Cookie Consent Banner and Consent Mode: Setup and Data Loss
How to set up a cookie consent banner, pass the visitor's choice to Google tags with Consent Mode, and what drop in measurement to expect afterwards.
rabbitclip teamPublished: 7 min read
Short answer
A cookie consent banner is the interface that tells a visitor which cookies and similar technologies a site uses, and lets them make a choice. Consent Mode is a signalling system that passes the result of that choice to Google tags: if the visitor has refused analytics or advertising storage, the tags behave accordingly. In short, the banner asks the question and Consent Mode carries the answer to your measurement tools. Either one without the other is half a job.
The commonest mistake in practice is to build the banner as a purely visual element. It appears on screen, but the tags never hear the visitor's choice and fire regardless. The opposite also happens: the banner and the signal are set up correctly, but nobody explained the expected drop in measurement, so the team reads the fall in its reports as a fault.
This article covers the practice: how to set up the banner, how to pass the choice to your measurement tools and what to expect in your numbers. Legal obligations depend on what your site does and where it operates, so please take advice from a lawyer on your own situation.
What does a cookie banner do, and what does it not do?
A banner does three things. It tells the visitor which kinds of cookies are in use, it offers an accept or reject choice, and it stores the answer as a record. The banner itself does not stop any cookies. For cookies to really stop, the tags and scripts on the site have to be wired so that they run according to the answer the banner has recorded.
This gap is the part of the setup that is most often skipped. On a site with a banner where the analytics and advertising tags keep running after the visitor has pressed reject, the banner is decoration and nothing more. The first thing to test when the setup is finished is exactly this: what runs when I reject?
How should the banner be designed, and what should you watch for?
A good banner is short, clear and gives the choices equal weight. The accept and reject buttons should sit on the same first screen with similar prominence; burying the reject option several clicks deep misleads the visitor and can be risky from a compliance point of view. The categories, such as necessary, analytics, advertising and preferences, are explained in plain language.
The banner also needs a way back. Visitors must be able to change their choice later, so leave a permanent cookie preferences link in the footer. For managing consent you can use a ready-made consent management platform. When choosing one, judge it by whether it works with Google's Consent Mode, which languages it supports and whether it keeps a record of choices.
- Accept and reject buttons on the same screen with similar prominence
- Categories explained in plain, simple language
- A permanent link for changing the choice later
- A record and date of each choice is kept
- The banner does not cover the whole content on mobile
How is the choice passed to measurement tools?
Consent Mode works with four main signals: ad_storage, analytics_storage, ad_user_data and ad_personalization. Each represents one type of permission and takes the value granted or denied. In a correct setup, default values for these signals are set as the page loads; when the visitor makes a choice, the banner sends a command that updates the values.
The order matters. The default must be set before any tag runs, otherwise tags may behave in the first moments as if permission had been given. If you use a tag management tool, make sure the tag that sets the default signal fires before all the others. Test the setup in a browser developer tool or in the tag preview mode, with both an accept and a reject scenario.
There are two ways of working. In one, tags do not load at all until permission arrives. In the other, tags load but, without permission, send limited and anonymous signals without using cookies. The second approach can leave fewer gaps in measurement; which one suits you is something to decide together with legal advice.
How much loss in measurement should you expect?
Expect some loss, but how much depends on your audience, the design of the banner and what visitors choose, so no general figure can honestly be given. The behaviour of visitors who refuse cannot be measured directly. After the banner goes live, the sessions, conversions and advertising-attributed numbers in your reports may therefore look lower than in the period before.
Where certain conditions are met, Google can use statistical modelling to fill part of the gap left by refusing visitors; the thresholds and conditions are set out in the product's help documentation and may change over time. Modelling reduces the gap but does not close it. Explain this to the team early: figures after the banner cannot be compared one for one with those before.
What mistakes are common when setting up cookie consent?
The first mistake is a banner that is visible while the tags run regardless of consent. The second is setting the default signal too late. The third is a banner that appears only on the home page or only in one language; a visitor who lands on another page may never see the choice at all.
The fourth mistake is not testing after setup. Accept, reject and partial-choice scenarios should each be tried separately, and the moment cookies are written should be watched in the browser. The free cookie and tracker check on our site can be a starting point for seeing what your site runs in the reject scenario.
- A banner is present, but tags are not tied to the answer
- The default signal is set after the tags
- The banner appears only on some pages or in some languages
- Reject and partial-choice scenarios are not tested
- A fall in measurement is read as a fault and wrongly corrected
When is a cookie banner or Consent Mode unnecessary, or the wrong choice?
If your site uses only technologies that are strictly necessary for it to work, meaning no analytics, advertising or third-party tracking, a consent banner may be an unnecessary obstacle. Which technologies count as strictly necessary is a legal question; please take advice from a lawyer on your own situation.
Consent Mode gains you nothing on a site that does not use Google tags. For a small site with no advertising account and only basic analytics, a ready-made consent platform and a simple setup are enough, and building an elaborate signal architecture would be out of proportion. First take an inventory of what actually runs on your site, then choose the tool.
Cookie consent has two parts: a banner that gives the visitor an honest choice, and a correctly ordered signal that carries that choice to the tags. The real weight of the work lies in expecting some loss in measurement, explaining it to the team beforehand and testing the setup under every scenario. If you would like to review together what runs on your site today, we can arrange a discovery call with rabbitclip.
FAQ
Does Consent Mode replace a cookie banner?
No. Consent Mode is not a banner; it is a signalling system that carries the banner's answer to Google tags. You still need a separate interface that offers the visitor a choice.
If a visitor refuses, is no data collected at all?
It depends on the setup. Where tags never load, cookie-based measurement stops; in the other approach, limited and anonymous signals may still be sent. Decide which suits you together with a lawyer.
Why did my numbers fall after the consent banner went live?
Refusing visitors cannot be measured directly, so session and conversion counts can look lower. This is an expected result rather than a fault; make comparisons from the date the banner went live onwards.
Should I use a ready-made consent platform or build one by hand?
For most businesses a ready-made platform is safer, since choice records, language options and Consent Mode support come built in. A hand-built setup only makes sense when there is a very specific need.
