Cookies are not the only measurement future

What to watch

Attribution Reporting for conversion measurement, Topics for interest, Protected Audience for remarketing. Each has a registration and a report delay. None of them accept a 1x1 GIF with a query string as the primary API. You will not paste a Topics value into Pixellint and get a Floodlight ord check.

CHIPS partitions remaining third-party cookies by top-level site. Bounce tracking mitigations expire storage that looks like a tracker. None of that is a GTM bug. Read it as: identifiers you do not collect on your site will not be there later. If your only story was a cookie set by facebook.com or doubleclick.net, match rates are a history lesson.

Not a pixel host

Do not traffic a conversion pixel into a Topics slot. Do not expect Topics to fire Purchase with value and currency. If you need conversion measurement without a third-party cookie, that is Attribution Reporting, enhanced conversions, CAPI, or a first-party collect path, not an interest API.

The google-ads-conversion pack still checks conversion_id and label on the conversion image. That image may keep working for users who still have a cookie. It is not Attribution Reporting. Do not delete the image because a Sandbox demo worked in one Chrome profile. Do not keep only the image because Safari already killed the cookie.

https://www.googleadservices.com/pagead/conversion/123456789/?label=AbC-D_efG&guid=ON&script=0

What not to pause

First-party click ids, hashed PII CAPI, Consent Mode. Those keep working while Sandbox ships, unships, and ships again. First-party cookies on your domain still exist, and ITP still shortens them. Chrome did not flip a single kill switch. It gave users a choice and shipped APIs for sites that cannot rely on the old cookie.

Privacy Sandbox does not waive GDPR or a CMP. A Topics request is still processing. A Protected Audience auction is still ads. Wire the banner and the IAB strings for the hops that still fire. Pixellint core will still decode gdpr_consent on those hops. It will not decode a Topics header.

QA across browsers

Do not use a Chrome-only Sandbox success as proof the Meta pixel is healthy on iOS. Different browsers, different graphs. QA still starts with the fired URL for the tags that still fire. Sandbox reports will not look like a pixel HAR.

If your only debugger is the Network tab, you will declare Attribution Reporting broken because you cannot see a g/collect twin. Budget a different debug path, and keep the pixel contract for facebook.com/tr and the Ads tag. Your Meta pixel did not become illegal. It became incomplete.

CAPI is the portable layer

A server Purchase with hashed email, click id, and a real event_time works in Safari, Firefox, and Chrome. Topics does not. Attribution Reporting does not replace Meta CAPI. Run them as layers. Dedup is a dictionary you write, not a Sandbox feature that deletes facebook.com/tr.

pixellint validate json on the CAPI body. pixellint validate url on the pixel that still fires. Neither call is a Privacy Sandbox conformance test. Google ships those APIs. Pixellint is not affiliated with that program. Keep the contracts you can still paste.

Check the artifact

Paste the pixel URL or JSON body into the playground. Same engine as pixellint validate. Nothing leaves the tab.