What cookie banners really do, in real scans

Real scans from our study of 168 online shops, with the names removed. Three banners that looked fine and weren’t, and one that works, on the same consent tool.

Real scans from the October 2026 study. Each was checked by hand against its own evidence. Shops aren’t named without permission. The full numbers are in the study, and you can open a complete sample report.

Questions people ask

Are these real shops?

Yes. Each case is a real scan from our October 2026 study of 168 European shops. We checked every finding by hand against its own evidence: the request, the cookies and the screenshot.

Why are the shops not named?

A scan is one visit on one date, and shops fix things. We show the pattern so you can check your own shop, not a list of offenders. We name a shop only with its permission.

Can my shop have the same problem with a big consent tool?

Yes. Two of the cases use OneTrust: one failed and one got grade A. The tool matters less than whether every tag waits for the visitor’s choice.

What does your own banner do?

See every tracker before a choice, after Reject and after Accept. Free, about a minute.