Analytics without a consent banner

A consent banner is the only part of your site that every single visitor sees before anything else, and the only part designed to be dismissed. It costs you conversion, it costs you goodwill, and it costs you a share of your measurement data that almost nobody has quantified.

That last cost is the interesting one, because it is measurable in an afternoon and it usually turns out to be larger than people expect.

Measuring what the banner costs you

Run a cookieless tool alongside your existing consented setup for a month and compare the totals. The cookieless tool needs no prompt, so it counts everyone; the consented tool counts only those who accepted, minus those blocked before the prompt appeared.

The difference is your measurement gap. It is not an estimate or a model, it is two numbers from two tools on the same site over the same period. Simple Analytics has built an entire enterprise product on selling this comparison to large organisations, and its published claim is that the recoverable share runs from 20 to 60 per cent. That is a vendor figure rather than an independent one, but the direction is not controversial.

Do this before you change anything. Whatever you buy afterwards, knowing the size of the gap changes how you read every report you have looked at for the last two years.

When you can actually remove it

The consent requirement attaches to storing or accessing information on a visitor's device, and to processing personal data. An analytics tool that writes nothing to the device and retains no identifier does neither, which is why the genuinely cookieless tools do not need a prompt for analytics purposes.

The catch is everything else on the page. Embedded video, chat widgets, advertising pixels, A/B testing tools, font loaders that phone home and social embeds all carry their own obligations. Removing the banner means auditing the whole page, not just swapping the analytics.

This is also why adding a free heatmap tool to a cookieless setup is a trap. Microsoft Clarity and Hotjar both set cookies, so bolting either onto a cookieless stack puts the banner straight back and undoes the work.

The banner-free stack

In practice a site that wants no banner needs four decisions to go the same way. Analytics must be cookieless. Behaviour tooling, if you want it, must come from the same cookieless tool rather than a separate cookie-setting one. Fonts must be self-hosted rather than loaded from a third-party CDN. And embeds must be either removed or wrapped in a click-to-load placeholder.

The second of those is where most attempts fall down, because until recently the cookieless category simply did not offer heatmaps. Plausible, Fathom, Pirsch and Simple Analytics still do not, at any tier. That leaves you choosing between seeing where people click and not having a banner, which is a choice nobody should have to make.

If you must keep one, at least make it honest

Some sites genuinely cannot remove the banner, usually because advertising or personalisation depends on it. If that is you, two things are worth doing.

Make reject as easy as accept, on the same screen and in the same visual weight. Dark patterns here are both an enforcement risk and a conversion problem, because a visitor who feels tricked at the doorway reads the rest of the site differently.

And run a cookieless tool underneath the banner anyway, purely as a measurement baseline. It tells you what your consented analytics is missing, which is the only way to interpret the numbers you do have.

The rest of the field

Common questions

Related guides