Competition Law and Strategic Data Asset Concentration

1. Introduction

Strategic data asset concentration refers to a situation in which a firm acquires, accumulates, controls, or combines large quantities of commercially valuable data in a manner that may strengthen its market power or make effective competition more difficult.

Modern competition law increasingly treats data as a strategic competitive resource, particularly where data provide advantages in:

• improving algorithms and artificial intelligence;
• targeted advertising;
• search and recommendation systems;
• consumer profiling;
• credit assessment;
• pricing and demand forecasting;
• platform matching;
• personalization;
• fraud detection;
• cloud and digital services;
• training machine-learning models; and
• developing complementary products and services.

Data concentration does not automatically constitute an antitrust violation. Competition authorities generally examine whether control over data creates or reinforces market power and whether that power is used to exclude competitors, exploit users, foreclose markets, or impede innovation.

2. Meaning of Strategic Data Asset Concentration

Strategic data asset concentration can arise through several mechanisms.

A. Organic accumulation

A dominant platform may continuously collect:

• search queries;
• transaction histories;
• location information;
• browsing behaviour;
• purchase histories;
• social interactions; and
• device information.

The resulting dataset can become difficult for competitors to replicate.

B. Acquisition of data-rich firms

A company may acquire another business primarily because the target possesses:

• unique consumer datasets;
• proprietary datasets;
• transaction histories;
• behavioural information;
• valuable APIs;
• industry-specific databases; or
• datasets necessary for AI development.

This creates a potential data-driven theory of harm in merger control.

C. Data combination

Competition concerns may arise when two previously separate datasets are combined.

For example:

Company A possesses search data, while Company B possesses purchasing data. Combining both datasets may create a substantially more comprehensive consumer profile.

D. Exclusive data arrangements

A dominant undertaking may negotiate agreements preventing suppliers, customers, or partners from sharing data with competing platforms.

E. Data interoperability restrictions

A platform may prevent competitors from obtaining data through:

• API restrictions;
• data portability limitations;
• technical barriers;
• contractual restrictions; or
• discriminatory access conditions.

3. Why Data Can Become a Strategic Asset

Data possess several economic characteristics relevant to competition law.

3.1 Economies of scale

Large datasets can improve:

• recommendation accuracy;
• advertising targeting;
• fraud detection;
• search relevance; and
• machine-learning performance.

3.2 Economies of scope

A dataset collected for one service may be useful for another service.

For example:

payment data → consumer profile → advertising → credit assessment → personalized offers

This creates opportunities for cross-market leveraging.

3.3 Network effects

More users generate more data, which can improve the service, attracting more users and generating still more data.

This can create a feedback loop:

Users → Data → Better service → More users → More data

3.4 Replication difficulties

A competitor may technically be capable of creating an alternative service but lack the historical data necessary to reproduce the incumbent's performance.

3.5 Data uniqueness

Not every large dataset is competitively significant.

The important questions include:

• Is the data unique?
• Is it commercially valuable?
• Is it difficult to reproduce?
• Is it continuously updated?
• Can competitors obtain substitutes?
• Does access to the data materially improve the product?

4. Competition-Law Theories of Harm

A. Abuse of dominance

A dominant undertaking may potentially abuse its position by:

• refusing access to indispensable data;
• imposing discriminatory data-access conditions;
• tying data access to another service;
• using customer data to disadvantage competitors;
• restricting data portability;
• imposing exclusivity;
• engaging in self-preferencing based on privileged data access.

B. Data-driven exclusion

A dominant platform may use data obtained from business users to compete against those same businesses.

For example:

A marketplace obtains detailed sales data from independent sellers and subsequently uses that information to launch competing private-label products.

The competitive concern is not merely possession of data. It is the strategic use of privileged data to weaken competitors.

C. Leveraging

A company possessing significant data in Market A may use that data advantage to enter Market B.

Example:

dominant search data → advertising advantage → adjacent digital market

The authority may examine whether the conduct transfers market power from one market into another.

D. Data-based predation

A company might offer a service at apparently low monetary prices because it obtains valuable user data.

Competition analysis must therefore recognize that:

Price is not necessarily the only competitive parameter.

Quality, privacy, data collection, innovation, and consumer choice can also constitute competitive dimensions.

5. Data Concentration and Merger Control

Data concentration has become particularly significant in merger analysis.

Traditional merger analysis asks:

Will the transaction increase market concentration and reduce competition?

Data-driven merger analysis additionally asks:

Will the transaction combine datasets in a way that creates an important competitive advantage unavailable to rivals?

Authorities may investigate:

1. uniqueness of the datasets;
2. substitutability;
3. size and quality of the data;
4. ability to combine datasets;
5. privacy effects;
6. barriers to entry;
7. network effects;
8. interoperability;
9. access to alternative datasets;
10. potential innovation effects.

6. Six Major Case Laws

1. Facebook/WhatsApp — European Commission

Case: Facebook/WhatsApp merger decision, European Commission, 2014.

The Commission examined the competitive implications of Facebook's acquisition of WhatsApp.

A significant issue concerned the potential combination of WhatsApp user data with Facebook's existing data resources.

The Commission considered whether the transaction would give Facebook a significant competitive advantage in online advertising through additional user information.

Principle

The case demonstrates that data concentration can form part of merger analysis even when the transaction involves apparently different digital services.

The availability and usefulness of alternative data sources are important in determining whether additional data materially strengthens market power.

2. Microsoft/LinkedIn — European Commission

Case: Microsoft/LinkedIn, European Commission, 2016.

Microsoft's acquisition of LinkedIn raised concerns concerning the combination of:

• Microsoft's software ecosystem;
• LinkedIn's professional-network data;
• customer information;
• professional profiles; and
• business-user relationships.

The Commission examined whether LinkedIn's data could provide Microsoft with a significant competitive advantage in various markets.

Principle

The case illustrates the importance of examining whether a transaction involving a data-rich platform could create data-related competitive advantages.

It also demonstrates that authorities may consider remedies or behavioural commitments where a transaction creates potential foreclosure concerns.

3. Google/DoubleClick — European Commission

Case: Google/DoubleClick, European Commission, 2008.

The transaction combined Google's significant online presence with DoubleClick's advertising technology and data resources.

The Commission investigated whether the transaction could strengthen Google's position in online advertising.

Principle

The case is important for understanding the relationship between:

data + advertising technology + network effects + market power.

It demonstrates that control over advertising-related information can become strategically important when combined with complementary technological infrastructure.

4. Bundeskartellamt — Facebook Data Combination Decision

Case: Bundeskartellamt v Facebook/Meta, German Federal Cartel Office, 2019.

The German competition authority addressed Facebook's practice of combining user data collected from Facebook with information obtained from other services and websites.

The case was significant because the authority considered the relationship between:

• dominance;
• data collection;
• terms of service;
• privacy-related conditions; and
• exploitative abuse.

Principle

The decision established an important competition-law concept:

The collection and combination of personal data may constitute a competition issue when imposed by a dominant undertaking under exploitative conditions.

It therefore expanded the practical relationship between data protection and competition law.

5. Google Shopping — European Commission

Case: Google Search (Shopping), European Commission, 2017; General Court judgment, 2021.

The case primarily concerned Google's treatment of comparison-shopping services in search results rather than data concentration alone.

However, it is highly relevant to strategic data concentration because Google's search ecosystem generated enormous quantities of:

• search information;
• user interaction data;
• behavioural signals; and
• commercial information.

Google's control over a major gateway to online users enabled it to influence competition in an adjacent market.

Principle

The case illustrates the broader concept of ecosystem-based leverage.

Where an undertaking controls a strategically important digital infrastructure and possesses substantial information about users and market behaviour, competition concerns may extend beyond the original market.

6. Google Android — European Commission

Case: Google Android, European Commission, 2018; General Court, 2022.

The case concerned Google's conduct involving Android devices and related services, including contractual arrangements surrounding Google's search and application ecosystem.

Although not exclusively a data case, it demonstrates how:

operating system → search → applications → users → data

can produce reinforcing ecosystem effects.

Principle

Competition authorities may consider how control over one technological layer facilitates expansion or reinforcement of market power in connected markets.

Data can function as an important component of this ecosystem feedback loop.

7. Additional Important Case: Google/Alphabet AdTech Investigations

Digital advertising investigations involving Google have increasingly focused on the relationship between:

• user data;
• advertising technology;
• exchanges;
• publisher information;
• advertiser information; and
• vertically integrated infrastructure.

The competition concern is that an undertaking controlling several layers of the advertising stack may obtain information unavailable to competitors.

Principle

Vertical control + privileged data access + information asymmetry can potentially create foreclosure advantages.

8. Data as an Essential Facility

One of the most difficult questions is whether a particular dataset can constitute an essential facility.

The argument becomes stronger where:

1. the dataset is indispensable;
2. competitors cannot reasonably reproduce it;
3. access is technically feasible;
4. refusal eliminates effective competition; and
5. there is no legitimate justification for refusal.

However, courts and authorities generally apply essential-facility principles cautiously.

Not every valuable dataset is an essential facility.

9. Data Portability and Competition

Data portability can reduce switching costs.

Consider:

Platform A → user's historical data → Platform B

If users cannot transfer their:

• transaction histories;
• contacts;
• reviews;
• playlists;
• preferences; or
• professional information,

switching may become more difficult.

Consequently, competition authorities may consider whether restrictions on portability create artificial switching barriers.

10. Data Interoperability

Interoperability can be particularly important in digital ecosystems.

A dominant platform might restrict competitors' access to:

• APIs;
• authentication systems;
• messaging protocols;
• payment information;
• technical interfaces; or
• user-generated data.

Competition concerns increase where the restriction prevents competitors from developing interoperable services.

11. Data Advantage and Artificial Intelligence

Strategic data concentration has become particularly significant with AI.

A large platform may possess:

User data + computing infrastructure + algorithms + distribution + capital

This combination can create cumulative competitive advantages.

For AI markets, authorities may examine:

• access to training datasets;
• proprietary datasets;
• data quality;
• real-time data;
• feedback data;
• synthetic-data generation;
• data-sharing arrangements;
• exclusive licensing;
• model-training access; and
• interoperability.

The relevant competitive asset may therefore be not merely "data," but the entire data-to-model feedback loop.

12. Data Concentration and Algorithmic Competition

A dominant firm controlling a large dataset may develop superior algorithms.

For example:

More transaction data → better demand prediction → better recommendations → more transactions → more data.

This can produce a self-reinforcing data advantage.

Competition authorities must distinguish between:

• legitimate innovation resulting from better data;
• economies of scale;
• superior product quality; and
• exclusionary strategies designed to prevent rivals from obtaining competing datasets.

13. Data Sharing and Cartel Risks

Data concentration does not always concern unilateral conduct.

Competitors sharing commercially sensitive data may create risks of:

• price coordination;
• output coordination;
• customer allocation;
• market allocation;
• algorithmic coordination.

For example, exchanging detailed real-time pricing data between competitors may facilitate coordinated behaviour.

Therefore:

Data sharing can promote competition when appropriately structured, but it can also facilitate collusion.

14. Data Pooling and Joint Ventures

Data pools may have procompetitive benefits.

They can facilitate:

• fraud detection;
• cybersecurity;
• scientific research;
• interoperability;
• industry standards;
• financial risk assessment.

However, competition concerns arise where participation becomes:

• exclusive;
• discriminatory;
• compulsory without justification;
• unavailable to competing firms; or
• structured to coordinate prices or output.

15. Privacy as a Parameter of Competition

Privacy can itself become a competitive dimension.

Consumers may prefer:

Service A: extensive data collection

over

Service B: limited data collection.

If a dominant company worsens privacy conditions because users have no realistic alternatives, competition authorities may examine whether this represents a form of non-price exploitation.

The Facebook German case is particularly important in this context.

16. Relevant Market Definition

Data concentration makes market definition more complicated.

Traditional markets can be assessed through:

• price;
• output;
• substitution;
• geographic boundaries.

Digital services may instead involve zero monetary prices.

Consequently, authorities may examine:

• quality;
• privacy;
• data practices;
• switching costs;
• user attention;
• interoperability;
• innovation; and
• multi-homing.

17. Barriers to Entry

Strategic data concentration may create entry barriers where a new entrant requires large quantities of historical data.

For example:

Established platform

10 years of consumer behaviour data
↓
Better algorithm
↓
Better recommendations
↓
More users
↓
More data

New entrant

Few users
↓
Limited data
↓
Lower-quality recommendations
↓
Difficulty attracting users

This is sometimes described as a data network effect.

18. Competitive Remedies

Authorities may consider several remedies where data concentration creates anticompetitive effects.

Structural remedies

• divestiture;
• separation of business units;
• restrictions on data combination.

Behavioural remedies

• data-access obligations;
• interoperability;
• API access;
• data portability;
• non-discrimination;
• firewalls;
• restrictions on combining datasets.

Merger-specific remedies

Authorities may require commitments concerning:

• continued access to datasets;
• restrictions on exclusive use;
• preservation of interoperability;
• separation of databases;
• non-discriminatory access.

19. Key Legal Questions for Analysis

A competition-law assessment of strategic data asset concentration should ask:

20. Difference Between Data Concentration and Anticompetitive Data Concentration

This distinction is fundamental.

Data concentration alone

Company possesses a very large dataset.

Not necessarily unlawful.

Strategic data concentration

Company possesses a dataset that materially strengthens its competitive position.

Requires competition analysis.

Anticompetitive data concentration

Company obtains or exploits data in a manner that substantially restricts competition, forecloses rivals, exploits consumers, or reinforces dominance without sufficient legitimate justification.

Potential competition-law violation.

21. Six Core Case-Law Principles at a Glance

22. Conclusion

Strategic data asset concentration is becoming an important dimension of modern competition law. Data may function simultaneously as an input, competitive advantage, entry barrier, network-effect accelerator and source of market power.

The central legal question is not:

"How much data does a company possess?"

but rather:

"How does control over that data affect competitive conditions?"

Competition authorities therefore increasingly need to examine the interaction between data accumulation, market power, network effects, interoperability, switching costs, privacy, innovation, algorithmic advantages and ecosystem control.