Banner study 2026: cookie banners on 366 Slovak and Czech sites
We measured what cookie banners look like and what they do on 366 Slovak and Czech sites from the Tranco list (4 Nov 2026). On 39% of them analytics or ad cookies are set before the visitor decides (a technical indicator, not a legal assessment), and 25% show Reject as prominently as Accept.
Norbert Kovalčín · Conseto, technical measurement,
- 39%
- of sites with a banner set analytics or ad cookies before the visitor decides (144 of 366).Tranco 746WX, N = 366 sites with a banner (.sk and .cz), captured from Slovakia on 4 Nov 2026
- 68%
- of sites have a free Reject button on the first layer of the banner (250 of 366).Tranco 746WX, N = 366 sites with a banner (.sk and .cz), captured from Slovakia on 4 Nov 2026
- 25%
- of sites show Reject as prominently as Accept (90 of 366). 43% show it less prominently.Tranco 746WX, N = 366 sites with a banner (.sk and .cz), captured from Slovakia on 4 Nov 2026
- 29%
- of the desktop screen is covered by the banner, median.Tranco 746WX, N = 330 sites with a banner and a measured desktop area, 4 Nov 2026
What we measured
From the Tranco list we took the 260 top-ranked .sk domains and 260 .cz domains, 520 sites in total. 441 loaded and 366 of them (83%) have a consent banner. Every share in this article is out of those 366 sites unless stated otherwise.
We opened each site on desktop and on mobile in a clean browser context, with no cookies and no clicks. We measured what a visitor sees after the page loads: which buttons the banner has, how large and how prominent they are, how much of the screen it covers and which cookies were set before any decision.
Cookies before the decision
After loading, without a single click, 144 of 366 sites with a banner (39%) set analytics or ad cookies. The gap between Slovak and Czech sites is small: 39% on .sk and 40% on .cz.
The most common are Google Analytics (the _ga cookie, 74 sites) and the Meta pixel (the _fbp cookie, 50 sites), followed by Gemius, Criteo and Google ads. This is a technical indicator, not a legal assessment: the measurement cannot see the legal basis or a site's configuration.
| Domain | Sites with a banner | Cookies before the decision |
|---|---|---|
| .sk | 184 | 72 (39%) |
| .cz | 182 | 72 (40%) |
| Total | 366 | 144 (39%) |
Reject: is it there and how does it look
A free Reject button directly on the first layer of the banner appears on 250 of 366 sites (68%). On 88 sites (24%) it is missing, and on 12 Czech sites rejecting is offered only with a subscription. The gap between domains is large: 80% on .sk, 57% on .cz.
The second question is how Reject looks. It is as prominent as Accept on 90 sites (25%) and weaker on 158 (43%): a link, an outline button or a smaller button.
For comparison we hand-picked 50 sites (a hand-picked set of 50 sites, not a market sample). Reject is equally prominent on 14 of 16 banner vendors' own sites, 11 of 19 large EU sites and 4 of 15 Slovak and Czech ones. It shows what tool vendors do, not what the market looks like.
| Domain | Sites with a banner | Free Reject | As prominent as Accept |
|---|---|---|---|
| .sk | 184 | 147 (80%) | 53 (29%) |
| .cz | 182 | 103 (57%) | 37 (20%) |
| Total | 366 | 250 (68%) | 90 (25%) |
What banners look like
On desktop the most common type is a centered window (162 of 366, 44%), and 129 of those dim the rest of the page. On mobile a bar at the bottom leads (140, 38%), and 57 sites show a full-screen banner.
The typical banner is not small. On desktop it covers a median of 29% of the screen (N = 330), on mobile 63% (N = 331). The text has a median of 76.5 words (N = 348). 200 sites show three buttons on the first layer, 126 show two.
| Type | Desktop | Mobile |
|---|---|---|
| Centered window | 162 (44%) | 129 |
| Bar at the bottom | 103 | 140 (38%) |
| Window at the bottom | 39 | 18 |
| Full screen | 3 | 57 |
Where the Conseto banner stands
We measured the Conseto banner with the same script, on the Demo pražiareň sample site (4 Nov 2026). It is the banner that runs in production today. The measurements show a small area, short text and Reject as prominent as Accept. The measurement found no dark pattern.
| Measurement | Conseto | Wide sample (N = 366) |
|---|---|---|
| Screen area, desktop | 15% | median 29%; 15% or less on 61 of 330 |
| Screen area, mobile | 38% | median 63%; 38% or less on 63 of 331 |
| Words of text | 19 | median 76.5; 19 or fewer on 34 of 348 |
| Reject as prominent as Accept | yes | yes on 90 of 366 (25%) |
What this means for a site owner
Four things you can check in a few minutes. This is not a legal assessment, it is about what happens on your site.
- Open your site in a private window and, without clicking the banner, look at the cookies in your browser's developer tools. If _ga or _fbp are already there, your site is among the 39% in our sample.
- Look at how Reject compares to Accept: colour, size, fill. Only a quarter of the sites in the sample make them equally prominent.
- Open the site on mobile too. The banner covers a median of 63% of the screen there, and one tap may be all that separates it from your content.
- Conseto's free check does the first two steps for you and shows what was set before consent.
The Conseto banner: what we learned
When we measured it, the Conseto banner used its own system font and looked like a foreign element on a site with another font. That is why, since version 2.4.0, the banner takes the font, surface, text colour and corner radius of the site it runs on.
Method
Sample: Tranco list 746WX (created 3 Nov 2026). The top 260 .sk domains and top 260 .cz domains were selected, and only 2 domains with adult content were excluded. Of the 520 sites, 441 loaded, 24 blocked the automated browser (the protection was not bypassed) and 55 did not load. 366 of the loaded sites have a banner.
Capture: 4 Nov 2026 from Slovakia, each site's home page once on desktop (1280 × 800) and once on mobile (375 × 812), always in a fresh clean browser context. Nothing was clicked and the second layer was not captured.
Evaluation: banner position, size and text were measured from the DOM. A model then reviewed every screenshot and decided whether a banner is visible. Cookies set after loading were recorded by name and domain (never values) and classified by cookie name and ad network domain. Technical cookies (bot protection, load balancing, stored consent) are not counted.
Limits
- The sample is the top-ranked .sk and .cz domains from the Tranco list that show a banner on the home page, not the whole market. It describes 366 measured sites.
- The hand-picked set of 50 sites (banner vendors, large EU sites, Slovak and Czech sites) is not a random sample and its shares cannot be read as market shares.
- DOM measurement cannot see into a closed shadow DOM or into some frames. A banner that loads late or only after mouse movement may be missed. Whatever could not be measured is marked as not measured in the data.
- Some sites block an automated browser and some do not show it the banner. A real visitor may see a different banner.
- The screenshot review is done by a model, not a person. Cookies before the decision are a technical indicator: the measurement cannot see the legal basis or a site's configuration, so it says nothing about whether the rules are broken.
How to cite
On 366 Slovak and Czech sites with a banner from the Tranco list (4 Nov 2026), 25% show Reject as prominently as Accept.
On 366 Slovak and Czech sites with a banner from the Tranco list (4 Nov 2026), 39% set analytics or ad cookies before the visitor decides (a technical indicator, not a legal assessment).
Find out what your site does before consent