Blog

What Is a Good Cookie Consent Rate? Real Benchmarks for 2026

If you ask five sources what an average cookie consent rate is, you will get five numbers, anywhere from 25% to 85%. They are all technically correct. They are just measuring different things.

Key Takeaways

  • 01The median ConsentStack site sees 72% of visitors grant full consent, measured strictly: every optional category accepted, essential excluded.
  • 02Published averages disagree because definitions differ. All-visitor studies report 42% to 47% acceptance; explicit-choice benchmarks run higher.
  • 03European visitors say yes to well-designed banners: 63% full opt-in in Germany and 72% in France, versus commonly cited norms below 25%.
  • 04Banner design is the biggest variable. Industry studies show acceptance ranging from 4% to 85% based on design alone, and the best-designed ConsentStack sites reach 89% full opt-in.
  • 05An equal-weight Reject button does not tank consent rates, and buried reject options are now a regulatory liability.

Here is the short answer, then the long one.

The short answer: across ConsentStack customer sites, the median site sees 72% of visitors grant full consent, meaning they accept every optional tracking category. Just under 7 in 10 of all consent decisions recorded across our customer base last month were full opt-ins. Top sites reach 89%. And that is measured the strict way: essential cookies do not count, and a visitor who allows only essential cookies counts as a rejection.

For context, aggregate studies across the industry put average acceptance between 42% and 47%, with individual sites ranging anywhere from 4% to 85% depending almost entirely on how the banner is designed.

Consent rate sounds like one metric. It is at least three, and vendors rarely say which one they are reporting.

1. What counts as consent? Some vendors count any positive interaction, including partial consent (the visitor allowed analytics but rejected advertising). Some count a visitor who clicked OK on a notice-only banner. The strictest version counts only full opt-in: every optional category accepted.

2. What is the denominator? Some studies divide by every visitor, including the large share of people who never touch the banner at all. Others, like Didomi's widely cited annual benchmark, divide by visitors who made an explicit choice. The same site can report 45% under the first method and 75% under the second.

3. What happens to essential cookies? Essential categories are always on, because the site does not function without them. A measurement that includes them can classify a visitor who rejected every single optional category as partial consent. That inflates the number while telling you nothing.

So when you see a consent rate with no definition attached, it is not a benchmark. It is a vibe.

How we measure, and what we found

Our numbers use the strict definition on every axis:

  • Full consent means full consent. The visitor accepted every optional category. Analytics, marketing, personalization, all of it.
  • Essential is excluded from the math. A visitor who keeps only essential cookies is counted as a rejection, not a partial win.
  • The denominator is explicit choices. We count visitors who actually made a decision on the banner.
  • Every banner has a real Reject button, equal in weight and visibility to Accept. No buried links, no gray-on-gray second option.

Across every live customer site over the last month:

ConsentStack consent rate benchmarks
MetricValue
Median site, full opt-in rate72%
All decisions pooled69%
Top-performing site89%

The gap between the median site and the top is worth pausing on. Same consent engine, same compliance behavior, very different results. The variable is design: how the banner looks, when it appears, and how much it feels like part of the site versus a legal notice taped over it.

The most repeated claim in this space is that European visitors reject everything. One frequently cited study puts acceptance in Germany and France below 25%.

That is not what we see:

Full opt-in rate by country, ConsentStack customer sites
CountryFull opt-in rate
Ireland84%
Belgium73%
Canada73%
Netherlands72%
France72%
United Kingdom72%
Spain71%
United States68%
Germany63%

Germany, the most privacy-conscious market in the world, lands at 63% full opt-in on ConsentStack sites. That is roughly two and a half times the commonly cited German norm. France lands at 72%, roughly three times the cited norm.

Under GDPR traffic as a whole, ConsentStack sites see 68.6% full opt-in. Under the newer US state laws it is a similar story: 84% in Colorado, 75% in Virginia.

These are opt-in jurisdictions with an equal-weight Reject button on screen. Nobody is being tricked into these numbers.

Why design moves the number this much

Regulators decided years ago what a consent banner must do. They never decided what it must look like, and that is where almost all of the variance lives.

Most banners fail the same way: they look like they were bolted on by the legal department. Default vendor styling, the consent vendor's own logo stamped in the corner (a backlink for them, a foreign object on your site), fonts that match nothing else on the page. Visitors treat foreign objects the way they treat popup ads: reject, dismiss, leave.

The design choices that separate a native-feeling banner from a bolted-on one are boring and repeatable:

  • The banner inherits the site's identity. Same fonts, same colors, same border radius, same voice. And no vendor logo, no powered-by link: nothing on the banner announces a third party. When the consent prompt looks like the site asking, people answer it the way they answer the site. When it looks like a third party interrupting, they refuse the third party.
  • It loads instantly. A banner that pops in two seconds after the page settles is an interruption. One that renders with the page is part of the page.
  • The copy is human, in every language. A cookie notice written in your brand's voice outperforms boilerplate legalese pasted from a template. And that holds across languages: instead of settling for a generic machine translation, ConsentStack lets you refine the wording per language, so your German banner sounds like your brand speaking German, not a translated English notice.
  • Reject is a real option. Counterintuitively, an honest, visible Reject button helps. Visitors who feel cornered look for the escape hatch: close the tab, ignore the banner, block it. Visitors given a fair choice mostly say yes.

That last point matters beyond conversion. Regulators in the EU and several US states are actively fining dark patterns: buried reject links, unequal button styling, nagging re-prompts. The high-pressure banner is now a compliance liability that also converts worse than a fair one.

Benchmarks are a starting point, not a target. But as rough guidance from our data:

  • Above 70% full opt-in: you are doing well. The median ConsentStack site lives here.
  • 50% to 70%: normal, with room to improve. Look at banner timing, styling, and copy before anything else.
  • Below 50%: in an aggregate-study world this is average. In practice it usually means the banner reads as foreign to the site, or it is slow, or the choices feel loaded.

And measure yourself honestly. If your CMP reports an 85% acceptance rate, check the definition. If it counts essential-only choices, partial consent, or notice acknowledgments as acceptance, you do not actually know your number.

Frequently Asked Questions

See what your consent rate could be

ConsentStack banners look and feel native to your site: your fonts, your colors, your voice, rendered instantly. Full compliance, no compromise.