Consent-free analytics: how to measure your audience without a banner

Half the web shows a consent banner solely because of its audience measurement tool. That is a choice, not a fate: EU privacy rules explicitly allow an exemption for respectful audience measurement. Here is how to activate it, cleanly.

Quick Answer: can you measure without a banner?

Yes β€” EU rules explicitly allow it. The ePrivacy framework (article 82 of the French Data Protection Act, PECR in the UK) exempts audience-measurement trackers that meet strict criteria:

  • strictly statistical purpose;
  • anonymous data, no cross-site tracking;
  • no cross-referencing or sharing with third parties;
  • trackers capped at 13 months, data at 25 months.

Two roads to get there:

  1. Configure a generic tool along the regulator's guide β€” possible but fragile (Matomo loses e-commerce tracking in exempt mode).
  2. Pick a tool exempt by design, with nothing to configure.

The immediate gain beyond compliance: you measure 100% of your traffic instead of the 50–70% who accept banners.

Mirage Analytics dashboard: complete audience measurement without a consent banner

Why the banner exists β€” and why it can go

Prior consent applies to any tracker that is not strictly necessary for the service (the ePrivacy directive; PECR in the UK, article 82 of the French Data Protection Act). Google Analytics, with its identification cookies and advertising ecosystem, falls squarely inside β€” hence the banner.

But the same rules allow an exemption for audience-measurement trackers, under strict conditions. The regulator's logic: knowing how many people read a page, anonymously, does not threaten privacy. The condition: the measurement must be incapable of following individuals.

The exemption conditions, concretely

Taking France's CNIL β€” the most explicit European regulator on this β€” the measurement tool must:

  1. Serve audience measurement exclusively β€” traffic statistics, content performance, technical issue detection. No advertising, no scoring.
  2. Produce anonymous data β€” no identifier that can follow a person across visits, anonymised IP.
  3. Not cross-reference the data with other processing (CRM, ad platforms) nor share it with third parties.
  4. Not track visitors across sites (no cross-site tracking).
  5. Limit tracer lifetime (13 months max) and data retention (25 months max).

Our dedicated guide covers the history and the tools assessed by the CNIL. The crucial point: exemption is a property of the tool and its configuration, not a statement of intent. An exemptible tool badly configured is not exempt.

Two roads to exemption

The configuration road: possible, but fragile

Some generic tools can be configured into compliance β€” Matomo is the documented case, with a dedicated CNIL guide: IP anonymisation, visitor logs disabled, opt-out, no User ID, no e-commerce tracking. The paradox: configured correctly, the tool loses key features; and one mistakenly unticked box takes you out of the exemption with no warning.

The by-design road: exemption by default

The alternative: a tool built for exemption β€” anonymous by architecture, cookieless, with nothing to configure. Impossible to "misconfigure" since there is nothing to configure. That is the approach of modern privacy-first analytics, covered in our Google Analytics alternatives comparison.

In practice with Mirage β€” Mirage is exempt by design: no cookies, no persistent identifiers, anonymised IPs, data in France, and conversion tracking works in exempt mode β€” e-commerce included. Proof by example: this site runs Mirage and shows no banner. See how it works.

What exemption changes in your numbers, concretely

Migrating to exempt measurement produces an immediate and often unanticipated effect: your traffic curves go up. Nothing changed on the site β€” you simply see the visitors the banner was hiding. Practical consequences:

  • Recalibrate your references. Monthly targets, year-on-year comparisons and alert thresholds were set on a measurement missing 30–50%. Plan a methodology note at the switch date.
  • Channels rebalance. Banner refusal rates vary by audience: social and mobile traffic refuse more than professional desktop. Your "real" channel mix may differ noticeably from what GA showed.
  • Conversion rates drop optically. Same conversions, a larger and more honest denominator: your rates get closer to reality. That is a correction, not a regression.

The classic objections, and what they are worth

"Without cookies I lose returning-visitor tracking." Partially true: individual-level return tracking disappears. But aggregates (estimated new vs returning, anonymous cohort loyalty) remain measurable, and the core of steering β€” pages, channels, conversions, journeys β€” doesn't depend on it.

"My marketing team needs remarketing." Advertising remarketing was never covered by the exemption: it requires consent, with or without your measurement tool. Separating measurement (exempt, 100% of traffic) from advertising (consented, opt-in) is precisely what clarifies that debate.

"Google Analytics is free, why pay?" Because "free" is paid elsewhere: a CMP to maintain, 30–50% of data lost, a legal risk documented by European regulators, and your traffic data feeding the advertising ecosystem of the world's largest ad seller. The full calculation is in our comparison.

Beyond France: the European map

The analytics exemption is a French reading of ePrivacy, but the logic spreads: the German authority accepts consent-free audience measurement under similar conditions, and several authorities (Austria, Italy, Denmark) have mostly hardened the other side β€” the unlawfulness of transfers to US tools. For a multi-country EU site, the robust strategy depends on no local tolerance: anonymisation by design + EU hosting, the base that satisfies the CNIL and its counterparts simultaneously. Our cookieless analytics guide covers the technical architectures.

The 5-step action plan

  1. Inventory your trackers. Open DevTools (Application β†’ Cookies) on your key pages: who sets what? Often, analytics is the only non-essential tracker.
  2. Pick your exempt measurement tool β€” by design, or configured along the regulator's guide if you keep a generic tool.
  3. Migrate properly: both tools in parallel for 2–4 weeks to validate the numbers (expect to see MORE traffic: you recover the banner refusers).
  4. Adjust the banner. Analytics was the only cause β†’ remove it. Other trackers remain β†’ remove analytics from the purposes list.
  5. Document. Record of processing, updated privacy policy, mention of the tool and its exemption compliance.

What you gain beyond compliance

  • Complete data: no more 30–50% hole from refusals. Your conversion rates and channel analyses become representative again.
  • A faster, cleaner page: without a CMP and ad scripts, loading gets lighter.
  • A frictionless experience: the banner is the first thing every new visitor sees. Its disappearance is an immediate UX win β€” some even make it a brand statement.

FAQ

Yes, provided the tool meets the exemption criteria that EU regulators derive from the ePrivacy directive: a purpose strictly limited to audience measurement, anonymous data, no cross-site tracking, no sharing with third parties. France's CNIL documents these conditions precisely; other EU regulators follow similar logic.

Only if analytics was your only consent-requiring tracker. If your site also loads advertising pixels, tracking social buttons or embedded videos with cookies, consent remains required for those. Many brochure and editorial sites can indeed drop the banner entirely.

Setting trackers without consent is one of European regulators' most frequent sanction grounds, with fines ranging from tens of thousands of euros for an SME to hundreds of millions for the giants. Enforced compliance orders with daily penalties are the common scenario.

Is conversion tracking possible in exempt mode?

Yes, if conversions are counted anonymously and in aggregate, without individual profiling or cross-referencing with other processing. That is an advantage of tools designed for exemption; by contrast, the CNIL's Matomo guide explicitly excludes e-commerce tracking from the exempt perimeter.