In short
Ad retargeting shows ads to people who have already interacted with a business: site visitors, cart abandoners, past customers, app users. It works because intent is highest among people who have shown interest, so the return on retargeting spend is usually strong. What changed is the mechanism. Retargeting once ran on third-party cookie pools that followed users across the web, and as browsers blocked those cookies, that approach decayed. Post-cookie retargeting runs on first-party audiences instead: hashed customer lists uploaded as Customer Match or Custom Audiences, and site events delivered server-side rather than through a browser pixel alone. The practical consequence is that retargeting quality now depends on the same data discipline as everything else, namely clean identifiers, server-side event delivery, and consent carried through, because the audiences are built from data you own rather than cookies you borrow. Retargeting did not disappear with third-party cookies; its plumbing moved to first-party signal, and its quality moved with it.
How it works now
The cookie pool era is over. Today’s retargeting audiences are built from your own data: hashed emails and phone numbers synced to platform audiences, and site or app events delivered server-side so the platforms can match visitors to logged-in users.
What decides retargeting quality
Match rate. An audience of 100,000 customers that matches 40% reaches less than half the retargeting pool of the same list matched at 85%. Identifier hygiene and server-side delivery are where that difference comes from.
How does retargeting work without third-party cookies?
A brand’s cookie-based cart-abandon retargeting shrank as browsers blocked third-party cookies. Rebuilding it on server-side site events and a hashed customer list restored the audience without relying on cross-site tracking.
Reference: Google Ads Help, your data segments (remarketing)