Selective distribution systems and UK competition law: the Court of Appeal’s final word in Up & Running v Deckers

As explained in more detail in our November 2024 briefing, the case involved a dispute between Deckers, a manufacturer of HOKA-branded running shoes, and its retail partner. Under the terms of Deckers’ SDS, retailers had to obtain Deckers' approval to supply HOKA-branded products (both current season and out of season, clearance products), whether in physical stores or online.

Up & Running was already approved to sell HOKA products (both in bricks and mortar stores and on its existing associated website) but wanted to start a new anonymised website to sell excess stock at discounted prices, due to stock volumes accumulated during the COVID-19 pandemic. Deckers declined to approve this second website.

An SDS is a distribution model where suppliers sell only to authorised distributors meeting minimum, specific criteria (e.g., quality standards or providing after-sales services). In return, the distributors commit to only supply end users or other approved distributors. The rationale for using a SDS is typically to restrict distribution to achieve a legitimate purpose, such as maintaining the supplier's brand image and achieving service consistency, while preventing free-riding by distributors who are not prepared to invest in a high-quality retail environment.

It is established EU case law that certain types of SDS fall entirely outside the prohibition on anti-competitive agreements (Ch/apter I of the Competition Act 1998 / Article 101 of the Treaty on the Functioning of the EU (TFEU)), if they meet the criteria set out in Metro v Commission (Case C-26/76) (see our November 2024 briefing for further detail on what those entail).

Otherwise, the SDS must comply with competition law – i.e., not have the "object or effect" of restricting competition, unless it can be exempted under either the Vertical Agreements Block Exemption (VBE) or the individual conditions in s.9 of the Competition Act 1998 / Article 101(3) of the TFEU. The VBE is permissive of selective distribution as long as (i) the relevant market share thresholds are met; and (ii) the system does not contain “hardcore” restrictions.

The present case was an appeal against a 25 October 2024 ruling of the CAT (CAT Judgment). The CAT was asked to decide whether Deckers' refusal to approve the second website amounted to an object infringement of competition law. As explained in more detail in our November 2024 briefing, the CAT found that: 

• the SDS did not meet the Metro criteria as Deckers’ approval criteria (e.g., relating to consumer experience, presentation, service and product range) were neither objective nor applied in a uniform and non-discriminatory fashion;
• the refusal to approve the second website, and the termination of the contractual relationship with the claimant distributor amounted to a "by object" restriction of competition. In particular, the CAT concluded that because the only plausible objective of Deckers' approval requirements and its refusal to authorise the website was to prevent retailers from significantly discounting stock on a clearance basis, this was a form of setting retail prices (retail price maintenance) which infringed competition law “by object”, without the need for the tribunal to carry out any wider economic assessment; and
• the agreement did not benefit from the VBE. Although the parties' market shares fell squarely within the VBE thresholds, the CAT held the VBE was not available as the arrangements amounted to a "hardcore” restriction, by preventing a buyer from selling to end users (other than from an unauthorised place of establishment) and limiting their ability to set retail prices. 

Given the significant impact of the CAT Judgment on Deckers’ approach to distribution, Deckers appealed the CAT’s findings to the CoA. Addleshaw Goddard represented Deckers in its appeal.

The key points of appeal were that the CAT had: 

• mischaracterised Deckers’ conduct as a form of “hardcore” resale price maintenance;
• incorrectly found that the relevant legal test for “by object” infringements could be limited to considering the objective or purpose of the restriction and did not require a consideration of whether there was a sufficient degree of harm to competition by reference to the content of the restriction and its legal and economic context (in addition to the objective);
• not applied the correct law to the facts, in that the restriction on the distributor (preventing sales through an additional website) was not capable of having sufficient impact on competition in the relevant product market given the relevant economic context; and
• failed to apply the VBE correctly, in that Deckers should have benefitted from the exemption due to the parties’ low market shares as the restriction was not hardcore.

Interestingly, the CMA intervened in the case on account of its precedent value and sided with Deckers in arguing that the CAT had misinterpreted the law and applied the relevant test for by object restrictions incorrectly. The CMA submitted that the CAT had “wrongly elevated one relevant factor… into the whole legal test” by only considering the objective pursued by the supplier as relevant, without considering other factors, including the economic and legal context, as part of its assessment. It added that the loss of intra-brand competition is only problematic where inter-brand competition is already limited and that therefore considering the parties’ market power as part of the overall assessment will always be relevant. The CMA also noted that its position was consistent with both the established caselaw of the Court of Justice of the European Union and the European Commission’s Guidelines on Vertical Restraints. (1)

(1) CoA Judgment, paragraph 48.

The CoA upheld Deckers’ challenge on all counts and established some key principles that provide more latitude to a supplier seeking to operate a form of SDS.  In particular, the CoA:

(a) explicitly rejected the view that an SDS falling outside of the Metro safe harbour amounted to or was very likely to amount to a restriction by object, arguably putting an end to a lack of clarity on this point arising from prior case law; (2)

(b) rejected the inference that if a clause was “hardcore” as defined by the VBE, this alone is sufficient to classify it as a restriction by object (3) or that a clause conferring unfettered discretion on the supplier was a restriction by object because of its potential to be used for an unlawful, restrictive purpose; (4)

(c) helpfully distilled the relevant competition law cases (5), making clear that an assessment of a potential by object restriction requires the application of a cumulative four-part test, which must always include some assessment of the economic context (i.e., nature of goods or services affected, market conditions and structure, including market shares), alongside the content of the restriction (i.e., its actual scope), its objective/purpose (with the analysis being objective in nature, albeit evidence of the supplier’s subjective intention can also be taken into account) and the legal context within which it takes place (e.g., prior case law, whether agreement is horizontal or vertical or any legal or regulatory barriers to entry);

(d) emphasised that the law on restrictions by object needs to be interpreted strictly and restrictively, as to do otherwise can hinder effective competition. In the words of the CoA, “competition law should not interfere unless there is a need to interfere”; (6) and

(e) helpfully reinforced that a restriction would only amount to a hardcore restriction under the VBE where it had a “real and practical” impact on a retailer’s ability to discount or on customers’ ability to freely access goods passively or actively. (7)

(2) CoA Judgment paragraph 108(iii).

(3) CoA Judgment paragraph 108(iv).

(4) CoA Judgment, paragraph 108(v).

(5) CoA Judgment, paragraph 105.

(6) CoA Judgment, paragraph 115.

(7) CoA Judgment, paragraph 156.

A selective distribution that meets the Metro criteria permits SDS operators to limit sales to a closed system of third-party retailers meting the SDS operator’s chosen criteria. But achieving consistent adherence to the requirements of a Metro compliant system is often a practical challenge and comes at the cost of commercial flexibility for many suppliers. 

However, the CoA has now unequivocally confirmed that failing to meet the Metro criteria does not make an SDS invariably unlawful and that such agreements can still very much be capable of benefitting from the VBE safe harbour. Overall, the CoA Judgment confirms that there is a materially wider variety of circumstances where SDS operators could place restrictions on their distributors without having a fully rigorous SDS in place and places them in a more legally secure position than would have been the case had the CAT Judgment stood. 

For product and geographic markets where the supplier’s share is below 30% and there is no relevant retailer with a share of over 30%, SDS operators can much more confidently rely on the VBE. It is only necessary to avoid hardcore restrictions. The CoA Judgment confirms that what amounts to a hardcore restriction is not infinitely elastic. As long as retailers can, “in a real and practical sense” discount freely (or customers access the goods passively or actively) there will be no hardcore restrictions. Under this reasoning, a restriction which in practice is very narrow in its scope and application (e.g., because it applies only to a niche channel or particular individual items within a wider product market) stands a good chance of not being considered “hardcore” even if one of the objectives of that conduct is to restrain pricing.

Where SDS operators do not have the benefit of the VBE (either because retailer shares, or their own, are high, but also where there is a hardcore restriction under their terms and conditions), the CoA Judgment is explicit that an assessment of the relevant restrictions has to be conducted. This must always include some consideration of the economic context, even to identify an anti-competitive “object”. The higher the shares, the more likely it is that the SDS operator’s conduct will have a material economic impact and that vigorous attempts to influence retail prices, or availability of products will be considered by object infringements of competition law by a regulator or court. 

An SDS operator with higher market shares can still limit its risk exposure, without having to operate a fully Metro compliant system, including by appropriate consideration of the impact of its actions within the economic context in which it operates.

Experienced SDS operators may recall the 2017 CMA decision against golf club manufacturer PING and subsequent appeals to the CAT and CoA. In that case, the SDS operator had applied a policy which required all its distributors to sell PING golf clubs face-to-face after a custom fitting assessment and banned all online sales. The CMA had found that this practice amounted to a breach of competition law by object and fined PING £1.45m, requiring it to end its online sales ban. Both the CAT and CoA upheld the CMA’s decision on appeal.

The CoA Judgment did consider and apply the PING case, suggesting the difference in outcomes came down to differences in facts and how these affected the assessment of the practices’ respective impacts on the market – in particular: 

• PING had substantial market shares nearing 30% in the UK and wider EU;
• its restrictions applied to distributors’ use of online sales channels across PING’s entire UK network; and
• online sales channels were deemed of economic importance for this particular type of products.

In contrast, Deckers’ restriction only affected a small one-off tranche of clearance stock (not all stock) and was targeted at a single retailer and only in relation to one anonymised website. Product and market shares also indicated healthy levels of inter-brand competition.