Exclusivity, tying and loyalty rebates are the dominant firm’s classic distribution weapons — and the most frequently condemned abuse category. Exclusive dealing forecloses rivals from customers; tying leverages a must-have product into a contested market; retroactive loyalty rebates replicate exclusivity through price structure. Post-Intel, all are judged on foreclosure capability against as-efficient competitors — form matters less, effects analysis and coverage arithmetic more.
Exclusive dealing, tying and loyalty rebates account for more abuse decisions than any other conduct family, because they grow directly out of ordinary commercial practice: every supplier wants committed customers, every multi-product firm wants to bundle. This guide sets out where each practice crosses the line for dominant firms, the Intel-era tests, and the design patterns that keep commercial programs defensible — completing the dominance pillar of our Competition & Antitrust hub.
Why are these practices singled out for dominant firms?
Because with dominance, locking distribution forecloses rivals from the customer base they need for scale: the ‘non-contestable’ share of demand that must be bought from the incumbent becomes leverage over the contestable remainder.
What is the difference between quantity and loyalty rebates?
Quantity (incremental) discounts reward volume as such and are broadly lawful. Loyalty rebates condition the discount on share-of-wallet or exclusivity, and retroactive versions apply the discount to all past purchases once a threshold is hit — creating a ‘suction effect’ that can make small contestable volumes impossible for rivals to price against.
What decides modern cases?
Foreclosure arithmetic: covered share of the market, the effective price an as-efficient rival must beat over the contestable portion, duration and lock-in mechanics, plus evidence of strategy. Intel made this analysis mandatory when the defendant raises it.
Why does exclusivity foreclose when a dominant firm uses it?
Because customers cannot avoid the incumbent entirely: distributors must stock the must-carry brand, manufacturers must buy the dominant input for part of their needs. That non-contestable share gives the incumbent a lever rivals lack — it can price the committed relationship as a package, effectively spending the profits of the assured volume to lock the contestable remainder. A rival competing only for the contestable slice must compensate the customer for losing the whole package’s benefits — an artificially steep hill unrelated to relative efficiency.
Foreclosure becomes anticompetitive when its scale deprives rivals of the customer access needed to reach efficient scale: the analysis therefore counts coverage (what share of the market is tied up), duration and staggering of the contracts, switching penalties, and the practical availability of distribution alternatives. Modest, short, terminable exclusivity in a market with open channels is routinely defensible even for dominant firms; extensive networks of long, auto-renewing exclusives in concentrated channels are the standing fact pattern of condemnation — in EU decisions, US Section 2 cases and the Turkish Board’s distribution docket (beer, soft drinks, ice cream freezer exclusivity and fuel-station cases built the local canon).
When does tying become abuse?
The elements, stable since Microsoft: separate products (by consumer demand), dominance in the tying product, coercion (contractual, technical or economic — the untied option unavailable or irrational), and foreclosure of rivals in the tied market without adequate justification. Technical tying — building the tied product in — draws the modern battles, since unbundling remedies fight product design itself.
The case line traces the doctrine’s evolution: Windows Media Player (unbundling ordered, famously ineffective), Android’s app-suite conditions (€4.34 billion — pre-installation as the tie), and the Teams-Office investigation resolved by commitments unbundling Teams and opening interoperability — a sign that the Commission now negotiates product-architecture fixes rather than litigating a decade. Mixed bundling (pricing the bundle below the sum of components) shades into rebate analysis: lawful when the bundle price covers incremental costs and reflects genuine economies, dangerous when discount attribution makes rival single-product entry unpriceable. Multi-product firms with one dominant component should run the discount-attribution test — regulators and claimants certainly will.
Intel’s rebates to OEMs — conditioned on 80-100% Intel share of their CPU purchases — were fined €1.06 billion in 2009 under a form-based approach. Fifteen years of litigation later, the exclusivity-rebate findings were annulled: once Intel submitted an as-efficient-competitor analysis, the Commission was obliged to engage with whether the rebate structure could actually foreclose an equally efficient AMD over the contestable volumes — and its analysis failed judicial scrutiny. The saga’s legacy is procedural discipline: rebate cases are now won and lost on effective-price economics, coverage data and contestable-share estimates, prepared to litigation grade on both sides.
How do loyalty rebates create the ‘suction effect’?
Retroactivity is the engine. A rebate of 10% on all annual purchases once 90% share-of-wallet is reached means that, near the threshold, the last increment of rival purchases costs the customer the discount on the entire year’s volume — the effective price a rival must beat on the contestable units can fall below zero. The structure converts a headline discount into a fidelity mechanism stronger than contract exclusivity, invisible in list prices.
The compliant redesigns are known: incremental rebates (discounts on marginal volume bands only), individualised thresholds set on objective capacity grounds rather than share-of-wallet, shorter reference periods, and transparency that lets customers compare offers unit by unit. The AEC test is the audit instrument: compute the effective price over the realistically contestable share for your top twenty customers; where it dips toward or below your AAC, the scheme needs restructuring regardless of litigation appetite — because that spreadsheet, in an authority’s hands, is the case. Türkiye’s Board applies the same framework, and its distribution decisions add a local nuance: cumulative effect across a market’s parallel exclusivity networks can condemn practices each modest alone.
What about MFNs, parity clauses and English clauses?
Adjacent devices, actively policed. Most-favoured-nation / retail parity clauses — the platform demanding suppliers never price lower elsewhere — soften price competition and entrench incumbents: wide parity (covering rival platforms) has been condemned or banned across EU jurisdictions and in Türkiye’s platform decisions (the Board’s food-delivery and marketplace cases), while narrow parity (own-website only) survives in some regimes and falls in others. The 2022 EU Vertical Block Exemption removed wide retail parity from safe harbour.
English clauses — the customer may buy cheaper elsewhere only after giving the incumbent a matching right — function as exclusivity plus a rival-price surveillance system and are treated accordingly. Exclusivity payments to distribution gatekeepers (slotting exclusivity, default-position purchases) complete the family: the Google search defaults condemned in Washington are, doctrinally, loyalty payments at ecosystem scale. The cross-regime pattern for all of them: assess by coverage and effect, presume badly of retroactivity and wideness, and mind the sector-specific overlays — the DMA simply prohibits several outright for gatekeepers.
How should dominant firms design compliant commercial programs?
By replacing lock-in with earned preference. The defensible toolkit: incremental volume discounts priced above cost; service and investment-based differentiation (training, tooling, category support) tied to verifiable costs; short terms with real termination rights; capacity-based supply commitments rather than share-of-wallet conditions; and honest tender behaviour where customers run competitive processes.
Governance makes it durable: a rebate-scheme register with AEC screening, contract-clause libraries with dominance-mode variants, deal-desk escalation for exclusivity requests (customers often ask for exclusivity — the request does not immunise the grant), and periodic coverage mapping across the distribution network, including parallel networks’ cumulative effect. Layer the market-share trigger from our abuse overview on top, and the program scales with risk: what is merely commercial at 30% share is regulated conduct at 50%, and the firms that track the transition in real time are the ones that never make case law.
How do these theories play out in litigation and damages?
Foreclosure cases generate distinctive private litigation: excluded rivals claim lost profits and market-position damages (harder to quantify than overcharges, but juries and courts award them — US exclusive-dealing verdicts and EU follow-on claims after rebate decisions both show nine-figure potential), while customers claim the price effects of softened competition. Interim relief matters more than in cartel cases: an excluded rival’s business may not survive the decision timeline, so injunction applications and authority interim measures are the real battlefield — Türkiye’s Board has used them in platform exclusivity matters.
For the dominant defendant, the litigation-economics lesson mirrors Intel: the AEC and coverage analyses must be litigation-grade from the first submission, because the case will be decided on whose arithmetic survives expert cross-examination. For complainants, the file to build is customer-level: coverage maps, effective-price computations on real contract data, and evidence of scale denial — the elements authorities triage complaints by.
How should customers and distributors respond to lock-in offers?
With their own arithmetic. A retroactive rebate’s headline generosity often prices below the flexibility it removes: compute the true exit cost at your projected volumes, compare against dual-sourcing’s resilience value, and negotiate structure — incremental bands, shorter references, portability. Customers hold more power here than they use: dominant suppliers need coverage, and coverage is bought from customers who can price what they are selling.
Where lock-in is imposed rather than negotiated, the complaint route is a genuine commercial lever: authorities prioritise customer complaints over rival complaints (less self-interest suspicion), interim measures exist, and even an informal authority inquiry recalibrates a supplier’s posture. Documented refusals, comparative offers and effective-price computations make the file; our guidance on complaining about abuse and the deal-desk patterns above apply symmetrically from the buying side.
Frequently Asked Questions
Are exclusivity agreements illegal for everyone?
No — for non-dominant firms they are generally lawful (within vertical-agreement rules and the VBER’s 5-year exclusivity guidance). Dominance changes the analysis: the same clause becomes an abuse candidate judged on foreclosure effect.
Is asking customers to sign exclusivity at their request still risky?
Yes — the origin of the request does not neutralise foreclosure; authorities examine effects, not etiquette. Document the customer’s rationale and keep terms short and terminable if you proceed.
How do free products figure in tying analysis?
Zero-price tied goods (bundled software, free delivery) still tie: coercion and foreclosure are assessed on the structure, and ‘free’ funded by the dominant product is precisely the leverage pattern — Android’s free app suite anchored a €4bn decision.
What sectors does Türkiye police hardest here?
Fast-moving consumer goods distribution (beverage and ice cream exclusivity lines), fuel retail, platform parity clauses and pharmaceutical distribution — plus growing platform self-preferencing enforcement under the amended Article 6 practice.
Discover more from Kurums | Business Intelligence
Subscribe to get the latest posts sent to your email.
