The death of third-party cookies, one year in
What actually broke, what did not, and why contextual targeting is suddenly interesting again to buyers.
A full year after third-party cookie deprecation finally stuck, the post-mortem is far less dramatic than either side predicted. The web did not break. It also did not get noticeably better for users. It just got more uneven.
What broke
Frequency capping and view-through attribution took the hardest, most obvious hits. Retargeting degraded but did not vanish — it simply got more expensive and less precise.
What quietly got better
Two things improved. Contextual targeting became genuinely interesting again, with real investment behind it rather than the apologetic afterthought it had been. And first-party data finally got the board-level attention publishers had been asking for since 2019.
The cookie did not die so much as it stopped being an excuse not to do the harder work.
The winners and losers
The outcome split cleanly:
- Publishers with logged-in users and real first-party data gained leverage.
- Pure arbitrage and commodity inventory lost their cheapest targeting signal.
- Walled gardens, predictably, were largely unaffected.
What comes next
The interesting question for 2026 is not how to replace the cookie. It is whether the privacy sandbox alternatives are worth the integration cost, or whether contextual plus first-party is simply good enough. For most independent publishers, good enough is winning.