We scanned 400 Danish webshops. 4 in 10 track visitors before consent
A cookie banner is a promise: nothing tracks you until you say yes. We wanted to know how often that promise holds, so we checked 400 Danish webshops.
How we tested
We started from a public research ranking of the million most-visited websites, took every Danish .dk site on it that works as a webshop, and picked 400 of them at random. That gives a mix of big chains and small specialist shops, across every kind of product.
On 20 November 2026 we ran each homepage through our free cookie scanner. A fresh browser with an empty profile loads the page, clicks nothing, scrolls once, waits, and writes down everything that happened before the visitor had a chance to answer the banner. 396 of the 400 shops loaded and are counted.
With that many shops, the percentages below are accurate to within about 5 percentage points either way.
What we found
42% of the shops set tracking cookies before consent. That is roughly 4 in 10.
78% of those showed a consent banner while doing it. The banner was on the screen, waiting for an answer, and the tracking had already started behind it.
The cookies fell into four groups:
- Advertising cookies on 28% of the shops. These tie the visit to an ad platform, so the visitor can be followed and targeted later.
- Analytics cookies on 18%. A persistent ID that recognises the visitor on their next visit.
- Email marketing cookies on 14%. An ID that can later be matched to a newsletter subscriber.
- Session recording cookies on 2%. A recording tool that captures how the visitor moves, scrolls and clicks.
The shops that set no tracking cookies were not all spotless either. 71% of them still contacted known tracking services before consent. Some tools now load without setting cookies until the visitor agrees, which is a real improvement, but the visitor's browser still calls the tracking service, and that call carries the visitor's IP address.
The typical shop contacted 17 outside domains before consent. The busiest contacted 64.
Why this happens
None of this looks deliberate, and it is not down to one bad consent tool. The shops used more than 40 different ones, and every tool used by five or more shops had at least one of them tracking before consent. The usual causes are small and easy to miss:
The banner is installed, but the tags were never told to wait for it. A consent banner only records the visitor's answer. Each tracking tag still has to be set up to check that answer before it runs, and one tag that skips the check is enough.
An app or plugin adds its own tracking. Webshops install apps for reviews, newsletters, upsells and chat. Many of them bring their own tracking script, and it often bypasses the banner because it was never connected to it.
Your own browser hides the problem. You accepted your own banner months ago, so every time you visit your shop, you see the version for someone who has already said yes. The first-time visitor's version is the one the rules are about, and it is the one you never see.
Check your own shop
Scan your site. It is free, needs no sign-up and takes under a minute. The report lists every cookie, every outside domain and every storage key from before consent, so you can hand it to whoever runs your tags.
If you want to check by hand, open your shop in a private window, don't touch the banner, and look at the cookies in your browser's developer tools.
The simpler fix
Most of the effort above goes into making tracking wait for permission. The alternative is to measure without anything that needs permission.
Galytics gives you visitors, traffic sources, conversions and session replay without cookies, so there is nothing for the banner to block and nothing that can slip past it. You can try it free for 30 days.
Website analytics and session replay in one cookieless script.
Try Galytics free for 30 days