1. Introduction

Ecosystem modelling is an analytical approach used in modern competition law to examine how a powerful firm operates not merely in one relevant market, but across a network of interconnected products, services, technologies, data resources, distribution channels and user groups.

Traditional competition analysis often asks:

What is the relevant market, who are the competitors, and what is the firm's market share?

Ecosystem analysis asks a broader question:

How does the firm's position in one part of a commercial ecosystem strengthen its position in other markets, and does that interconnected structure create or reinforce durable competitive advantages?

This is particularly important for digital platforms because an operating system may be connected to an app store, search engine, advertising business, browser, payments system, cloud service, data infrastructure and AI service.

The European Commission and courts have increasingly recognised the relevance of the "ecosystem" concept, particularly in the Google Android litigation. The General Court expressly discussed multi-sided platforms and the concept of an ecosystem, while the 2026 CJEU judgment continued to examine the interaction between Android, Google Search, Chrome and Play Store.

2. Meaning of an Ecosystem in Competition Law

An ecosystem may be understood as:

A group of economically interconnected products, services, platforms, technologies and complementary activities controlled or coordinated by one or more firms, where competitive conditions in one component influence competitive conditions in other components.

For example:

Google ecosystem

                         GOOGLE                            │          ┌─────────────────┼─────────────────┐          │                 │                 │       Android           Search            Chrome          │                 │                 │      Play Store         Ads/Data         Web users          │       Apps          │    Developers          │    Consumers

The competitive significance is not simply that Google has a large share in Android or search.

The deeper question is whether:

Android → Play Store → Search → Chrome → Data → Advertising → AI

creates reinforcing feedback effects that make entry into one or more markets substantially more difficult.

3. Ecosystem Modelling versus Traditional Market Analysis

Traditional competition analysis generally proceeds through:

1. Relevant market definition.
2. Market shares.
3. Market power.
4. Conduct.
5. Effects.
6. Efficiency or justification.
7. Remedy.

Ecosystem modelling adds another layer:

1. Identify ecosystem components.
2. Identify relationships between components.
3. Identify dependencies.
4. Measure network effects.
5. Identify cross-market leverage.
6. Model feedback loops.
7. Assess foreclosure.
8. Examine switching costs.
9. Analyse entry barriers.
10. Determine whether competitive advantages become self-reinforcing.

Thus:

4. Why Ecosystem Modelling Matters

Ecosystem modelling is especially important where competition is affected by:

• network effects;
• economies of scope;
• data advantages;
• interoperability;
• switching costs;
• default settings;
• bundling;
• tying;
• self-preferencing;
• exclusive agreements;
• app-store restrictions;
• technical restrictions;
• interoperability limitations;
• cross-subsidisation;
• vertical integration;
• acquisitions of complementary firms.

The principal concern is cumulative market power.

A firm might not have overwhelming dominance in every individual market but may nevertheless possess considerable power because its activities reinforce one another.

5. Core Components of an Ecosystem Model

A. Nodes

The first stage is to identify the different components of the ecosystem.

For example:

• operating system;
• search engine;
• browser;
• app store;
• payment service;
• advertising platform;
• cloud service;
• hardware;
• AI assistant;
• marketplace;
• social network.

Each is a node.

B. Links

The second stage identifies the relationships between nodes.

For example:

Operating system → App store

The operating system may determine access to the app store.

App store → Developers

The app store controls developer distribution.

Search → Advertising

Search activity generates data and advertising opportunities.

Hardware → Operating system

Hardware may provide the distribution mechanism for the software ecosystem.

These relationships determine how power can be transmitted across the ecosystem.

6. Ecosystem Dependency

A central metric is dependency.

Suppose 70% of app developers depend upon one app store for distribution.

The app-store operator therefore possesses influence beyond the app-store market itself.

The analytical question becomes:

What happens to competition in adjacent markets if access to the ecosystem is restricted?

This can be represented as:

Dependency = ecosystem users relying on platform ÷ total relevant ecosystem users

High dependency increases the potential competitive significance of platform conduct.

7. Network Effects

Network effects occur when the value of a platform increases as participation increases.

Direct network effect

More users → greater platform value.

Indirect network effect

More users → more developers → more applications → greater consumer value → more users.

For example:

More consumers      ↓ More developers      ↓ More applications      ↓ More consumer value      ↓ More consumers

This creates a feedback loop.

Ecosystem modelling asks whether the feedback loop becomes so strong that competitors cannot realistically reproduce it.

8. Cross-Market Network Effects

Ecosystem power becomes particularly significant when network effects operate across markets.

For example:

Search users     ↓ Search data     ↓ Better search results     ↓ More users     ↓ More advertising inventory     ↓ More advertising revenue     ↓ Investment in ecosystem     ↓ Better products

This creates a reinforcing competitive advantage.

Consequently, competition authorities should not necessarily examine search, advertising, browser and operating-system markets as completely isolated structures.

9. Ecosystem Modelling and Relevant Market Definition

Ecosystem analysis does not automatically eliminate traditional relevant-market analysis.

Relevant markets remain important under competition law.

However, the ecosystem may affect how the market is understood.

For example, an investigation might identify:

• Market A: mobile operating systems;
• Market B: app stores;
• Market C: general search;
• Market D: mobile advertising;
• Market E: browsers.

The authority can then examine whether conduct in Market A affects competition in B–E.

This is sometimes described as cross-market leveraging.

10. Ecosystem Power and Leverage

Suppose a company has substantial power in Market A.

It may use that position to strengthen its position in Market B.

The basic model is:

Market A dominance → leverage mechanism → Market B foreclosure → stronger ecosystem

Examples include:

• tying;
• bundling;
• exclusive pre-installation;
• default agreements;
• technical restrictions;
• preferential ranking;
• discriminatory access;
• interoperability restrictions.

The concern is not simply that the firm competes aggressively.

The concern is whether its conduct changes the competitive structure of an adjacent market.

11. Ecosystem Entrenchment

Ecosystem modelling is particularly useful for analysing entrenchment.

Entrenchment occurs when existing ecosystem advantages reinforce one another over time.

For example:

Large user base      ↓ More data      ↓ Better product      ↓ More developers      ↓ More applications      ↓ Higher switching costs      ↓ Fewer users leave      ↓ Larger user base

The result can be a self-reinforcing ecosystem.

This creates an important distinction between:

Market share today

and

ability to preserve market power tomorrow.

12. Six Major Case Laws

Case 1: Google Android — Google LLC and Alphabet Inc. v European Commission

Case T-604/18; subsequently C-738/22 P

This is perhaps the most important modern authority for ecosystem modelling.

The case concerned Google's Android ecosystem and its relationships among:

• Android operating system;
• Google Play Store;
• Google Search;
• Chrome;
• mobile-device manufacturers;
• mobile network operators.

The General Court expressly referred to the concepts of multi-sided platform and ecosystem.

The Commission had identified several forms of conduct, including:

• tying;
• exclusivity payments;
• anti-fragmentation obligations.

The importance of the case is that the competitive assessment could not sensibly be understood by examining Android independently from the other Google services connected to it.

Ecosystem lesson

A platform can use control over one technological layer to influence competition at other layers.

Competition-law significance

The case demonstrates the importance of analysing:

OS → App Store → Search → Browser → Developers → Users

as an interconnected competitive structure.

The CJEU's 2026 judgment continued to examine the contractual restrictions, tying, exclusivity payments, anti-fragmentation obligations and exclusionary effects involved in Google's Android arrangements.

13. Case 2: Google Shopping — Google Search (Shopping)

Google and Alphabet v European Commission

The Google Shopping litigation concerned Google's treatment of its own comparison-shopping service within general search results.

The essential ecosystem problem was:

General Search → Search ranking → Shopping service → Traffic

Google's position in general search could therefore affect competitive conditions in comparison shopping.

Ecosystem modelling lesson

A dominant platform may possess a distribution advantage that competitors cannot reproduce.

The important economic question is not merely:

Does Google compete with comparison-shopping services?

It is:

Does Google's control over the principal access point to consumers allow it to advantage its own downstream service?

Principle

Control of an important ecosystem gateway can have competitive effects in adjacent markets.

14. Case 3: Microsoft — Microsoft/Commission

Case T-201/04, Microsoft Corp. v Commission

Microsoft is one of the foundational cases for modern ecosystem analysis.

The Commission examined Microsoft's dominance in PC operating systems and its conduct concerning:

• interoperability information;
• work-group server operating systems;
• Windows Media Player.

The case illustrates how control over a foundational technological layer can be used to affect adjacent markets.

Ecosystem structure

Windows operating system          ↓ Software compatibility          ↓ Application developers          ↓ Users          ↓ Adjacent software markets

Microsoft's operating-system position provided strategic advantages in neighbouring markets.

Ecosystem lesson

A firm controlling an essential technological interface may have the ability to influence downstream or complementary markets.

15. Case 4: United States v Microsoft Corp.

The U.S. Microsoft case is another foundational authority.

The case involved Microsoft's Windows operating-system monopoly and conduct affecting browser competition.

The browser was not simply an independent product.

Its competitive position was affected by:

• operating-system distribution;
• default arrangements;
• technical integration;
• distribution advantages.

Ecosystem modelling insight

The case demonstrates why competition authorities must analyse distribution architecture, rather than merely comparing the standalone quality of products.

A theoretically superior competitor may still struggle if it lacks access to the dominant ecosystem's distribution channels.

16. Case 5: Qualcomm — Qualcomm (Commission Decision)

The European Commission's Qualcomm decision concerning exclusivity payments illustrates another ecosystem mechanism.

A dominant firm can potentially use commercial arrangements to limit rivals' access to strategically important customers.

In an ecosystem model, exclusivity can be represented as:

Platform power → exclusive arrangement → reduced rival access → reduced scale → higher rival costs → weaker competitive constraint

The significance is therefore dynamic.

The question is not merely:

How many customers are currently excluded?

It is:

Does exclusion prevent competitors from reaching the scale necessary to compete effectively?

17. Case 6: Intel v Commission

Case C-413/14 P, Intel Corp. v Commission

Intel is important for analysing exclusionary rebates and competitive effects.

The CJEU required careful consideration of whether rebates were capable of producing exclusionary effects, particularly where the dominant firm relied upon an economic assessment.

Ecosystem relevance

In a broader ecosystem, rebates can interact with:

• distribution relationships;
• customer dependence;
• scale economies;
• switching costs;
• network effects.

Thus, the competitive assessment may need to consider how contractual incentives affect the entire competitive structure, rather than looking solely at the nominal discount.

18. Case 7: Apple — App Store / Digital Platform Investigations

Apple-related competition proceedings provide a particularly useful ecosystem model because Apple's activities connect:

iPhone  ↓ iOS  ↓ App Store  ↓ Developers  ↓ Consumers  ↓ Payments  ↓ Digital services

The App Store therefore operates not merely as a distribution service but as a gateway into a broader technological ecosystem.

The European Commission's current Digital Markets Act framework expressly designates Apple among the gatekeepers subject to ecosystem-oriented obligations, alongside Alphabet, Amazon, Booking, ByteDance, Meta and Microsoft.

Ecosystem lesson

Control over an intermediary layer can generate significant power over downstream businesses.

19. Case 8: Amazon Marketplace

Amazon provides another important ecosystem model.

Its ecosystem can be represented as:

Consumers    ↓ Amazon Marketplace    ↓ Third-party sellers    ↓ Transaction data    ↓ Amazon logistics    ↓ Amazon advertising    ↓ Prime membership

The competitive concern is that control of the marketplace may give Amazon access to information and commercial advantages that influence competition between Amazon and third-party sellers.

This illustrates the concept of platform-as-infrastructure.

The platform may simultaneously act as:

1. intermediary;
2. competitor;
3. advertiser;
4. logistics provider;
5. data collector;
6. payment intermediary.

That combination is particularly important for ecosystem modelling.

20. Mathematical Ecosystem Model

An ecosystem can be represented using a network model.

Let:

E = {N, L}

where:

• N = nodes;
• L = links between nodes.

Suppose the ecosystem contains:

• operating system;
• app store;
• search engine;
• advertising service;
• cloud service.

Each node has a degree of market power.

We can define an illustrative ecosystem-power score:

EP = Σ(Mᵢ × Dᵢ × Cᵢ × Nᵢ)

where:

• Mᵢ = market power of node i;
• Dᵢ = dependency on node i;
• Cᵢ = connectivity with other ecosystem components;
• Nᵢ = network-effect intensity.

This is not a statutory legal test.

It is an economic modelling framework for organising evidence.

21. Ecosystem Centrality

A particularly useful concept is centrality.

A node with high centrality connects many other components.

For example:

              Search                │                │ Apps ─────── Android ───── Hardware                │                │            Play Store                │             Payments

Android may therefore have greater strategic importance than a product with an equivalent market share but fewer ecosystem connections.

Possible measures include:

• degree centrality;
• betweenness centrality;
• network density;
• dependency ratios.

22. Switching Costs in Ecosystem Modelling

Switching costs should be measured across the ecosystem, not merely at the product level.

A consumer changing phones might have to transfer:

• applications;
• subscriptions;
• passwords;
• cloud storage;
• payment information;
• purchased content;
• data;
• accessories;
• contacts;
• device integrations.

Therefore:

Total switching cost = product switching cost + ecosystem switching cost

This can substantially increase barriers to entry.

23. Data as an Ecosystem Asset

Data may operate as a connecting resource.

For example:

Users ↓ Search queries ↓ Data ↓ Algorithm improvement ↓ Better service ↓ More users

Data can therefore produce feedback effects.

Competition analysis should ask:

1. What data does the firm possess?
2. How frequently is the data generated?
3. Can competitors obtain equivalent data?
4. Is the data proprietary?
5. Does data improve product quality?
6. Does the advantage compound over time?

24. Ecosystem Expansion

Ecosystem modelling is particularly important when a dominant firm expands into adjacent markets.

Expansion may be:

Pro-competitive

A firm enters a new market because:

• it innovates;
• lowers costs;
• improves interoperability;
• introduces new products.

Potentially exclusionary

Expansion may become problematic where the firm:

• leverages dominance;
• forecloses rivals;
• restricts interoperability;
• ties products;
• discriminates against rivals;
• uses proprietary data to disadvantage competitors.

Thus, ecosystem expansion is not inherently anticompetitive.

The legal question is the mechanism and effect.

25. Ecosystem Foreclosure

Foreclosure occurs when rivals are denied meaningful access to an important ecosystem component.

The model can be represented as:

Gateway control → restricted access → reduced rival scale → increased rival costs → reduced innovation → ecosystem entrenchment

This is particularly relevant where the gateway is difficult to bypass.

Examples include:

• app stores;
• operating systems;
• payment systems;
• search engines;
• marketplaces;
• cloud infrastructure.

26. Ecosystem Modelling in Merger Analysis

Ecosystem analysis is increasingly important in mergers.

Traditional merger analysis may ask:

Will Company A and Company B compete less after the merger?

Ecosystem analysis additionally asks:

Will the acquisition strengthen an existing ecosystem and make future competition less likely?

This is especially important for:

• nascent competitors;
• complementary products;
• data-rich firms;
• AI companies;
• interoperability technologies;
• cloud services;
• cybersecurity;
• digital advertising.

27. Killer Acquisitions and Ecosystem Modelling

Suppose a dominant platform acquires a small innovative firm.

The target may have:

• low current revenue;
• low market share;
• substantial future potential.

Traditional market-share analysis may underestimate the transaction.

Ecosystem modelling asks:

Would the target have become an important independent ecosystem node?

If yes, acquisition could eliminate a potential competitive pathway.

28. Ecosystem Innovation Effects

Competition law should examine innovation dynamically.

A platform may appear highly efficient today but reduce innovation competition if it prevents rivals from developing complementary products.

The relevant question becomes:

Does ecosystem control increase innovation or reduce the possibility of independent innovation?

This is a dynamic-efficiency analysis.

29. Ecosystem Modelling and Self-Preferencing

Self-preferencing occurs when a platform favours its own downstream service.

Example:

Platform   ↓ Ranking algorithm   ↓ Own service ← preferential treatment   ↓ Consumer traffic

The competitive concern arises because rivals depend on the platform while the platform itself competes with them.

The European Commission's 2026 DMA enforcement against Google provides a contemporary example: the Commission found Google non-compliant regarding preferential treatment for its own services in Google Search.

30. Ecosystem Modelling and Interoperability

Interoperability is crucial because it determines whether consumers and businesses can move between ecosystems.

Low interoperability:

Ecosystem A → high switching costs → limited migration → stronger lock-in

High interoperability:

Ecosystem A ↔ Ecosystem B → lower switching costs → greater competitive pressure

The European Commission's 2026 DMA measures concerning Google's Android ecosystem illustrate this modern concern: the Commission required measures aimed at enabling competitors' AI services to access relevant Android functionalities and improving access to certain Google Search data.

31. Ecosystem Lock-In Model

An ecosystem lock-in index can conceptually be represented as:

LI = S + D + N + I

where:

• S = switching costs;
• D = dependency;
• N = network effects;
• I = interoperability restrictions.

The higher the combined score, the greater the potential for ecosystem entrenchment.

Again, this is an analytical tool rather than a statutory formula.

32. Counterfactual Analysis

One of the most important aspects of ecosystem modelling is the counterfactual.

The authority should ask:

What would the competitive structure look like if the challenged conduct had not occurred?

For example:

Actual world

Google controls:

Android → Play Store → Search → Chrome.

Counterfactual

Android remains available, but:

• alternative search engines receive greater access;
• alternative app stores can compete;
• manufacturers can distribute rival services;
• users face lower switching costs.

The difference between these two worlds helps identify the competitive effect of the conduct.

The Google Android litigation itself involved analysis of exclusionary effects and counterfactual issues.

33. Ecosystem Modelling and Consumer Welfare

Ecosystem analysis should ultimately connect to competition-law objectives.

Possible harms include:

Price

Higher prices or commissions.

Quality

Reduced service quality.

Innovation

Reduced innovation incentives.

Choice

Fewer competing products.

Privacy

Potential deterioration in privacy conditions.

Access

Reduced access for businesses or developers.

Entry

Higher barriers to entry.

Therefore, ecosystem analysis should not become an abstract study of corporate size.

It must demonstrate a competition-relevant effect.

34. Important Distinction: Ecosystem ≠ Automatic Dominance

The existence of an ecosystem does not automatically establish:

• dominance;
• monopoly;
• abuse;
• foreclosure;
• anticompetitive intent.

A large ecosystem may result from:

• superior innovation;
• economies of scale;
• consumer preference;
• legitimate integration;
• efficient distribution;
• technological compatibility.

Competition authorities must therefore distinguish between:

efficient ecosystem integration

and

strategic ecosystem foreclosure.

35. Practical Ecosystem Competition Test

A competition authority can use the following sequence:

Step 1 — Map the ecosystem

Identify all major products and services.

Step 2 — Identify gateways

Determine which components control access.

Step 3 — Measure dependencies

Identify consumers, developers and businesses dependent upon those gateways.

Step 4 — Measure network effects

Determine whether scale creates reinforcing advantages.

Step 5 — Identify cross-market leverage

Determine whether power in one market can affect another.

Step 6 — Identify conduct

Examine:

• tying;
• bundling;
• exclusivity;
• self-preferencing;
• discriminatory access;
• interoperability restrictions.

Step 7 — Model foreclosure

Estimate how rivals' ability to compete changes.

Step 8 — Analyse counterfactual

Compare the actual ecosystem with a competitive alternative.

Step 9 — Examine efficiencies

Consider:

• integration;
• security;
• innovation;
• quality;
• cost savings.

Step 10 — Determine remedy

Potential remedies include:

• interoperability;
• access obligations;
• non-discrimination;
• data portability;
• restrictions on self-preferencing;
• separation of functions;
• structural remedies in exceptional cases.

36. Indian Competition-Law Perspective

Ecosystem modelling is particularly relevant to Section 4 of the Competition Act, 2002, which prohibits abuse of dominant position.

Potentially relevant forms of conduct include:

• unfair or discriminatory conditions;
• denial of market access;
• leveraging;
• tying;
• discriminatory treatment;
• practices resulting in foreclosure.

Indian digital-market enforcement has increasingly required examination of interconnected platform activities rather than only conventional single-product markets.

For Indian analysis, ecosystem modelling can therefore complement:

• relevant-market definition;
• dominance analysis;
• AAEC assessment;
• foreclosure analysis;
• network-effect analysis;
• data analysis;
• entry-barrier analysis.

37. Overall Analytical Framework

A sophisticated ecosystem competition model can therefore be expressed as:

                    ECOSYSTEM                        │        ┌───────────────┼───────────────┐        ↓               ↓               ↓     Products          Data          Infrastructure        │               │               │        └───────────────┼───────────────┘                        ↓                 Network Effects                        ↓                  User Dependency                        ↓                 Switching Costs                        ↓                  Market Power                        ↓              Cross-Market Leverage                        ↓                  Foreclosure                        ↓             Ecosystem Entrenchment                        ↓             Reduced Competition?                        ↓                 Legal Assessment

38. Key Principles Emerging from the Case Law

The case law collectively supports several important propositions.

1. Market boundaries remain important

Ecosystem analysis supplements rather than necessarily replaces relevant-market analysis.

2. Technological architecture matters

The structure of the platform can determine how market power is exercised.

3. Distribution is competitive power

Control over consumer access can affect adjacent markets.

4. Network effects can reinforce dominance

Large ecosystems may become increasingly difficult to challenge.

5. Interoperability is strategically important

Restrictions on interoperability may increase ecosystem lock-in.

6. Cross-market effects matter

Conduct in one market can have competitive consequences in another.

7. Dynamic effects should be considered

Competition authorities should consider future entry and innovation.

8. Ecosystem power is not automatically unlawful

The legal focus remains on dominance, conduct, effects and applicable statutory standards.

39. Conclusion

Ecosystem modelling represents an important evolution in competition analysis.

Traditional competition law often treats markets as relatively discrete units. Digital and technology-driven businesses increasingly operate differently: an operating system may support an app store; the app store may support payments and advertising; user activity may generate data; data may improve services; improved services may attract more users; and the resulting scale may make the ecosystem more difficult for rivals to challenge.

The central analytical insight is therefore:

Competition may be determined not merely by who controls a particular market, but by who controls the connections between markets.

The Google Android litigation is particularly significant because the EU courts expressly confronted the relationship between multi-sided platforms, interconnected products and the concept of an ecosystem. Microsoft, Google Shopping, Intel, Qualcomm and platform cases involving Apple and Amazon further illustrate how distribution, interoperability, exclusivity, vertical integration and technological gateways can affect competitive conditions.

Modern ecosystem analysis should therefore combine market definition + network economics + dependency analysis + switching costs + data analysis + interoperability + counterfactual modelling + foreclosure analysis.