Competition Law and Feedback-Driven Market Dominance

1. Introduction

Feedback-driven market dominance describes a situation in which an undertaking's existing market position generates a self-reinforcing cycle that makes the position progressively stronger.

A typical digital-market feedback loop is:

More users → more transactions/interactions → more data → better algorithms/services → greater user attraction → more users → greater market power

The loop may also operate through sellers, advertisers, developers, or complementary services:

More users → more sellers → greater product variety → more users → more sellers

Competition law becomes particularly important when a dominant undertaking uses control over the feedback mechanism to exclude rivals, rather than merely benefiting from superior performance.

Indian competition jurisprudence has expressly recognised that large online platforms can benefit from positive feedback loops: in All India Online Vendors Association v. Flipkart Internet Pvt. Ltd., the CCI noted that network effects can increase a platform's value as its user base expands, thereby generating market power.

2. Meaning of Feedback-Driven Market Dominance

Feedback-driven dominance arises where several reinforcing factors operate simultaneously:

A. Network effects

The value of a platform increases as more users join it.

For example:

Users ↑ → Sellers ↑ → Product variety ↑ → Consumers ↑

This creates a positive feedback loop.

B. Data feedback

More users generate more behavioural data.

Users → searches/clicks/purchases → data → algorithmic improvement → better service → users

This is particularly significant for:

• search engines;
• social networks;
• e-commerce;
• online advertising;
• AI services;
• recommendation platforms;
• financial technology.

Academic analysis has specifically described this phenomenon as a positive feedback loop in which firms with more customers obtain more data, enabling them to improve their services and attract still more customers.

C. Economies of scale

A platform may spread fixed technological costs over an increasingly large user base.

D. Switching costs

Users may accumulate:

• transaction histories;
• contacts;
• ratings;
• playlists;
• purchased applications;
• seller reputation;
• stored data.

The larger the accumulated ecosystem, the harder it can become to switch.

E. Ecosystem effects

A platform may connect several markets:

Search → advertising → shopping → maps → payments → cloud → AI

Dominance in one market may therefore strengthen the undertaking in adjacent markets.

3. Why Feedback Loops Matter Under Competition Law

A feedback mechanism is not itself anti-competitive.

Competition law generally does not condemn a company merely because:

• it has many users;
• its algorithms improve with data;
• consumers prefer its service;
• it benefits from network effects; or
• it achieves economies of scale.

The legal concern arises where a dominant undertaking artificially reinforces the loop through exclusionary conduct.

Examples include:

1. self-preferencing;
2. tying and bundling;
3. exclusive dealing;
4. discriminatory access;
5. denial of interoperability;
6. discriminatory ranking;
7. data leveraging;
8. refusal to provide essential inputs;
9. restrictive API policies;
10. acquisition of emerging competitors;
11. contractual restrictions preventing multi-homing;
12. exploitative or exclusionary use of platform-generated data.

4. The Feedback-Loop Theory of Harm

A useful competition-law framework is:

Stage 1 — Initial advantage

The undertaking acquires a significant customer base.

↓

Stage 2 — Data/network accumulation

The enlarged user base produces additional data, sellers, advertisers or developers.

↓

Stage 3 — Service improvement

The undertaking uses these additional resources to improve:

• ranking;
• recommendations;
• targeting;
• pricing;
• fraud detection;
• search;
• logistics;
• AI models.

↓

Stage 4 — Further user acquisition

Improved service attracts additional users.

↓

Stage 5 — Rival disadvantage

Competitors have:

• fewer users;
• less data;
• fewer sellers;
• fewer advertisers;
• less liquidity;
• weaker network effects.

↓

Stage 6 — Entry barriers

Potential entrants cannot easily reproduce the accumulated ecosystem.

This can convert an ordinary competitive advantage into a self-reinforcing structural advantage.

5. Relevant Competition-Law Doctrines

A. Abuse of dominance

Under abuse-of-dominance rules, the central question is whether the dominant undertaking is using its position to restrict competition.

In India, relevant provisions include Sections 4(1) and 4(2) of the Competition Act, 2002.

Particularly relevant forms of conduct include:

• unfair/discriminatory conditions;
• denial of market access;
• tying;
• leveraging dominance;
• discriminatory treatment.

B. Self-preferencing

A platform may use the data and traffic generated by its marketplace to favour its own products.

Example:

Marketplace data → identification of profitable product → platform launches competing product → platform ranking advantage → rival sellers lose visibility

This can strengthen the feedback loop.

C. Data leveraging

Data can become an important competitive input.

Where a dominant platform possesses data unavailable to rivals, combining data across services may strengthen its competitive position.

The German Bundeskartellamt has expressly treated collection, processing and combination of data as an important source of market power for large digital firms.

D. Denial of market access

A dominant platform may restrict rivals' ability to reach users.

Examples:

• restricting APIs;
• limiting interoperability;
• preventing alternative payment systems;
• restricting data portability;
• manipulating rankings;
• imposing discriminatory access conditions.

E. Network-effect foreclosure

If a dominant undertaking prevents competitors from acquiring sufficient users, competitors may never reach the scale necessary to activate their own positive feedback loop.

Thus:

Dominant platform's network effect → rival's inability to scale → reduced competitive pressure → dominant platform's network effect becomes stronger

6. Case Laws

1. Google Shopping — European Commission / General Court

Case: Google Search (Shopping), Commission Decision AT.39740; General Court, Google and Alphabet v Commission, T-612/17.

Google was found to have systematically favoured its comparison-shopping service in general search results while demoting competing comparison-shopping services.

The General Court upheld the central infringement finding.

Relevance to feedback-driven dominance

Search traffic itself becomes a competitive resource.

The feedback mechanism can be represented as:

More searches → more search data → stronger search service → more searches

Google's control over search visibility could simultaneously affect competing comparison-shopping services because those services depended substantially on traffic from the search engine.

The EU's subsequent digital-market framework continues to treat self-preferencing as a significant competition concern.

Principle

A dominant platform's control over an important gateway can allow it to reinforce its position by favouring its own downstream service.

7. Google Android — Google and Alphabet v Commission

Case: T-604/18, Google and Alphabet v Commission, General Court, 14 September 2022.

The case concerned Google's practices involving Android, including contractual arrangements concerning search and app distribution.

Feedback-loop relevance

Android can be viewed as part of a wider ecosystem:

Android users → Google services → more data/use → stronger Google ecosystem → greater attractiveness of Android

Competition concerns therefore cannot always be analysed by looking at a single product in isolation.

The EU's current digital-market materials continue to recognise interconnected products and ecosystems as important characteristics of digital-platform markets.

Principle

Where products are technologically and commercially interconnected, competitive effects may arise from the interaction between multiple markets.

8. Facebook/Meta Data Case — Bundeskartellamt

Case: Bundeskartellamt proceedings concerning Facebook/Meta and the combination of user data.

The German competition authority examined Facebook's combination of data obtained from Facebook with data from other services and sources.

Feedback-loop relevance

The theory is:

More users → more data → more effective services/advertising → more users → more data

The authority considered the accumulation and combination of data to be capable of strengthening Facebook's market power.

The German authority has described the Facebook proceeding as an important example of competition enforcement involving the interaction between data and market power.

Principle

Data advantages can contribute to market power where they reinforce an existing ecosystem and create competitive disadvantages for rivals.

9. Amazon Marketplace — Bundeskartellamt

The Bundeskartellamt investigated several aspects of Amazon's marketplace practices.

Earlier proceedings concerned price-parity clauses, while later proceedings examined seller conditions, pricing mechanisms and Amazon's broader cross-market significance.

The authority ultimately determined that Amazon possessed paramount significance for competition across markets under Section 19a GWB; that finding was upheld by Germany's Federal Court of Justice in 2024.

Feedback-loop relevance

Amazon's ecosystem can operate through:

Consumers → sellers → product selection → consumers

and:

Transactions → marketplace data → improved platform operations → more consumers/transactions

The platform can therefore occupy multiple roles:

• marketplace;
• retailer;
• logistics provider;
• advertising provider;
• cloud provider;
• service provider.

Principle

A platform's ecosystem-wide significance may be relevant when assessing the competitive consequences of individual practices.

10. Amazon Buy Box / Amazon Marketplace — European Commission

The European Commission's Amazon proceedings examined the use of marketplace data and the relationship between Amazon's marketplace activities and its own retail operations.

Feedback-loop relevance

Third-party sellers generate information through marketplace activity.

The competitive concern is potentially:

Seller activity → commercially valuable data → platform insight → platform's own retail activity → stronger platform position

This demonstrates why data generated within a platform can have competitive significance beyond the immediate transaction.

EU competition materials have specifically discussed network effects and the role of marketplace/search ecosystems in cases involving Amazon and Google.

Principle

A vertically integrated platform may face competition concerns where its control of an intermediary platform generates informational advantages in downstream competition.

11. All India Online Vendors Association v. Flipkart Internet Pvt. Ltd.

CCI, 2021

This Indian case is particularly relevant to the concept.

The allegations concerned practices involving:

• exclusive launches;
• preferred sellers;
• deep discounting;
• preferential listing/promotion;
• private labels.

The CCI expressly recognised that strong network effects can generate market power because a larger user base can make a platform more valuable and thereby attract still more users—a positive feedback loop.

Principle

Network effects are an important consideration in assessing the competitive characteristics of digital platforms.

However, the existence of a positive feedback loop does not automatically establish abuse. The specific conduct and its effect on competition must still be examined.

12. Meta Platforms Inc. v Competition Commission of India

Competition Appeal No. 1 & 2 of 2025, Supreme Court proceedings

The dispute concerns, among other matters, the relationship between WhatsApp's position in messaging and Meta's broader ecosystem and data practices.

The record discusses the possibility that data obtained through WhatsApp can strengthen Meta's position in online advertising and create barriers for rivals that lack comparable data access.

Feedback-loop relevance

The theory can be expressed as:

Messaging users → user data → ecosystem advertising capabilities → advertising advantage → stronger ecosystem → continued user attraction

Principle

In ecosystem markets, market power in one service can potentially be leveraged into related markets where the undertaking possesses informational, technological or distribution advantages.

13. Google AdSense

European Commission, AT.40411 — Google Search (AdSense)

Google's online-search advertising practices provide another example of how a dominant platform can control an important interface between advertisers and publishers.

The broader feedback mechanism is:

Advertisers → advertising revenue → platform investment → user engagement → more advertising inventory → advertisers

Restrictive contractual practices affecting competing advertising channels can potentially make it harder for rivals to develop sufficient scale.

The EU's subsequent treatment of digital ecosystems expressly identifies Google Search, advertising and interconnected markets as relevant examples of network and ecosystem effects.

Principle

Control over a key intermediary can allow a platform to reinforce its position across connected sides of a multi-sided market.

14. Legal Test for Feedback-Driven Dominance

A competition authority or court can examine the following questions.

Step 1 — Relevant market

Identify:

• product market;
• geographic market;
• platform sides;
• complementary markets.

Step 2 — Dominance

Consider:

• market share;
• network effects;
• switching costs;
• entry barriers;
• data advantages;
• ecosystem integration;
• economies of scale;
• countervailing buyer power.

Step 3 — Identify the feedback mechanism

Determine whether the undertaking benefits from:

users → data → improved service → users

or:

users → sellers → variety → users

or:

transactions → liquidity → participants → transactions

Step 4 — Identify the intervention

Ask whether the dominant undertaking has:

• restricted interoperability;
• excluded rivals;
• self-preferenced;
• tied products;
• discriminated;
• restricted data access;
• imposed exclusivity;
• manipulated rankings;
• used contractual restrictions.

Step 5 — Competitive effects

Examine whether the conduct:

• forecloses rivals;
• raises entry barriers;
• reduces multi-homing;
• prevents rival scale;
• reduces innovation;
• increases switching costs;
• transfers dominance to adjacent markets.

Step 6 — Counterfactual

Compare the actual market with a plausible competitive scenario without the allegedly exclusionary conduct.

15. Feedback Loops and the "Tipping" Problem

One major concern is market tipping.

A market may initially contain several competitors:

A ↔ B ↔ C

But network effects can eventually create:

A → more users → more data → better service → more users

If B and C cannot reach comparable scale, competition may weaken substantially.

This is particularly important in:

• social networking;
• search;
• marketplaces;
• payment systems;
• app stores;
• digital advertising;
• AI platforms;
• cloud ecosystems.

German policy analysis has specifically identified the self-reinforcing relationship between economies of scale, network effects and data access as a competition challenge in digital markets.

16. Feedback Loops and Multi-Sided Markets

Feedback effects can operate simultaneously across several sides.

Example: E-commerce

Consumers ↑

↓

Sellers ↑

↓

Product variety ↑

↓

Consumers ↑

↓

Transactions ↑

↓

Marketplace data ↑

↓

Better recommendations/advertising ↑

↓

Consumers ↑

This creates multiple interacting feedback loops.

Consequently, traditional market-share analysis may not fully capture the competitive significance of the platform.

17. Data Feedback and AI Markets

The concept becomes particularly important with generative AI.

Potential loop:

Users → prompts/feedback → training/evaluation data → better model → more users → more feedback

A second loop may arise:

More users → more developers → more applications → greater ecosystem value → more users

Competition concerns may arise if a dominant undertaking:

• restricts access to essential APIs;
• prevents interoperability;
• bundles AI with an incumbent platform;
• uses proprietary data unavailable to rivals;
• discriminates against competing AI services;
• uses an operating system to preference its own AI assistant.

The European Commission in 2026 issued measures concerning access to Google Search data for competing search engines and interoperability involving AI services on Android, illustrating the regulatory relevance of access to data and platform functionality in feedback-driven markets.

18. Remedies

Potential competition-law remedies include:

Structural remedies

• divestiture;
• separation of business units;
• restrictions on acquisitions.

Behavioural remedies

• non-discrimination;
• transparent ranking;
• prohibition of self-preferencing;
• data portability;
• interoperability;
• API access;
• restrictions on tying;
• restrictions on exclusive dealing.

Data-related remedies

• user choice;
• data portability;
• interoperability;
• data-access obligations where legally justified;
• restrictions on combining datasets.

Procedural remedies

• algorithmic transparency;
• independent monitoring;
• compliance reporting;
• audit mechanisms.

The EU's Digital Markets Act now provides ex ante obligations for designated gatekeepers, while Germany's Section 19a GWB permits enhanced intervention concerning undertakings of paramount significance across markets. The German authority has identified Alphabet, Amazon, Apple, Meta and Microsoft as subject to this enhanced framework.

19. Distinguishing Legitimate Feedback From Anticompetitive Feedback

The key distinction is therefore not whether a feedback loop exists, but how the loop is generated, maintained and exploited.

20. Important Competition-Law Principles Emerging From the Cases

Principle 1 — Network effects can strengthen market power

A growing user base can make an established platform increasingly difficult to challenge.

Principle 2 — Data can be a competitive advantage

Large-scale data accumulation can reinforce service quality and market position.

Principle 3 — Ecosystem dominance matters

Competition authorities increasingly examine interactions between connected markets rather than isolated products.

Principle 4 — Self-preferencing can reinforce feedback loops

Preferential treatment of the dominant firm's own services can divert traffic and data away from rivals.

Principle 5 — Market access is critical

If competitors cannot obtain sufficient users, data, interoperability or distribution, they may never achieve the scale necessary to compete.

Principle 6 — Feedback effects do not automatically equal abuse

A positive feedback loop is generally a market characteristic, not by itself an infringement.

21. Short Hypothetical

Suppose Platform X operates a dominant agricultural marketplace.

It receives:

• farmer purchasing data;
• crop information;
• machinery usage information;
• supplier prices;
• transaction histories.

X then uses that information to develop its own agricultural-input business.

The cycle becomes:

Farmers → transactions → data → better prediction → more farmers → more transactions → more data

If X additionally:

• gives its own products preferential rankings;
• restricts competing sellers' access to data;
• blocks interoperable APIs;
• imposes exclusivity;
• penalises farmers who use competing platforms,

the feedback loop may become an important theory of competitive harm.

The legal inquiry would then focus on dominance, the specific exclusionary conduct, foreclosure effects, efficiencies and consumer impact.

22. Conclusion

Feedback-driven market dominance is one of the central competition-law issues in digital and data-intensive markets. Network effects, data accumulation, economies of scale, switching costs and ecosystem integration can create self-reinforcing competitive advantages.

The principal legal concern arises when a dominant undertaking controls or manipulates the feedback mechanism to prevent rivals from reaching sufficient scale.

The leading cases—particularly Google Shopping, Google Android, Facebook/Meta, Amazon Marketplace, Google AdSense and All India Online Vendors Association v. Flipkart—illustrate different dimensions of the problem.

The central analytical formula is:

Existing market power → users/data/network advantages → improved ecosystem → additional users → greater data/network advantages → stronger market power

Competition law therefore increasingly examines not merely who has market share today, but also whether the structure of the market permits rivals to generate their own competitive feedback loop tomorrow.