Two years to £260m. Rogers v Alphabet Inc

Google settled the Play Store developer claim a month before trial. The settlement figure has been covered. For funders, the interesting part is the two years of rulings that produced it, and what they say about where a defendant now chooses to fight.

Google settled the Play Store developer claim a month before trial. The settlement figure has been covered. For funders, the interesting part is the two years of rulings that produced it, and what they say about where a defendant now chooses to fight.

Professor Barry Rodger v Alphabet Inc and others   ·   Case No. 1673/7/7/24   ·   Competition Appeal Tribunal

The claim

Professor Barry Rodger, a competition law academic at Strathclyde, issued collective proceedings against seven Alphabet and Google entities at the Competition Appeal Tribunal on 23 August 2024. His solicitors were Geradin Partners. Bench Walk Guernsey PCC Limited funded the proceedings through the GPS UK Funding Cell, whose beneficial owner is Bench Walk 23t LP, a Delaware fund.

The pleaded case alleged two abuses of dominance. First, exclusionary conduct preventing developers from distributing Android apps by routes other than the Play Store. Second, unfair pricing through commissions of up to 30% on Play Store transactions, which the Tribunal found to be excessive and unfair in itself and as a system.

The class was UK-domiciled third-party app developers with qualifying sales. At certification, the Tribunal worked from an estimate of roughly 2,200 developers, of whom between 70% and 77% had losses under £10,000. By the time of the disclosure ruling a year later, the figure in use was around 2,600. The certification judgment put aggregate damages at £374m to £859m before interest, or £425m to £1,036m with compound interest.

That range matters, because the press has framed the £260m settlement as a quarter of a £1bn claim. The £1bn is the top of the interest-inclusive range. Measured against the mid-point of the pre-interest range, the £160m going to the class is closer to a fifth.

The chronology

Six substantive decisions shaped this claim between filing and settlement. Each carried something a funder should note.

23 Aug 2024   ·   Claim filed

Rodger files, three years after Coll

The claim was issued into a market where Google was already defending two related actions in the same tribunal: Epic Games’ standalone claim, filed in 2020, and Elizabeth Coll’s consumer collective proceedings, filed in 2021 and certified in August 2022. Rodger was the late entrant, and the class period ran back six years from the filing date to 22 August 2018.

FOR FUNDERS   Entering an established docket is a strategy, not an accident. Rodger arrived after the disclosure had been fought over and after the market definition had been argued. The economics of that choice run through everything that follows.

24 Mar 2025   ·   Order; [2025] CAT 25, 30 Apr 2025

Consolidation, and the price of a trial slot

The three actions were ordered to be case-managed and tried together. Coll opposed it. Her October 2025 trial date was lost; she said Rodger had been dilatory, and she put the additional cost of adjournment at approximately £6.4m. Rodger and Epic each offered a £1.5m indemnity. The Tribunal held that £3m was a reasonable sum to redress the additional cost to a significant degree, and adjourned to a joint trial in late 2026, observing that three separate trials would have consumed twenty or more weeks of tribunal time.

FOR FUNDERS   A funder in a later claim wrote a cheque to compensate the parties in an earlier one for the delay its arrival caused. This is a real and quantifiable exposure that sits outside the usual adverse costs model, and it is rarely priced. It bought Rodger a trial date roughly four years earlier than a standalone claim would have secured.

23 May 2025   ·   CPO made; reasons [2025] CAT 45, 6 Aug 2025

Certification, on an opt-out basis

At the hearing on 6 March 2025, the Tribunal indicated it would certify subject to outstanding points. Once those were resolved, the collective proceedings order was made on 23 May 2025, and unanimous written reasons followed on 6 August. Google’s objections are set out in the table below, and four of the six concerned funding rather than the merits.

On opt-out, the Tribunal applied the Rule 79(3) factors. It found the claim strong, noting that Google had not applied to strike out or for summary judgment, and found an opt-in impracticable given the small individual sums and the developers’ continuing commercial relationships with Google.

FOR FUNDERS   Nine months from filing to a certified opt-out class is fast for this regime. The absence of any strike-out or summary judgment application was treated by the Tribunal as evidence of the claim’s strength, which is a useful reminder that a defendant’s procedural silence is read against it.

25 Sep 2025   ·   Ruling; [2025] CAT 58

Disclosure, and who pays for pass-on

By this point around 3.5 million documents had been disclosed, with a further 70,000 agreed. Google put its own UK disclosure costs at over £7m in solicitors’ fees and £1.1m in disbursements. The Tribunal took a selective approach to Rodger’s outstanding requests: it granted OEM payment data and material from the CMA’s mobile browsers work, conditionally granted material on the suspension of Indian developers’ apps as evidence of Google’s general enforcement pattern, and largely refused a reconstruction of Play Store profitability as impracticable at that stage, with an adverse inference warning attached.

The more consequential ruling went the other way. Google sought transaction, pricing and mitigation data from a sample of class members to run its pass-on defence. The Tribunal directed that it must apply formally under Rules 63 and 89(1)(c) against identified represented persons, rather than route the request through the class representative, so that Google bore the associated cost. It cited the guidance in Gutmann and Le Patourel against expecting individualised class member evidence in opt-out proceedings.

FOR FUNDERS   Pass-on disclosure is one of the largest uncontrolled cost risks in a business-to-business opt-out claim, because the defendant can otherwise drive spend through the class representative’s budget. The Tribunal put that cost back on the defendant. That is a meaningful protection for a funded budget, and worth citing in underwriting.

10 Jun 2026   ·   Judgment; [2026] CAT 49

Google tries to break the class, and the funding with it

Fifteen months after certification, Google applied to vary the CPO so that the 25 largest developers by claim size would have to opt in, relying on its own transaction data to identify them. The Tribunal refused. It held that claim sizes sat on a sliding scale rather than showing a qualitative difference, that the vast majority of class members held claims worth no more than a few hundred pounds, that removing the largest claims would jeopardise the viability of the funding and risk the proceedings collapsing, and that a change at that stage would derail the trial.

FOR FUNDERS   The most directly useful authority to come out of this claim. The Tribunal treated the continued viability of the funding as a reason not to restructure a certified class. A defendant strategy of carving out the claims that carry the economics has now been tested and refused.

27 Aug 2026   ·   Settlement announced

Settlement, one month before trial

Google agreed to pay £260m, of which £160m goes to developers with qualifying sales between 22 August 2018 and 31 July 2026, and £100m to the cost of bringing and funding the claim. There is no admission of liability. The agreement records Google’s belief that it has strong defences. The trial had been listed to begin on 28 September 2026 and to run for up to eleven weeks.

Under the litigation funding agreement published at certification, the funder’s entitlement was 100% of capital outlay for the first eighteen months, 200% thereafter, and 400% from the first day of a liability trial. The settlement therefore landed while the 2x tier was still running, about a month before the step-up.

FOR FUNDERS   At certification, Google argued the uplift gave the funder a perverse incentive to avoid settlement. The Tribunal rejected that, reasoning that the class representative controlled settlement and that Google had its own incentive to settle before the return stepped up. That is what happened. The ratchet operated as a settlement-forcing device, and the class captured the lower multiple.

15 Sep 2026   ·   Approval hearing

The part that is not yet decided

The Tribunal must approve the settlement as just and reasonable. Represented persons had until 4pm on 10 September 2026 to file written submissions. Until that approval is given, the allocation between the class and the costs and funding is a proposal, not an entitlement.

What Google actually fought

Google never applied to strike out the claim or for summary judgment. Of the six heads of objection it ran at certification, four attacked the litigation funding agreement or the adverse costs cover. The pattern held into 2026, when its application to vary the CPO was refused partly because it would have broken the funding.

Colour indicates the outcome from the class representative’s perspective. Green, objection rejected. Amber, provision amended by consent. Clay, relief granted to Google.

The funding architecture that survived

The structure that came through two years of attack is now close to a template, and worth setting out plainly.

The contracting funder was a Guernsey-protected cell, established solely for these proceedings. Behind it sat a Delaware limited partnership that assumed obligations directly, with uncalled limited partner commitments exceeding the LFA commitment and a record of never having failed a capital call. Adverse costs were covered by a £15m ATE policy carrying an anti-avoidance endorsement that Google could enforce directly against the insurer. The Tribunal accepted, following UK Trucks Claim, that it need not be satisfied at the outset that every possible cost could be met, only a substantial level of them.

The funder’s return was structured as a multiple of capital outlay rather than a share of damages, which is the standard response to PACCAR and remains the market’s working solution. Budgeted disbursements were £3.336m including VAT, of which £316,615 was allocated to the consultative panel. Solicitors’ and counsel’s fees were provided to the Tribunal separately and are not public.

Google spent two years attacking the funding rather than the merits and lost almost all of it. Underwriting should assume the funding agreement, not the pleaded case, is where the fight will be.

The time taken

Twenty-four months from filing to settlement is not what this regime is known for. The comparison with the other actions in the same courtroom, and with the regime’s best-known case, is stark.

The speed was bought, not found. Rodger attached itself to proceedings in which 3.5 million documents had already been disclosed, the market definition had already been argued, and a trial date already existed. The entry price was a £1.5m indemnity to the party whose trial date was displaced, and acceptance of a compressed timetable in which its own disclosure requests were repeatedly refused as impracticable at that stage.

That is a distinct funding proposition, and one worth naming: the late entrant into a mature docket. It shortens duration risk materially. It also means the claim inherits a timetable it did not set and gives up the ability to build its own evidential record.

What funders should take from it

1.  The £100m is not yet the funder’s money

This is the risk to watch on 15 September. In Merricks v Mastercard, Innsworth had spent roughly £45.5m. Its contractual entitlement under the funding agreement has been reported at over £520m, and it is reported to have sought £179m from the settlement fund. The Tribunal allowed it its expenditure plus a 50% profit, around £68m, in a case it described as having produced a very poor result. Innsworth’s judicial review of that decision was dismissed by the divisional court in June 2026, which upheld the Tribunal’s discretion as an expert body, and Innsworth was ordered to pay the class representative’s costs.

Contractual entitlement under an LFA is not the same as recoverable entitlement from a settlement fund. Any model that treats the LFA multiple as the return has a missing variable.

2.  The ratchet works, and there is now evidence for it

Defendants routinely attack step-up multiples as creating perverse incentives against settlement. Here the Tribunal rejected that argument, predicted that the defendant would have its own incentive to settle before the step-up, and was proved right within eighteen months. Funders defending uplift structures at certification now have a worked example, not just a theory.

3.  Underwrite the funding agreement as hard as the merits

Four of six certification objections, and the 2026 variation application, went to the money rather than the case. The provisions that were conceded are instructive: a termination right triggered by failure to follow settlement advice, and a material adverse change definition that referred to the solicitor’s and the representative’s own commercial return. Both are the kind of drafting that reads as ordinary commercial protection to a funder and as control of the class representative to a tribunal.

4.  Duration assumptions can be revised, in the right structure

Two years from filing to settlement is a genuine data point, but it is not general. It reflects a claim that entered an existing docket and took the timetable it found there. A standalone claim in the same market would still be looking at five or six years.

5.  Recovery ratios are worse than the headlines

Press coverage has anchored on £260m against a claim of over £1bn. The certification range before interest was £374m to £859m, and the class receives £160m of the £260m. Anyone calibrating expected recovery from press figures will be optimistic by a wide margin.

6.  The regulatory position is still unresolved

The Civil Justice Council recommended reversing PACCAR in June 2025. The Ministry of Justice announced on 17 December 2025 that it would legislate and would introduce proportionate regulation of third-party funding. No bill appeared in the King’s Speech on 13 May 2026. Multiple-of-outlay structures, as used here, remain the working solution rather than a settled legal position.

Caveats to consider

• The composition of the £100m is not public. Whether and to what extent it comprises the funder’s return, the ATE premium, and solicitors’ and counsel’s fees is an inference from the certification judgment and press reporting, not a disclosed fact. The Tribunal will see the breakdown on 15 September.
• The neutral citation [2026] CAT 49 for the variation judgment is taken from the Tribunal’s case page and should be checked against the judgment itself before being cited in writing.
• Reported figures for what Innsworth sought in Merricks differ between sources, with £179m and a contractual entitlement of more than £520m both reported. The £68m awarded and the £45.5m spent are consistently reported.
• No appeal against certification appears on the Tribunal’s case page for this claim. That is an observation from the public record rather than a confirmed absence.
• The class size of roughly 2,200 comes from the 2025 certification. The settlement class period runs two years after the certified period, to 31 July 2026, so the class is larger than that figure.
• This note is drawn from the Tribunal’s published judgments and from press reporting. It is not based on the settlement agreement, which we have not seen.

Sources

• Competition Appeal Tribunal, case page 1673/7/7/24, Professor Barry Rodger v Alphabet Inc and others
• Judgment (CPO Application) [2025] CAT 45, 6 August 2025
• Judgment (Joint case management) [2025] CAT 25, 30 April 2025
• Ruling (Disclosure) [2025] CAT 58, 25 September 2025
• Judgment (CPO Variation Application), 10 June 2026
• Fountain Court Chambers, “Tribunal refuses Google’s application to vary certification”, June 2026
• Law Society Gazette, “Funder fails in JR challenge to Mastercard claim distribution”, June 2026
• Macfarlanes, “Merricks v Mastercard settlement: a landmark judgment and its implications for funders”
• Akin, “UK Government omits PACCAR fix from 2026 King’s Speech”; Dechert, “UK Government to legislate on litigation funding”, January 2026
• Press reporting of the settlement: Reuters, Financial Times, Bloomberg Law, City AM, EU Today, 27 to 31 August 2026

Ashdown Litigation Partners Ltd is a commercial introducer and facilitator of litigation finance opportunities, registered in England and Wales, company number 16015628, at 71 St James’s Drive, London SW17 7RW.

Ashdown Litigation Partners is not a regulated company. It does not offer any regulated products or services. Nothing in this note is legal, investment or financial advice, and it should not be relied on as such. It is a summary of published tribunal judgments and press reporting prepared for general information. Anyone considering an investment in legal assets should take their own professional advice and verify the primary sources.

Prepared 1 September 2026. The settlement described has not been approved by the Competition Appeal Tribunal as at the date of publication.