For five years, display advertising planning was organised around a date that never arrived. Third-party cookies were going away, Privacy Sandbox would replace them, and every agency deck contained a slide about preparing for the transition. Neither half of that prediction held.
Google discontinued the Privacy Sandbox initiative in October 2025, retiring most of its core APIs — Topics, Protected Audience and Attribution Reporting — citing low adoption and continued regulatory pressure. Separately, in July 2024, Google had already announced it would not deprecate third-party cookies in Chrome, opting instead for a user-choice mechanism.
So the replacement was withdrawn and the thing it was replacing stayed. That is an unusual outcome, and it leaves advertisers in a position nobody planned for.
The number that matters
Third-party cookies remain functional in Chrome for most users, but their reliability has declined meaningfully. The practical measure is match rate, and it has settled at roughly 47% on authenticated traffic across the open web — well below the 68% typical of the cookie era, but workable.
That gap is the whole story. A 21-point decline in match rate does not break targeting; it changes the economics of it. Campaigns built on assumptions from 2021 will underdeliver against frequency caps, over-count unique reach, and misattribute conversions — not because anything failed, but because the denominator moved.
What buyers actually run now
The stack that emerged is hybrid rather than a single successor technology:
- Authenticated identifiers — UID2, ID5, RampID
- Contextual signals — the pre-cookie approach, rebuilt with better classification
- Clean-room measurement — for attribution that no longer survives naive last-click
- First-party data — now the anchor rather than a supplement
The shift in publisher sentiment is sharp. In Q1 2025, 71% of publishers identified first-party data as a key driver of advertising results, up from 64% the previous year, and 85% expect its role in monetisation to grow further. The importance attached to third-party data is declining rapidly — not because the cookies disappeared, but because their reliability did.
The market did not shrink
It is worth putting the disruption in proportion. Global programmatic ad spend reached an estimated $755 billion in 2025 and is projected to clear $821 billion in 2026, representing roughly 90% of all digital display investment worldwide.
An industry in existential crisis does not grow 9% year on year. What happened is narrower: the addressability layer got noisier, and the operators who adapted their measurement absorbed it while those who did not saw performance decline for reasons their dashboards could not explain.
What to change in your planning
Re-baseline your reach numbers. If your unique-reach figures come from a cookie-era model, they are wrong now. Not slightly — a 47% match rate against an assumed 68% inflates unique counts substantially.
Stop treating contextual as a fallback. It was the default before behavioural targeting and it is a first-class option again, particularly for upper-funnel work where identity resolution adds little.
Audit what your DSP is actually using. Some platforms still describe Privacy Sandbox integration in their documentation. Given the programme was retired, that is a question worth asking your account team directly.
Build the first-party asset now. Publishers moved first because the pressure hit them first. Advertisers are one step behind on the same curve.
The lesson is uncomfortable for anyone who spent five years preparing for a specific transition: the disruption was real, but it did not arrive in the shape that was forecast. It arrived as gradual signal decay, which is harder to notice and harder to plan around than a deadline.
