Skip to content

This translation has not been editorially reviewed yet. The German version is authoritative. Deutsche Fassung →

Data protection

Does clicking "Reject" in the cookie banner actually work?

Only a practical test can show that — and that's exactly what our check runs: it actually clicks the reject button, reloads the page, and measures whether tracking cookies are still set or tracking servers still contacted afterward. A banner whose rejection has no effect is worse than none at all, since it promises visitors a choice it doesn't honour.

The functional reject test, step by step

The checker opens your website in a fresh browser, finds the reject button on the first banner level, and clicks it — just as a real visitor would. Then it reloads the page and measures two things: what cookies are now in the browser, and what third-party servers get contacted. No tracking cookies should be created after a rejection, and no connections to tracking servers either. This acid test tells a functioning banner apart from a facade — settings on paper don't count, behaviour is what's measured.

Three possible outcomes of the measurement

The best result: after rejection, everything's quiet — no tracking cookies, no tracker contacts, full marks. The middle result: no more cookies are created, but the page still contacts tracking servers with cookieless signals, as Google Consent Mode provides for. The IP address is still transmitted then; the check awards half marks and recommends a manual review, since the legal classification is disputed. The worst result: tracking cookies are set despite rejection — the visitor's objection is technically ignored, and the rule counts as clearly failed.

An ineffective banner is worse than none at all

A website showing no banner at all promises nothing. One that shows a banner gives visitors its word: your decision counts. If the website still sets tracking cookies after a click on "Reject", it breaks exactly that word — and the visitor has no way of noticing. A common technical cause is a consent tool that stores the answer but isn't actually connected to the tracking scripts: the banner manages consent, while the scripts load regardless. The audit report makes this state visible so it can be fixed.

  • Open your website in a private window, click "Reject", and check the browser settings for what cookies exist afterward.
  • Reload the page after rejecting and watch the network tab (F12 key) for whether tracking servers are still contacted.
  • Ask your agency to technically link every tracking script to the consent tool — the stored answer alone blocks nothing.
  • Make a deliberate decision on whether cookieless signals after rejection (Consent Mode) are acceptable for you, and get legal advice on it.
  • Repeat the test after every change to the banner or the integrated services.

FAQ

What is Google Consent Mode?

A mode in which Google services stop setting cookies after a rejection, but keep sending cookieless signals to Google servers. The visitor's IP address is still transmitted in the process. Our check recognises this pattern, awards half marks, and recommends a manual review — the legal classification is disputed and belongs with your legal counsel.

How can I test myself whether my cookie banner actually implements rejection?

Open the website in a private browser window, click "Reject", and reload the page. Then check the browser settings for stored cookies and the network tab (F12 key) for contacted servers. If names like "_ga" or connections to analytics servers show up there, the rejection isn't working.

My service provider says the banner is set up correctly — is that enough?

A setting is an intention, not an effect. The banner and the tracking scripts are separate building blocks; whether they actually work together only shows up in a measurement in the browser. That's why our check actually clicks "Reject" and measures afterward. The result is in the report — as confirmation for your service provider, or as a concrete task to fix.

What happens if the check can't click the reject button?

Then the report notes exactly that: a reject button was found but couldn't be triggered automatically — the effect needs manual review. If the first banner level offers no rejection at all, the effectiveness test doesn't happen either, and both banner rules count as unmet.

Deutsche Fassung dieser Seite