exclusionary effects?
5. Is there an objective justification?
6. Are there less restrictive alternatives?

In the EU, the principal framework is Article 102 TFEU.

In India, the principal framework is Section 4 of the Competition Act, 2002.

In China, the Anti-Monopoly Law and platform-economy rules address discriminatory treatment, tying, exclusive dealing and other abuses.

5. Relevant Factors in Digital Ecosystems

Competition authorities commonly examine:

1. Market power

The platform does not necessarily have to be a conventional monopoly. Its importance as a digital gateway, network effects, switching costs and data advantages can be relevant.

2. Gateway control

The more important the platform is for reaching consumers, the greater the potential competitive significance of preferential placement.

3. Vertical integration

Self-preferencing becomes particularly important where the platform competes with the businesses that depend upon it.

4. Algorithmic discrimination

Authorities may examine whether algorithms systematically produce different treatment.

5. Network effects

More users → more data → better service → more users.

This feedback loop can make an initial preference increasingly difficult for competitors to overcome.

6. Switching costs

Users may face costs in moving between:

• operating systems;
• marketplaces;
• cloud services;
• payment systems;
• social networks;
• app ecosystems.

7. Foreclosure

The central economic question is often whether preferential treatment can substantially reduce rivals' ability to compete.

6. Major Case Laws

Case 1 — Google Shopping v European Commission

Google Search / Comparison Shopping Services — EU

This is the leading modern case concerning self-preferencing.

The European Commission found that Google had systematically given prominent positioning and display to its own comparison-shopping service while applying ranking algorithms to competing comparison-shopping services.

The General Court upheld the Commission's central findings in Google and Alphabet v Commission, T-612/17, judgment of 10 November 2021.

The case is important because the conduct was not simply a refusal to supply access. Google continued to provide search access to competitors but allegedly differentiated the manner in which its own and competing services were displayed.

The EU framework therefore demonstrated that discrimination in a dominant digital gateway can raise Article 102 concerns even where competitors technically continue to receive access.

The EU's later regulatory approach expressly builds upon this experience.

Principle

A dominant digital intermediary may infringe competition law by using its gateway position to systematically favour its own downstream service.

Case 2 — Naver Search Algorithm Case

Naver — Korea Fair Trade Commission, 2020

The Korean Fair Trade Commission found that Naver had manipulated its search algorithms in relation to shopping and video services, placing its own products and services more favourably while reducing the exposure of competitors.

The KFTC described the conduct as involving manipulation of search-result exposure through changes to the algorithm.

The case is especially important because it demonstrates that algorithmic design itself can become a competition-law issue.

The concern was not merely that Naver owned competing services. It was that the search algorithm allegedly treated those affiliated services differently from rival services.

Principle

Algorithmic neutrality can become a competition concern where a dominant platform modifies ranking mechanisms to favour its own affiliated services.

Case 3 — Google Search Bias Case

Google — Competition Commission of India, 2018

In Matrimony.com & CUTS v Google, the CCI examined Google's presentation of search results.

The CCI found that Google's prominent display of its Commercial Flight Unit, linking users to Google's specialised flight service, constituted an abuse in the circumstances examined. It also considered Google's earlier treatment of Universal Results.

The case is important for India because it recognised that search-result design can have competitive consequences.

The CCI specifically observed that Google occupied a gateway position for users seeking information on the internet.

Principle

Product design and search-result presentation can constitute a competition-law concern when a dominant search platform uses them to disadvantage competing services.

Case 4 — Amazon Marketplace and Amazon Buy Box

European Commission — Amazon, 2022

The European Commission investigated two related areas:

1. Amazon's use of non-public marketplace seller data; and
2. the operation of the Buy Box.

The Commission's investigation examined whether Amazon could use seller data obtained through its marketplace to benefit Amazon Retail and whether the Buy Box favoured Amazon's own retail offers or sellers using Amazon's logistics services.

The Commission's 2022 commitments included restrictions on the use of non-public seller data and measures concerning the selection of offers displayed through the Buy Box.

The case illustrates an important extension of self-preferencing:

Preferential treatment does not necessarily occur only through visible ranking. It can also operate through data access and marketplace selection mechanisms.

Case 5 — Alibaba Group

Alibaba — SAMR, China, 2021

China's State Administration for Market Regulation investigated Alibaba's practice commonly described as “二选一” (choose one from two).

SAMR found that Alibaba had required merchants to choose Alibaba's platform and restricted their ability to operate or participate in promotional activities on competing platforms.

The authority concluded that Alibaba had abused its dominant position in China's online retail platform service market and used platform rules, market power, data and algorithms to reinforce the restrictions.

SAMR imposed a fine of RMB 18.228 billion, equivalent to 4% of Alibaba's 2019 domestic sales.

Although this is more directly an exclusive-dealing case than a classic ranking self-preferencing case, it is highly relevant to digital ecosystems because it demonstrates how a dominant platform can use ecosystem control to disadvantage competing platforms.

Principle

Platform rules, data and algorithms can reinforce exclusionary treatment where a dominant ecosystem restricts merchants from dealing with rival platforms.

Case 6 — Meituan

Meituan — SAMR, China, 2021

SAMR also investigated Meituan's conduct in China's online food-delivery platform market.

The authority found that Meituan used measures including:

• differential treatment of merchants;
• delayed onboarding;
• exclusive cooperation requirements;
• deposits;
• data and algorithmic tools;
• punitive measures.

SAMR concluded that the conduct amounted to abuse of dominant position through imposing exclusive dealing without legitimate justification.

The authority ordered Meituan to cease the conduct, return approximately RMB 1.289 billion in exclusive-cooperation deposits and imposed a RMB 3.442 billion fine.

Principle

The case shows how preferential or discriminatory treatment can operate through the rules governing participation in an ecosystem, rather than only through search ranking.

Case 7 — Google Android

Competition Commission of India — 2022

In its Android decision, the CCI examined Google's arrangements concerning Android mobile devices.

The CCI considered, among other things, Google's control over important search-entry points, pre-installation and distribution arrangements. It found that the arrangements gave Google's search services significant competitive advantages and contributed to barriers for competing services.

The CCI imposed a penalty of ₹1,337.76 crore and ordered behavioural remedies.

This case is particularly relevant to preferential treatment because ecosystem preference can be embedded before a consumer makes a search choice.

Principle

Pre-installation, default status and ecosystem-level contractual arrangements can create competitive advantages for an affiliated service.

7. Comparative Case-Law Table

8. The Economic Theory Behind Preferential Treatment

A. Leveraging

The platform has power in Market A and uses that position to strengthen its position in Market B.

Example:

Search dominance → preferential ranking → shopping dominance.

B. Foreclosure

If competitors cannot obtain comparable visibility or access, their ability to compete may decline.

The analysis therefore focuses on whether the conduct can:

• reduce rival traffic;
• increase rivals' customer-acquisition costs;
• reduce innovation;
• increase barriers to entry;
• reduce multi-homing;
• reinforce network effects.

C. Data advantage

Digital platforms often possess information that independent competitors cannot obtain.

For example:

Third-party sellers → provide marketplace data → marketplace operator observes demand → operator competes against sellers.

This may produce a significant informational advantage.

9. Network Effects

Network effects are particularly important.

Suppose:

More users

↓

More transactions

↓

More data

↓

Better algorithms

↓

Better user experience

↓

More users

A platform's preferential treatment can accelerate this feedback loop.

Consequently, a relatively small ranking advantage may become more significant over time.

10. The Role of Algorithms

Traditional competition law often examined explicit contractual discrimination.

Digital markets require examination of algorithmic discrimination.

Potential evidence includes:

• ranking changes;
• A/B testing;
• source-code documentation;
• internal algorithmic instructions;
• search-result logs;
• click-through rates;
• traffic allocation;
• recommendation data;
• internal emails;
• product-management documents;
• changes in ranking following acquisition or launch of an affiliated product.

An authority may therefore need to reconstruct the algorithm's actual competitive effects, rather than simply examine written contractual terms.

11. Objective Justification

Preferential treatment should not automatically be treated as unlawful.

A platform may argue that its own service receives preferential treatment because of:

• greater relevance;
• better quality;
• enhanced security;
• technical integration;
• fraud prevention;
• privacy protection;
• consumer convenience;
• lower transaction costs;
• system compatibility.

The legal question is whether the explanation is genuine, objectively justified and proportionate, rather than a pretext for exclusion.

The Google Shopping litigation is particularly important because it demonstrates the need to distinguish ordinary product improvement from exclusionary discrimination.

12. Consumer Harm

Preferential treatment may harm consumers indirectly.

Possible consequences include:

Reduced choice

Consumers see fewer competing services.

Higher prices

Reduced competitive pressure may permit higher prices.

Lower quality

Reduced rivalry can diminish incentives to improve products.

Reduced innovation

Potential entrants may conclude that competing against the platform's ecosystem is commercially unrealistic.

Less privacy competition

If privacy-friendly competitors cannot obtain sufficient scale, consumers may have fewer alternative business models.

13. Remedies

Competition authorities may impose several types of remedies.

A. Behavioural remedies

Examples:

• non-discriminatory ranking;
• equal access;
• prohibition of discriminatory algorithms;
• restrictions on use of competitor data.

B. Transparency remedies

Platforms may be required to explain:

• ranking criteria;
• access conditions;
• algorithmic changes;
• eligibility requirements.

C. Data separation

A platform may be prohibited from using non-public competitor data for its own competing business.

D. Choice mechanisms

Consumers may receive:

• default-choice screens;
• alternative search engines;
• alternative payment options;
• alternative app stores.

E. Structural remedies

In particularly serious cases, authorities may consider structural separation, although this is considerably more intrusive than behavioural regulation.

14. Shift From Ex Post Antitrust to Ex Ante Regulation

One of the most important developments is the emergence of ex ante digital-platform regulation.

Under traditional abuse-of-dominance law, an authority normally has to establish dominance and an abuse.

The EU Digital Markets Act goes further for designated gatekeepers.

For example, the EU has specifically addressed self-preferencing under its gatekeeper regime. In March 2024, the European Commission opened an investigation into Alphabet concerning preferential treatment of Google's own vertical search services.

More recently, in July 2026, the Commission announced a €460 million fine against Google for DMA violations concerning self-preferencing in Google Search, finding preferential treatment for Google's own shopping, hotel, transport and sports results.

Thus, the regulatory approach is increasingly moving from:

“Prove abuse after the conduct occurs”

toward:

“Impose specific conduct obligations on designated digital gatekeepers in advance.”

15. China: Importance of Differential Treatment

China has developed an especially detailed platform-economy approach.

The platform-economy antitrust guidance addresses 差别待遇 (differential treatment) and recognises that a dominant platform may use:

• big-data analysis;
• algorithms;
• different standards;
• different rules;
• different payment conditions;
• different transaction methods.

The guidance specifically considers differences based on users' payment ability, consumption preferences and usage habits.

The Supreme People's Court's subsequent judicial interpretation also addresses differential treatment and considers factors such as competitive effects, disadvantage to counterparties and consumer/public-interest effects.

China's 2026 Internet Platform Antitrust Compliance Guidelines similarly address discriminatory treatment based on algorithms and other platform mechanisms.

16. India: Competition Act Approach

Under Section 4 of the Competition Act, 2002, preferential treatment may potentially fall within several forms of abuse, depending on the facts:

• unfair or discriminatory conditions;
• unfair or discriminatory pricing;
• limiting or restricting markets;
• denial of market access;
• leveraging dominance;
• tying or bundling;
• exclusionary arrangements.

The Google search-bias and Android proceedings demonstrate that the CCI is prepared to examine digital architecture, ranking, defaults, pre-installation and ecosystem arrangements as part of competition analysis.

17. Distinguishing Legitimate Integration From Anti-Competitive Preference

The distinction depends heavily on market power, purpose/effects, evidence and justification.

18. Key Legal Tests

A useful analytical framework is:

Step 1 — Define the relevant market

Identify:

• platform market;
• upstream market;
• downstream market;
• complementary services.

Step 2 — Establish market power

Consider:

• market shares;
• network effects;
• switching costs;
• data advantages;
• entry barriers;
• multi-homing;
• consumer dependence.

Step 3 — Identify the preferential conduct

Ask:

What exactly receives preferential treatment?

Examples:

• ranking;
• price;
• access;
• data;
• API;
• payment;
• default status;
• recommendation;
• logistics;
• advertising.

Step 4 — Compare treatment

Compare:

platform's affiliated service

with

similarly situated third-party competitors.

Step 5 — Establish competitive effect

Examine:

• foreclosure;
• traffic diversion;
• increased rival costs;
• entry barriers;
• reduced innovation;
• reduced consumer choice.

Step 6 — Examine justification

Ask whether the preference is:

• legitimate;
• objectively justified;
• necessary;
• proportionate.

Step 7 — Consider remedy

Possible remedies include:

• non-discrimination;
• algorithmic transparency;
• data separation;
• access obligations;
• choice screens;
• interoperability;
• structural separation.

19. Important Doctrinal Insight

The central issue is not simply whether a platform prefers its own products.

Businesses normally have legitimate reasons to promote their own products.

The competition-law question is:

Has a platform with substantial gateway power used control over that gateway to confer an artificial competitive advantage on its own downstream business, thereby restricting effective competition?

This distinction prevents competition law from becoming a prohibition on ordinary vertical integration.

20. Emerging Issues

Preferential treatment is expanding beyond search and marketplaces into:

AI ecosystems

AI assistants may preferentially recommend their owner's:

• shopping services;
• advertising services;
• payment systems;
• cloud infrastructure;
• applications.

App stores

Platforms can potentially favour affiliated apps or payment systems.

Cloud computing

Cloud providers may favour their own software, databases or AI models.

Advertising technology

A vertically integrated adtech platform may favour its own advertising exchange or inventory.

Digital payments

Wallets and payment platforms can potentially favour affiliated financial services.

Smart devices

Operating systems can give affiliated services preferential:

• defaults;
• APIs;
• hardware integration;
• notification access.

Generative AI search

AI-generated answers may become a new form of digital ranking, making algorithmic self-preferencing increasingly important.

Conclusion

Preferential treatment is becoming a central issue in modern competition law because digital platforms frequently perform two roles simultaneously:

Gateway operator + competitor

The major cases—Google Shopping, Naver, Google Search Bias, Amazon Marketplace/Buy Box, Alibaba, Meituan and Google Android—demonstrate different legal pathways through which preferential treatment can become problematic.

The most important analytical factors are:

1. market power;
2. control over a digital gateway;
3. vertical integration;
4. algorithmic discrimination;
5. access to non-public data;
6. network effects and switching costs;
7. actual or likely foreclosure;
8. consumer and innovation effects;
9. objective justification; and
10. proportionality of remedies.