Cookieless analytics on a landing page: what French regulators actually exempt, and what it changes for your numbers

On most French landing pages, Google Analytics 4 is installed by default, and the cookie banner comes with it — because GA4 drops an identifier that follows the visitor from one session to the next, which disqualifies it from the consent exemption. Two consequences pile up silently: some visitors close or ignore the banner without ever accepting, and another share blocks the google-analytics.com request outright with an ad blocker, before even seeing the banner. Either way, that visitor exists, may well have converted, and shows up in no report at all. There is a less-discussed alternative to 'make the banner prettier': configure an audience-measurement tool so it fits within the French data authority's (CNIL) exemption framework — and skip the banner entirely.

The double hole in GA4's numbers

The first hole is well known and covered in our article on the cookie banner: a meaningful share of visitors close the banner without choosing, which counts as a refusal. The second hole gets less attention: even among visitors who do accept, or who never saw a banner because it was implemented poorly, a portion of traffic never reaches GA4 because the browser or an extension blocks the request upstream. A 2017 study by Iqbal, Shafiq, and Qian presented at the ACM Internet Measurement Conference ("The Ad Wars: Retrospective Measurement and Analysis of Anti-Adblock Filter Lists") measured at scale how community-maintained filter lists (EasyList, EasyPrivacy) target the most widely used audience-measurement domains first — Google's leading the pack, precisely because they are the most widespread. In other words: the more popular a tracking tool is, the more it gets blocked, and GA4 sits mechanically at the top of that list. On a B2B or tech-savvy audience, the gap between actual and measured traffic can run into double-digit percentages — a bias no setting can fix, only a change of tool or collection method.

What French regulators actually exempt — and why GA4 doesn't qualify

The CNIL accepts that an audience-measurement tool can skip consent, provided it simultaneously meets several strict criteria:

• the data serves only to measure the site's own audience for its own publisher, never cross-referenced with other processing;
• no transmission to any third party, including for advertising or cross-site matching purposes;
• the tracker cannot follow the visitor across other sites (no cross-domain tracking);
• the visitor is informed of its existence and can easily object;
• lifespan capped at 13 months, with data retention capped at 25 months.

GA4 fails these conditions by design: the data flows through Google's servers, feeds into the group's advertising ecosystem (Google signals, remarketing, cross-site benchmarking), and potentially serves purposes beyond pure audience measurement for the site itself. The CNIL has explicitly confirmed this — GA4 cookies do not qualify for the exemption. By contrast, a tool like Matomo, run in a compliant configuration (EU hosting, anonymized IP, no data sharing, easy opt-out), is among the solutions recognized as compatible with this framework. The CNIL has even launched a voluntary evaluation program to determine objectively which market tools genuinely meet these criteria — a sign that the topic keeps moving and deserves re-checking with every tool change, not a one-time assumption.

What you actually gain on a landing page

The most visible benefit is the outright disappearance of the consent banner for audience measurement — one less visual and cognitive obstacle on the first screen, as described in our anatomy of a landing page that converts. The second benefit is less visible but matters more to anyone running campaigns: numbers that no longer depend on an 'Accept' click or a third-party filter list, so a measured conversion rate closer to the real one — a basic requirement for any reliable A/B test, since a measurement bias that varies by browser or by how a visitor feels about the banner distorts the comparison between variants. There's also a trust effect that's harder to measure but real: a 2017 review by Martin and Murphy published in the Journal of the Academy of Marketing Science ("The Role of Data Privacy in Marketing") brings together research showing that the perception of excessive data collection erodes consumer trust in a brand, and that this trust directly shapes purchase intent. A form that isn't preceded by an intrusive consent request fits that logic — not the main reason to switch tools, but a consistent effect with the rest of the page.

What you lose, and the honest trade-off

Moving audience measurement to an exempted tool doesn't get you out of a banner for everything else. If the page uses a Meta pixel or a Google Ads conversion tag for retargeting, those advertising trackers remain outside the exemption and still require explicit consent before firing — see our article on the Meta pixel and the Conversions API. The real gain applies to audience measurement itself (page views, traffic sources, overall conversion rate), not to fine-grained ad attribution, which stays subject to the same rules as before. Another trade-off: the Google ecosystem (direct import into Google Ads, Looker Studio, campaign machine learning) is built around GA4, and a third-party tool requires manual integration to recover those conveniences. Many teams end up with a hybrid setup: an exempted, banner-free tool for general audience measurement, plus a second, consent-gated tracker reserved for visitors coming from paid campaigns meant to be retargeted — consistent with the approach already described for a compliant cookie banner.

Migrating without losing track of your numbers

1. Check the exact configuration required for the exemption before removing the banner — a 'privacy-friendly' tool set up carelessly (unanonymized IP, data sharing turned on) doesn't automatically qualify.
2. Run the old and new tools in parallel for two to four weeks to compare orders of magnitude, without expecting a perfect match — every tool counts differently.
3. Redefine the main conversion event in the new tool using the same logic as for GA4: one event, not a page-view count.
4. Only remove the banner once compliance is verified — keeping it up for a while alongside the new tracking costs nothing and de-risks the transition.
5. Document the choice (tool, configuration, date) in your GDPR record of processing activities, as with any change covered in our GDPR-compliant form guide.

This work is independent of which template you use: the ten LanderKit templates (€89 each, €229 for the bundle) ship as Next.js projects with no database and no tracking script baked in — the slot to plug in GA4, Matomo, or any other solution is left open, and nothing stops you from starting with an exempted tool from launch instead of fixing it later. To see actual visitor behavior beyond aggregate numbers, a heatmap remains a useful complement, as long as it follows the same consent rules.