Chrome Turned Off Third-Party Cookies for 1% of Users. Here Is Our Plan

It actually happened. On January 4th, Chrome began restricting third-party cookies for one percent of users worldwide. After four years of delays, false starts, and regulatory back-and-forth, the deprecation clock is finally running, with Google saying the remaining 99 percent follows in the second half of the year.

One percent sounds small. It is roughly 30 million people, which makes it the largest measurement experiment in advertising history, and it is big enough that sloppy setups are already showing cracks. Here is how we are triaging for clients.

What breaks first

The immediate casualties are the obvious ones: third-party retargeting pools shrink, view-through attribution gets patchier, and audience match rates on the open web drift downward. If your media plan leans heavily on prospecting through third-party audiences, you will feel this before your analytics team can explain it.

What does not break: anything built on first-party data, contextual targeting, and the walled gardens, which run on logged-in identity and are, not coincidentally, positioned to absorb the budgets that flee the open web. Have opinions about that dynamic, but plan for it either way.

The work we are doing now

We have spent January running the same audit for every retained client, and we would suggest any team do the equivalent:

• Inventory every pixel and tag, and classify each as first-party, third-party, or mystery. The mystery pile is always bigger than anyone expects.
• Verify server-side tagging or a first-party collection endpoint for your core analytics, so your own measurement does not degrade alongside the ad tech.
• Get conversion APIs live for the major platforms rather than relying on browser pixels alone.
• Pressure-test attribution by comparing platform-reported conversions against actual revenue in a spreadsheet. The gap you find now is the baseline you will need later.

For Northwind, whose retail business generates genuinely rich first-party signals, the bigger opportunity is offense rather than defense. Loyalty data, purchase history, and on-site behavior become more valuable with every cookie that dies, provided the consent and infrastructure are in place to use them. They are further ahead than most, and it is starting to look like a competitive moat.

What can wait

You do not need to pick a Privacy Sandbox strategy this quarter. The Topics API, Protected Audience, and Attribution Reporting are all real and all immature, and the UK competition regulator is still actively reviewing whether the whole scheme is fair to the rest of the ad industry. Committing hard to sandbox-specific infrastructure before that resolves is a gamble with little upside. Watch, test small if you are curious, and keep your powder dry.

You also do not need to panic-buy an identity resolution vendor. Some of these products are good. Many are expensive ways to feel busy. The unglamorous first-party work above beats all of them on return.

The honest caveat

Google has delayed this deprecation three times, and the regulatory pressure has not gone away. A fourth delay would surprise nobody in this office. But here is the thing: every action on the list above is worth doing even if third-party cookies somehow live forever. First-party data, honest attribution, and consent infrastructure are simply better marketing hygiene. The deadline is just the excuse.

Plan as if H2 2024 is real. Build things that are valuable even if it is not.