Margin Squeeze in Digital Ecosystems With Multi-Layer Pricing

Introduction

A margin squeeze occurs when a vertically integrated dominant undertaking controls an essential or strategically important upstream input and sets the upstream price, downstream price, or both in a manner that leaves an equally efficient downstream competitor with an insufficient margin to compete.

In traditional markets, the analysis is relatively straightforward:

Upstream price → downstream price → competitor’s margin.

Digital ecosystems make the problem considerably more complicated because the dominant undertaking may operate several interconnected layers simultaneously, for example:

Cloud infrastructure → AI compute → foundation model → API → application platform → app distribution → payments → advertising/data monetisation.

Pricing at each layer may be monetary, zero-price, commission-based, usage-based, subscription-based, advertising-funded, or based upon data and access conditions. Consequently, the relevant margin may not be visible from any single price.

A useful conceptual model is:

Effective Downstream Margin=PD−PU−CD−FD−RD\text{Effective Downstream Margin} = P_D - P_U - C_D - F_D - R_D

where:

• PDP_D = effective downstream revenue;
• PUP_U = upstream access/input cost;
• CDC_D = downstream incremental costs;
• FDF_D = platform fees, commissions and mandatory complementary services;
• RDR_D = economically relevant regulatory, interoperability and switching costs.

In a multi-layer ecosystem, however, the calculation must potentially be performed across several layers:

M=PLn−∑i=1n−1PLi−CM = P_{L_n} - \sum_{i=1}^{n-1} P_{L_i} - C

The central competition-law question is therefore:

Does the vertically integrated ecosystem leave an economically viable margin for an equally efficient rival once all compulsory upstream and ecosystem-layer costs are taken into account?

1. What Is a Margin Squeeze?

A margin squeeze generally requires three elements:

1. Upstream dominance;
2. A sufficiently important upstream input controlled by the dominant firm; and
3. A spread between upstream and downstream prices that is insufficient to permit effective downstream competition.

The classic structure is:

Dominant firm

→ sells wholesale input to rivals

→ competes with those rivals downstream

→ sets the wholesale price sufficiently high, the downstream price sufficiently low, or both.

The dominant firm therefore controls both sides of the competitive interface.

2. Why Digital Ecosystems Create a Different Problem

Digital ecosystems frequently contain multiple vertically connected markets.

For example:

A competitor may technically have access to each layer while still being commercially unable to compete.

For example:

Cloud price = €10
Model/API price = €20
Platform commission = €30
Payment fee = €10
Effective downstream revenue = €60

A superficial analysis might conclude that every individual price is reasonable.

But if the competitor's total cost is €70, the ecosystem creates a negative effective margin.

This produces what can be described as multi-layer margin squeeze.

3. Single-Layer Versus Multi-Layer Margin Squeeze

Traditional squeeze

W→DW \rightarrow D

where:

• WW = wholesale price;
• DD = downstream price.

Multi-layer squeeze

W1+W2+W3+F+C→PDW_1 + W_2 + W_3 + F + C \rightarrow P_D

The dominant firm may control:

• infrastructure;
• APIs;
• data;
• distribution;
• payment;
• advertising;
• identity;
• ranking;
• interoperability.

The competitive harm can therefore emerge from the combined pricing architecture, rather than one obviously abusive price.

4. Multi-Layer Pricing Mechanisms

A. Wholesale Input Pricing

The dominant platform may charge competitors for:

• API calls;
• cloud compute;
• storage;
• authentication;
• data access;
• infrastructure;
• payment processing.

A high upstream price can compress downstream margins.

B. Platform Commission

The dominant platform may simultaneously charge a commission on downstream transactions.

For example:

Competitor revenue−API cost−platform commission\text{Competitor revenue} - \text{API cost} - \text{platform commission}

may leave insufficient remuneration.

This is particularly important where the same undertaking controls both the infrastructure and distribution layers.

C. Bundled Pricing

The dominant undertaking may bundle several services.

Suppose:

• cloud access;
• AI inference;
• identity;
• analytics;
• payment processing

are offered as one package.

The competitor cannot easily determine which component creates the squeeze.

This raises a significant cost attribution problem.

5. Zero-Price Markets

Digital markets frequently contain zero monetary prices.

A service may be offered to consumers for free while the undertaking monetises:

• advertising;
• data;
• subscriptions;
• commissions;
• cross-selling;
• ecosystem lock-in.

Therefore:

Pconsumer=0P_{consumer}=0

does not necessarily mean:

Economic Price=0Economic\ Price=0

The relevant economic value may be captured elsewhere in the ecosystem.

6. Data as an Implicit Price

Data can function as an economic input.

A digital service may receive:

• behavioural data;
• transaction data;
• search data;
• location information;
• interaction data;
• model-training feedback.

The competitor may therefore incur an implicit cost even where monetary access prices are low.

A broader margin calculation could become:

M=R−Pcash−Pdata−Paccess−Pdistribution−CM = R - P_{cash} - P_{data} - P_{access} - P_{distribution} - C

The difficulty is determining the appropriate economic valuation of PdataP_{data}.

7. Cross-Layer Subsidisation

A dominant ecosystem may deliberately price one layer below cost while recovering profits elsewhere.

For example:

Cloud Price<CostCloud\ Price < Cost

but

Advertising Revenue≫CostAdvertising\ Revenue \gg Cost

This creates a potential competitive problem if the low downstream price is funded by monopoly profits generated at another layer.

The analysis therefore cannot necessarily stop at the individual product's accounting margin.

8. The Equally Efficient Competitor Test

The classic approach asks whether an equally efficient competitor could profitably compete.

The basic test is:

Downstream Price−Upstream Price≥Dominant Firm′s Relevant Downstream CostDownstream\ Price - Upstream\ Price \geq Dominant\ Firm's\ Relevant\ Downstream\ Cost

If the inequality fails, a margin squeeze may exist.

In digital ecosystems, however, several modifications may become necessary.

Relevant costs may include:

• compute;
• API calls;
• licensing;
• cloud infrastructure;
• payment fees;
• distribution commissions;
• interoperability costs;
• data acquisition;
• cybersecurity;
• compliance;
• customer acquisition;
• switching costs.

9. Long-Run Versus Short-Run Costs

Digital markets frequently have:

• very high fixed costs;
• low marginal costs;
• network effects;
• substantial R&D;
• significant scale economies.

Consequently, using only short-run marginal cost can produce misleading results.

The assessment may require:

• incremental cost;
• average avoidable cost;
• long-run average incremental cost;
• downstream operating cost;
• appropriately allocated common costs.

The correct methodology depends upon the particular market and theory of harm.

10. The Key European Case: Deutsche Telekom

Deutsche Telekom AG v Commission

This is the foundational European margin-squeeze authority.

The Court of Justice confirmed that a margin squeeze can constitute an independent form of abuse where the spread between wholesale and retail prices is insufficient for an equally efficient competitor to compete downstream.

Importantly, the undertaking need not necessarily have imposed a separate predatory downstream price.

Digital relevance

The principle translates readily to:

platform input → platform retail/service

For example:

API access price → competing AI application price

or:

app-store distribution fee → competing application revenue.

11. TeliaSonera

TeliaSonera Sverige AB v Konkurrensverket

This case significantly clarified the relationship between margin squeeze and other forms of exclusionary abuse.

The Court explained that a margin squeeze can constitute an abuse even where the upstream product is not necessarily supplied to competitors under a conventional regulatory access obligation.

The judgment is especially important for digital ecosystems because it demonstrates that the analysis cannot automatically be reduced to an essential-facilities doctrine.

Digital application

A dominant platform controlling an API, data layer or infrastructure service cannot necessarily avoid scrutiny merely by arguing:

“There is no legal obligation requiring us to supply competitors.”

12. Telefónica

Telefónica and Telefónica de España v Commission

The Telefónica litigation concerned broadband markets and margin squeeze.

The case is particularly relevant to multi-layer digital infrastructure because broadband networks themselves exhibit vertically integrated characteristics:

network infrastructure → wholesale access → retail broadband.

The EU courts accepted the Commission's assessment of the relationship between wholesale and retail prices and the competitive implications of the resulting margin.

Digital relevance

It provides a useful analytical analogy for:

cloud infrastructure → platform access → downstream digital service.

13. Slovak Telekom

Slovak Telekom v Commission

Slovak Telekom involved telecommunications infrastructure and access arrangements.

The case is significant because the Court considered exclusionary conduct associated with access to a vertically integrated network and the relationship between wholesale access and downstream competition.

Digital significance

The principles can inform cases involving:

• cloud infrastructure;
• identity infrastructure;
• app distribution;
• digital payment rails;
• interoperability services;
• API access.

The central question remains whether the conditions imposed at the upstream layer effectively impair downstream competition.

14. Wanadoo / France Télécom

France Télécom SA v Commission

The case is principally associated with predatory pricing rather than pure margin squeeze, but it is useful in understanding the relationship between pricing strategies and exclusionary effects in digital-style network markets.

The case illustrates that pricing below an appropriate cost benchmark can become problematic where the strategy is capable of excluding competitors and strengthening dominance.

Digital relevance

A digital ecosystem may combine:

• upstream monopoly profits;
• subsidised downstream pricing;
• network effects;
• rapid user acquisition.

Therefore, a margin-squeeze analysis may sometimes need to be considered alongside predatory-pricing analysis.

15. Google Shopping

Google and Alphabet v Commission

The Google Shopping litigation concerns preferential treatment of Google's own comparison-shopping service rather than a classic margin squeeze.

Nevertheless, it is highly relevant to digital ecosystems because it demonstrates how a dominant platform can use control over an upstream or intermediary layer to advantage its downstream activity.

The relevant architecture can be conceptualised as:

Search infrastructure→ranking/distribution→comparison serviceSearch\ infrastructure \rightarrow ranking/distribution \rightarrow comparison\ service

Digital margin-squeeze significance

The case demonstrates why digital competition analysis cannot always isolate price from:

• ranking;
• visibility;
• access;
• traffic;
• data;
• self-preferencing.

A platform may economically compress a rival's margin without doing so exclusively through a posted wholesale price.

16. Android

Google Android

The Android litigation is relevant to ecosystem economics because Google operated multiple interconnected layers:

• operating system;
• app distribution;
• search;
• advertising;
• mobile services.

The case illustrates the importance of analysing ecosystem-wide leverage.

A rival may theoretically compete at one layer but remain dependent on access to another layer.

Margin-squeeze implication

The effective competitive constraint can therefore be assessed through:

Input access+distribution+default position+monetisationInput\ access + distribution + default\ position + monetisation

rather than a single price.

17. Microsoft

Microsoft v Commission

The Microsoft litigation provides another important ecosystem precedent concerning interoperability, tying and the strategic use of control over a platform layer.

Although it is not a classic margin-squeeze case, it is highly relevant to modern digital ecosystem analysis.

A dominant undertaking can leverage control over one layer to influence competition at another.

Modern analogy

Operating system → API → application → cloud service

or:

AI infrastructure → model → application → distribution.

18. Important Distinction: Margin Squeeze Is Not Automatically Predatory Pricing

These concepts must be distinguished.

Predatory pricing

The focus is generally:

Price<Appropriate CostPrice < Appropriate\ Cost

Margin squeeze

The focus is:

Retail/Downstream Price−Wholesale/Upstream Price<Relevant Downstream CostRetail/Downstream\ Price - Wholesale/Upstream\ Price < Relevant\ Downstream\ Cost

Thus, a dominant firm may potentially engage in a margin squeeze even though its downstream price is above its own downstream cost.

The problem lies in the relationship between the upstream and downstream prices.

19. Multi-Layer Squeeze: A More Sophisticated Model

Consider an AI ecosystem.

The competitor's effective margin is:

€60−€55−€10=−€5€60-€55-€10=-€5

The rival loses €5 per transaction.

Yet analysing only the API layer could produce:

€60−€20−€10=€30€60-€20-€10=€30

which appears highly profitable.

This illustrates the central problem of multi-layer ecosystems: the squeeze may only become visible when compulsory ecosystem charges are aggregated.

20. Sequential Versus Cumulative Squeeze

Sequential squeeze

The dominant firm squeezes competitors at one particular interface.

Example:

Cloud→AI APICloud \rightarrow AI\ API

Cumulative squeeze

Multiple layers independently impose commercially modest burdens which collectively become exclusionary.

Example:

Cloud+API+App Store+Payment+AdvertisingCloud + API + App\ Store + Payment + Advertising

Each charge may appear defensible in isolation.

Together:

∑Costs>Downstream Revenue\sum Costs > Downstream\ Revenue

This creates a cumulative margin squeeze theory.

21. Bundling and Common-Cost Allocation

One of the most difficult issues is determining how to allocate costs when several services are bundled.

Suppose a dominant undertaking offers:

Cloud + model + analytics + security + identity

for one price.

A regulator must determine:

1. Which component is the bottleneck input?
2. Which costs are incremental?
3. Which costs are common?
4. Which costs would an equally efficient rival incur?
5. Which revenues should be attributed to each layer?

Incorrect allocation can dramatically change the outcome.

22. Internal Transfer Prices

Digital conglomerates often transact internally.

For example:

Cloud Division→AI Division→Platform DivisionCloud\ Division \rightarrow AI\ Division \rightarrow Platform\ Division

The internal transfer price may not represent a competitive market price.

Therefore, regulators may need to reconstruct:

What would the relevant upstream and downstream economics look like under competitive conditions?

This makes accounting evidence particularly important.

23. Dynamic Pricing

Digital platforms can change prices continuously through algorithms.

The relevant margin may therefore vary by:

• customer;
• geography;
• time;
• usage volume;
• demand;
• device;
• subscription tier.

A static margin test may consequently miss the actual competitive effect.

A regulator may need to analyse:

M(t)=PD(t)−PU(t)−CD(t)M(t)=P_D(t)-P_U(t)-C_D(t)

over time.

Persistent negative or strategically compressed margins may be more probative than isolated observations.

24. Personalised Pricing

Personalisation creates another complication.

A dominant platform could offer:

• lower prices to strategically important customers;
• higher input charges to rivals;
• differentiated commissions;
• personalised discounts downstream.

The resulting squeeze might not appear in average prices.

Competition authorities may therefore need customer-level margin analysis.

25. Cross-Subsidisation Across Ecosystem Layers

Suppose:

Upstream Monopoly Profit=€100mUpstream\ Monopoly\ Profit=€100m

and:

Downstream Loss=€60mDownstream\ Loss=€60m

The ecosystem remains profitable overall.

A conventional product-level analysis could incorrectly conclude that the downstream business is simply competitive.

However, the strategic question is whether upstream market power finances exclusionary downstream pricing.

This creates an important connection between:

• margin squeeze;
• predatory pricing;
• tying;
• leveraging;
• bundling;
• cross-subsidisation.

26. Data and Non-Monetary Margin Squeeze

A modern digital platform might provide API access cheaply but require extensive data surrender.

For example:

API price = €5

but competitor must provide:

• transaction data;
• customer behavioural data;
• model-training feedback;
• metadata.

The true economic cost could therefore substantially exceed €5.

A modern margin analysis may need to consider:

Effective Price=Monetary Price+Data Cost+Access Cost+Switching CostEffective\ Price = Monetary\ Price + Data\ Cost + Access\ Cost + Switching\ Cost

The valuation of non-monetary consideration remains highly fact-specific.

27. Interoperability as a Hidden Price

Suppose a competitor nominally receives API access but must:

• use proprietary protocols;
• maintain separate certification;
• pay testing fees;
• undertake repeated compliance audits;
• purchase complementary software.

The nominal upstream price might be low.

The effective access price, however, may be substantially higher.

Thus:

Effective Wholesale Cost=Nominal Price+Technical Access Costs+Compliance Costs+Interoperability CostsEffective\ Wholesale\ Cost = Nominal\ Price + Technical\ Access\ Costs + Compliance\ Costs + Interoperability\ Costs 

28. Network Effects Magnify the Squeeze

Digital ecosystems frequently exhibit network effects.

A small pricing disadvantage can therefore create a much larger competitive effect.

For example:

Higher Input Cost→Higher Retail Price→Fewer Users→Less Data→Lower Quality→Even Fewer UsersHigher\ Input\ Cost \rightarrow Higher\ Retail\ Price \rightarrow Fewer\ Users \rightarrow Less\ Data \rightarrow Lower\ Quality \rightarrow Even\ Fewer\ Users

This is a dynamic margin squeeze.

The initial price disadvantage can therefore generate a feedback loop.

29. Multi-Homing

Multi-homing affects the assessment of competitive harm.

If developers or users can easily participate in multiple ecosystems, a squeeze may have a smaller exclusionary effect.

If switching is difficult, however:

Squeeze+Lock-in+Network EffectsSqueeze + Lock\text{-}in + Network\ Effects

can substantially increase foreclosure.

Therefore, the economic assessment should consider:

• switching costs;
• portability;
• interoperability;
• multi-homing;
• contractual restrictions;
• default positions.

30. Relevant Counterfactual

A sophisticated investigation should ask:

What would the downstream competitor's economics look like absent the allegedly abusive pricing structure?

Possible counterfactuals include:

Counterfactual A — Competitive wholesale price

What if the upstream input were priced competitively?

Counterfactual B — No compulsory bundle

What if the competitor could purchase only the input it actually needs?

Counterfactual C — Interoperability

What if technical restrictions did not impose additional costs?

Counterfactual D — Competitive distribution

What if the competitor could distribute directly without the dominant platform's commission?

The counterfactual can reveal whether the squeeze is genuinely exclusionary.

31. Efficiency and Justification

A dominant undertaking may argue that pricing reflects:

• genuine economies of scale;
• security investments;
• infrastructure costs;
• innovation;
• quality improvements;
• risk;
• congestion;
• legitimate volume discounts.

These arguments cannot automatically be dismissed.

The competition authority should distinguish:

Legitimate EfficiencyLegitimate\ Efficiency

from

Strategic ExclusionStrategic\ Exclusion

The fact that a competitor experiences a low margin is not, by itself, sufficient to establish abuse.

32. Objective Justification

Potential objective justifications may include:

• capacity constraints;
• security requirements;
• fraud prevention;
• infrastructure recovery;
• technical compatibility;
• genuine cost differences.

However, the justification should be:

1. objectively necessary;
2. proportionate;
3. consistently applied;
4. capable of producing the claimed benefit.

33. Remedies

A multi-layer squeeze can require remedies extending beyond price regulation.

Possible remedies include:

1. Price transparency

Require disclosure of relevant ecosystem-layer charges.

2. Accounting separation

Separate upstream and downstream accounts.

3. Non-discrimination

Prevent discriminatory access prices.

4. Interoperability

Require technically effective access.

5. API access

Provide equivalent access to rivals.

6. Unbundling

Permit competitors to purchase only necessary components.

7. Commission restrictions

Limit excessive or discriminatory platform commissions.

8. Structural separation

In extreme cases, separate infrastructure from downstream operations.

34. Relationship With the EU Digital Markets Act

The Digital Markets Act changes the regulatory environment because certain conduct by designated gatekeepers can be addressed through ex ante obligations rather than waiting for a conventional Article 102 TFEU investigation.

This matters because a digital ecosystem may exhibit exclusionary effects before traditional margin-squeeze litigation can be completed.

The practical relationship can therefore be understood as:

Ex Ante DMA+Ex Post Article 102Ex\ Ante\ DMA + Ex\ Post\ Article\ 102

rather than treating them as mutually exclusive systems.

35. Relationship With Article 102 TFEU

Article 102 TFEU remains central where the undertaking:

• occupies a dominant position;
• engages in exclusionary conduct;
• affects trade between Member States.

Margin squeeze is particularly relevant where the undertaking is vertically integrated and controls an upstream input needed for downstream competition.

The assessment should nevertheless remain grounded in the specific economic circumstances rather than assuming that every vertically integrated platform creates a squeeze.

36. German Competition-Law Perspective

Under German competition law, the analysis can become particularly significant for large digital ecosystems.

GWB §19 addresses abusive conduct by dominant undertakings, while §19a GWB provides a specialised framework for undertakings of paramount significance across markets.

This creates a potentially important distinction:

Traditional approach

Dominance→Abuse→EffectDominance \rightarrow Abuse \rightarrow Effect

Digital ecosystem approach

Cross-market Power→Dependency→Ecosystem Leverage→Potential ForeclosureCross\text{-}market\ Power \rightarrow Dependency \rightarrow Ecosystem\ Leverage \rightarrow Potential\ Foreclosure

The German framework is therefore particularly relevant where an undertaking controls several interconnected digital layers.

37. Six Core Case Laws — Exam Table

38. Additional Authorities Worth Knowing

For a stronger examination answer, the following cases can also be connected to the doctrine:

• AstraZeneca v Commission — exclusionary conduct and regulatory mechanisms;
• Intel v Commission — effects-based assessment and foreclosure;
• Post Danmark I — pricing and exclusionary effects;
• Servizio Elettrico Nazionale — abuse and competition on the merits;
• Bronner — essential-facilities/access framework;
• Google Android — ecosystem leverage, tying and defaults;
• Microsoft — interoperability and platform power.

These are not all pure margin-squeeze cases; they are useful because modern digital margin-squeeze analysis intersects with several neighbouring theories of harm.

39. Regulatory Evidence in a Digital Margin-Squeeze Investigation

A competition authority should obtain:

Pricing data

• wholesale prices;
• retail prices;
• commissions;
• discounts;
• rebates.

Technical data

• API usage;
• interoperability requirements;
• latency;
• access restrictions.

Accounting data

• incremental costs;
• common costs;
• transfer prices;
• cross-subsidisation.

Behavioural data

• customer switching;
• multi-homing;
• churn;
• foreclosure rates.

Algorithmic data

• personalised pricing;
• dynamic discounts;
• automated commissions;
• customer segmentation.

This creates an increasingly important role for algorithmic auditing and granular transaction-level data.

40. The Central Legal Test

A useful exam formulation is:

A multi-layer digital margin squeeze arises where a dominant vertically integrated digital undertaking controls one or more upstream ecosystem inputs and structures the combined upstream, intermediary and downstream prices or conditions so that an equally efficient rival cannot obtain an economically viable margin, taking account of the relevant monetary and non-monetary costs of participating in the ecosystem.

The authority should then examine:

Effective Downstream Revenue−Total Compulsory Ecosystem Costs−Efficient Downstream Costs\boxed{ Effective\ Downstream\ Revenue - Total\ Compulsory\ Ecosystem\ Costs - Efficient\ Downstream\ Costs }

If the resulting margin is persistently and strategically insufficient, the conduct may warrant investigation under the applicable abuse-of-dominance framework.

Conclusion

Margin squeeze in digital ecosystems is no longer adequately understood as merely a difference between a wholesale telecommunications price and a retail price.

Modern ecosystems can contain:

cloud → compute → AI model → API → application → distribution → payment → advertising/data.

A dominant undertaking can potentially influence competition through the combined architecture of these layers.

The most important analytical developments are therefore:

1. Cumulative rather than single-layer pricing analysis;
2. Recognition of zero-price and non-monetary transactions;
3. Consideration of data and interoperability costs;
4. Customer- and transaction-level margin analysis;
5. Dynamic rather than purely static assessment;
6. Counterfactual modelling;
7. Integration of pricing with network effects and switching costs; and
8. Coordination between traditional Article 102/GWB enforcement and digital ex-ante regulation.