Google Ad-Tech Antitrust Litigation: A Publisher’s Playbook

If you depend on programmatic advertising revenue, the Google ad-tech litigation creates a planning problem before it creates a financial opportunity. The wrong response is to put a recovery into your forecast or make a rushed platform change. The useful response is to determine whether your business touches the surviving claims and whether you can still explain, with records, how money moved through your ad stack.

Major claims remain alive, but that is not the same as a finding that every publisher was harmed. Your immediate job is to separate what the court has established, what the publishers still must prove, and what evidence your own legal and finance teams would need to evaluate any potential exposure or recovery.

The ruling preserved a path, not a payout

On Sept. 30, U.S. District Judge P. Kevin Castel issued an 88-page opinion denying Google’s requests for summary judgment on the publishers’ principal ad-tech claims. He also declined to exclude important expert testimony supporting their damages cases.

Summary judgment is a pretrial mechanism for resolving claims that do not require a trial to decide. Denying it means Google did not persuade the court to dispose of the principal claims on the pretrial record. It does not mean the publishers have won a damages award, that every expert assumption has been accepted, or that every remaining dispute will necessarily reach trial.

The mixed outcome matters. Castel ruled for Google on the New York General Business Law claims brought by Gannett and Daily Mail, on claims brought by The Progressive, and on Inform’s federal antitrust claims. Claims involving the AdX publisher class and Mikula Web Solutions were allowed to continue. A headline saying publishers cleared a major hurdle is accurate, but it is too broad to answer whether a particular company, legal theory, or alleged loss remains in play.

When you brief executives, use a claim matrix rather than a win-or-loss label. Give each claimant and legal theory its own row, then record whether the claim survived, which issues are already established, which issues remain disputed, and what procedural event comes next. That prevents a partial ruling from turning into an inaccurate company-wide assumption.

The economic dispute sits between inventory and demand

A publisher ad server manages advertising inventory and helps decide which demand source can fill an opportunity. An exchange provides a marketplace in which demand can compete for that inventory. When one company controls important infrastructure on both sides of that handoff, the rules connecting the products can affect which demand participates, how an auction operates, what fees are charged, and what reaches the publisher.

That connection is central here. The publishers allege that Google’s control over its publisher ad server and the AdX exchange, combined with practices governing ad auctions, reduced publisher revenue or produced excessive fees. Google contests those allegations. The disputed question is therefore not simply whether publishers used Google technology; it is whether challenged conduct caused a measurable economic injury.

The litigation also draws on the federal government’s separate ad-tech case in Virginia. Castel had already determined that Google could not relitigate certain findings from that proceeding, including the finding that Google unlawfully tied its publisher ad server to AdX. That gives the publisher plaintiffs an important established point, but it does not calculate the consequences for a particular publisher. Injury, causation, and damages still have to be connected to the conduct at issue.

For your business, the practical unit of analysis is an ad-monetization dependency map. It should show:

• The legal entities, sites, applications, and business units that sold digital inventory.
• The publisher ad server and exchanges used during each relevant period, including migrations and material configuration changes.
• Which demand paths were direct, exchange-based, mediated, or otherwise dependent on the publisher ad server.
• The contracts, amendments, fee schedules, invoices, and reporting accounts associated with each path.
• The identifiers that connect domains, properties, accounts, reports, and payment records across systems.
• The employees or vendors who understood auction configuration, yield management, billing, and reporting definitions at the time.

This map does not establish that you belong to a class or have a claim. It gives counsel the facts needed to assess those questions without relying on institutional memory. It also reveals whether a change in revenue coincided with traffic, inventory, auction, fee, or platform changes instead of treating every decline as one undifferentiated problem.

Do not mistake the damages estimates for recoverable amounts

The public figures are large because they are damages estimates prepared by experts retained by the plaintiffs. They are not court-awarded compensation:

The plaintiffs claim additional class damages after March 31, 2024, but no additional amount was provided. Do not extend the $1.72 billion estimate beyond that date, apply it as a percentage of industry revenue, or use it to derive a hypothetical recovery for your company. None of those calculations is supported by the disclosed figures.

An expert’s testimony can remain admissible while its assumptions, method, causal reasoning, and conclusions remain disputed. The publishers still need to prove that the challenged conduct injured them and that the requested damages are attributable to that conduct. Google can continue contesting those points.

Your finance team should therefore treat the amounts as allegations supported by the plaintiffs’ models, not as receivables or operating income. If you need an internal scenario, build it in layers:

1. Use zero recovery as the operating baseline unless legal and accounting advisers determine otherwise.
2. Ask counsel whether the relevant legal entity, products, time periods, and transactions could fall within a surviving claim or class.
3. Identify which revenue, fee, and auction records could support or contradict economic injury.
4. Document every assumption in any contingent scenario, including eligibility, time boundaries, allocation method, legal costs, and uncertainty.
5. Keep the scenario outside normal performance targets so an unresolved lawsuit does not distort hiring, content, or technology decisions.

The same restraint applies to vendor decisions. A surviving antitrust claim is not proof that your current contract is invalid, that a migration will improve yield, or that another stack will produce a particular result. Evaluate a change using your own fees, demand access, reporting quality, operational cost, and measured auction outcomes.

Build a counsel-led evidence pack while the systems are identifiable

This is an operational preparation checklist, not a determination that your company is part of the litigation or subject to a legal-hold obligation. If the surviving claims may be relevant to your business, ask qualified antitrust or litigation counsel to assess eligibility and preservation duties. Do that before changing retention policies or launching a broad data collection.

1. Create a system inventory. Record each ad server, exchange, reporting interface, billing system, data warehouse, and archive, along with its owner and available date range.
2. Preserve the commercial record. Locate contracts, order forms, amendments, invoices, payment statements, fee disclosures, account notices, and documents explaining platform migrations or material configuration changes.
3. Preserve the operational record. Identify ordinary-course auction reports, revenue reports, configuration histories, demand-partner lists, account identifiers, and metric definitions. Record where a field was renamed or calculated differently over time.
4. Build a dated chronology. Align platform changes with shifts in impressions, fill, auction participation, reported fees, and net publisher revenue. A chronology makes alternative explanations visible instead of assuming every movement came from the challenged conduct.
5. Reconcile money to activity. Where the data permits, connect inventory and auction records to invoices and net payments. Record unexplained gaps rather than backfilling them with estimates.
6. Document limitations. Note missing periods, expired logs, acquired properties, changed account IDs, inconsistent currencies, and reports that cannot be reproduced. A known limitation is more useful than false precision.
7. Control access. Keep the working set limited to the people who need it, and follow counsel’s directions for preservation, privilege, privacy, security, and collection scope.

Do not delete, rewrite, or normalize potentially relevant originals after counsel identifies a preservation obligation. At the same time, do not collect extra user-level information merely because it might be available. An indiscriminate collection can create privacy and security exposure without helping establish publisher-level fees or revenue. Preserve what is relevant, document what each field means, and let counsel define the defensible scope.

SEO, content, and audience teams also have a role. Keep acquisition performance separate from monetization performance in your reporting:

• Acquisition: visits or sessions from organic search, AI-search referrals, direct traffic, social platforms, and other channels.
• Inventory: ad opportunities, eligible impressions, ad load, and fill-related measures available in your systems.
• Monetization: auction outcomes, disclosed fees, and net publisher revenue, with the governing metric definitions attached.

Traffic can improve while monetization weakens, or monetization can improve while traffic falls. A single blended revenue-per-session figure hides that distinction. Separating the layers helps you evaluate content performance accurately now and gives legal and financial reviewers a cleaner record if they later need to isolate alleged ad-tech harm.

Key takeaways for publisher teams

• The Sept. 30 decision kept principal Google ad-tech claims alive and preserved key expert testimony; it did not award damages.
• Google cannot relitigate certain findings from the separate Virginia case, including unlawful tying of its publisher ad server to AdX, but publisher-specific injury and damages still require proof.
• The estimates of roughly $901 million for Gannett, $600 million for Daily Mail, and $1.72 billion for the publisher class are plaintiffs’ expert estimates, not payouts.
• The result varies by claimant and legal theory. Several claims were resolved for Google, while claims involving the AdX publisher class and Mikula Web Solutions continue.
• Your defensible next step is a counsel-led review of eligibility, systems, contracts, fees, and retained data—not an assumed recovery or an emergency platform migration.

Within your next reporting cycle, produce a one-page ad-stack dependency map and assign owners for the supporting contracts, reports, and payment records. Have counsel decide whether a deeper eligibility or preservation review is warranted. That gives you a decision-ready file without pretending the litigation has already produced money for publishers.

References

• Search Engine Land — Publishers’ Google ad-tech antitrust claims clear a major legal hurdle