Google's Impact on Attribution Models: What Advertisers Still Control

The complaint operators make about Google and attribution is usually framed as a grievance. It is more useful framed as a structural fact: the measurement layer for a large share of digital media is owned by a company that also sells the media, and that arrangement produces predictable effects whatever anyone's intentions are.

The question worth spending time on is not whether that is fair. It is which parts of your measurement you still control, and how to make those parts carry the weight.

What actually changed

Two things moved at once. Identifiers available to independent measurement became scarcer, and the reporting that replaced them moved inside platform environments where the buyer sees an output rather than the working.

Privacy is a legitimate reason for much of that, and some of the change would have happened regardless of who drove it. But the practical result for a buyer is the same either way: fewer places where an independent party can check a platform's account of its own contribution.

The cookie question itself has been less decisive than the industry expected — Google's 2024 reversal on the forced phase-out of third-party cookies left them in place. Building an attribution architecture that depends on them is still a bet on someone else's roadmap.

A platform grading its own work is a structural problem, not a moral one

This is not an argument about Google specifically. Any platform reporting on its own contribution has an interest in the result, and that is as true of a DSP as it is of a walled garden. It is true of the platform ThumbAd trades on.

Which is why the useful discipline is not choosing whichever platform you trust most. It is refusing to let any single platform's number be the one that governs a budget decision, and reconciling all of them against something you own.

What you do not control

You do not control what identifiers a platform exposes, what its models weight, or when either changes. Planning as though you might is how measurement architectures end up needing a rebuild every eighteen months.

What you do control — and it is more than it looks

• Your pixel and your audiences. Built in your seat, named to a documented taxonomy, portable when you leave. If that is a negotiation rather than a default, you have learned something about the partner.

• Which number governs. Platform-attributed or backend-confirmed, agreed before launch and written down.

• The attribution window. Fixed per objective in the contract, with changes requiring your signature. Window-shifting after results are visible is the oldest trick in performance reporting.

• The reconciliation. Monthly, in writing, platform figures against your own backend, with every gap explained: event definitions, time zones, currencies, deduplication, naming and UTM logic.

• The raw path data. Full journey exports on request. A partner who cannot produce them is not analysing them either.

None of that requires a platform's cooperation. All of it is available to an operator who asks for it in the contract rather than in the quarterly review.

What the missing measurement is worth

Across six independent measurements — five operators, two continents, six years — the post-view share of conversions landed between 89% and 96.2%. Every one was customer-ID-level or platform-verified. None was modelled. An operator reading clicks alone is seeing roughly one conversion in ten.

That is a statement about measurement completeness, not an incrementality claim — it says nothing about what would have happened with no media at all. The method is set out in click and view attribution in programmatic advertising.

The cost of leaving it unresolved is documented in The 2% Problem: six sportsbook operators, every account ended, and not one of them ended on performance. They ended because two systems disagreed and nobody reconciled them.

Testing a partner rather than trusting one

The 20-question partner scorecard runs ThumbAd's published due-diligence standard against any desk — including ours — and scores it against a bar of 24 out of 30. It runs in your browser, asks for no email, and nothing leaves the page. Our own answers to all twenty are published in full and ungated in How ThumbAd Answers the 20 Questions.

Control over attribution was never going to be handed back by a platform. It is assembled, clause by clause, by the operator who insists on it.

ThumbAd is an independent programmatic trading desk in Oslo, trading on The Trade Desk across Europe, Africa, North America and Latin America since 2012, with betting and gaming at the core.