Cookieless Conversions: How Much You're Really Losing, and How to Get It Back

Four independent forces are behind the conversions missing from your ad panel, not one Google decision: Safari deletes script-set cookies after 7 days (WebKit, ITP 2.1), ad blockers strip tags before they load, part of EU traffic declines marketing consent, and some customers click an ad on their phone and buy on a computer. How much you lose depends on your traffic mix, but Safari alone shows the scale: in Poland it accounts for 25% of mobile traffic (StatCounter, September 2026). For those users, a conversion that comes more than 7 days after the click loses its link to the ad if the click cookie is set by a script in the browser. And that is only the first of four sources of loss. In July 2024, Chrome officially abandoned its plan to remove third-party cookies, so the media narrative about the "end of cookies" quieted down. That changes nothing about how many conversions you see in Google Ads and Meta Ads. Here's where the data gap comes from, how GA4 and Google Ads try to model around it, and what brings the numbers in your panel closer to what's in the bank account.

What "cookieless measurement" actually means in 2026

Cookieless measurement is a situation where part of your site's traffic can't be tied to a specific user through a third-party cookie, regardless of whether the cause is a browser setting, an ad blocker, or missing consent. Contrary to popular belief, Chrome never removed third-party cookies: in July 2024, Google abandoned that plan and announced a choice screen instead, and in April 2025 it dropped the choice screen as well (Privacy Sandbox announcement). Third-party cookies work in Chrome, and users can turn them off in the browser's privacy settings.

So the media myth of the "end of cookies" is largely outdated, but the conversion-measurement problem has existed for years for other reasons: Safari (the WebKit engine, a sizeable share of mobile traffic) has for years deleted JavaScript-set cookies after a maximum of 7 days by default, Firefox blocks third-party trackers right out of the box, and European consent rules require that a marketing script simply not fire at all without user acceptance. None of these three forces depends on Chrome's decision, and none of them will go away even if Google changes its mind again tomorrow.

Why you lose conversions even though Chrome hasn't blocked cookies

Conversions disappear from your panel because the data about them never reaches Google Ads or Meta. The transaction happened, but something along the way blocked the signal before it reached the ad platform. In a typical online store, four independent sources of loss stack on top of each other, and each needs a different fix:

These four forces compound: a store with a large share of iOS traffic, some Firefox users, and EU customers can miss a sizeable part of the conversions that actually happened. That's why a single fix, such as Consent Mode alone, never closes the whole gap. Each of the four loss sources needs its own answer.

GA4 models the conversions it can't measure: how accurate is the estimate

GA4 models key events it can't measure directly, estimating them statistically from the behavior of users who did consent. The mechanism requires Consent Mode v2 in advanced mode and only kicks in once data thresholds are met: for at least 7 days the property must collect at least 1,000 events a day without consent and have at least 1,000 daily users with consent on 7 of the previous 28 days (Google Analytics Help). Below those thresholds, GA4 doesn't model at all and just shows a hard gap.

Modeling is an estimate, and its accuracy depends on consent-data quality and traffic volume. The more users accept cookies, the more accurate the model. So a banner designed to maximize lawful acceptance, with a visible reject button, improves modeling quality and legal compliance at the same time.

Enhanced Conversions in Google Ads — first-party data instead of cookies

Enhanced Conversions sends Google Ads hashed (SHA256) customer data, such as email, phone, or address, collected right on your site, and Google matches it to signed-in Google accounts (Google Ads Help). Matching a conversion back to an ad click therefore doesn't depend on whether the click cookie (gclid) survived in the user's browser. The same mechanism runs for contact forms (Enhanced Conversions for Leads): form data replaces the cookie as the identifier.

Correctly implemented, Enhanced Conversions lift reported conversions by recovering events that used to vanish through cookie loss. Accounts with heavily European traffic and lower cookie consent usually have a lower match rate than US-market accounts, so comparing yours to overseas benchmarks can be misleading.

Meta CAPI and server-side tracking — the same logic, different platforms

Google Ads isn't the only platform solving cookie loss with first-party data sent from a server. Meta solves the exact same problem through Conversions API. We cover it separately, since the deduplication mechanism and the break-even point differ from Google Ads. If you want to tie CAPI, Enhanced Conversions, and TikTok data together in one piece of infrastructure, server-side GTM is a single server feeding every channel at once. Also check whether it's worth it at your traffic scale.

How to build measurement that survives cookie loss — a checklist

No single tool closes the data gap on its own. Only a combination of a few independent mechanisms brings your ad panel closer to what's actually happening. Before you start, check your baseline: the Consent Mode Checker shows whether the tags on your site respect consent, and the GA4 audit points out gaps in your measurement setup. The order below reflects what usually delivers the biggest return for the least effort.

• Consent Mode v2 Advanced — the prerequisite for GA4 modeling; without it, Google has no signal at all to estimate unconsented traffic from.
• Enhanced Conversions for web and for leads — hashed customer data as an identifier independent of the click cookie.
• Meta CAPI with event_id deduplication — the same events from the browser and the server, without double-counting conversions.
• Capturing email or phone as early as possible in the funnel — the more first-party data reaches your form, the higher your match rate across every channel at once.
• Server-side GTM once your ad budget reaches PLN 4,000 a month per channel — a single server extends cookie lifespan and feeds GA4, CAPI, and Google Ads at the same time.

None of these pieces solves the problem alone. Only the combination of Consent Mode, first-party data, and server-side infrastructure brings the numbers in your panel closer to what's actually happening in your bank account. As a marketing agency we build this whole stack alongside the campaigns themselves, within the same engagement.

Want to find out how many conversions you're losing? Let's talk: we audit your measurement setup together with your campaigns and show you what to fix first.

FAQ