Cookie banner speed on a phone
A cookie banner loads on every page, so its weight matters. We measured what Consentl, CookieYes and Cookiebot add to real websites on a phone, with Google's own tool, and we show the method so that anyone can check it.
Results
| Banner | Data downloaded | Requests | Processor time |
|---|---|---|---|
| Consentl | 19 KB | 3 | under 50 ms |
| CookieYes | 33 KB | 10 | 91 ms |
| Cookiebot | 157 KB | 5 | 425 ms |
Lower is better in every column. Only the banner vendor's own files count, not the rest of the website.
Measured websites
Each value is the median of the runs on that website. The links open the website in PageSpeed Insights.
| Website | Banner | Data downloaded | Requests | Processor time |
|---|---|---|---|---|
| consentl.com (external site) | Consentl | 18 KB | 3 | under 50 ms |
| consentl.com/pricing (external site) | Consentl | 19 KB | 3 | under 50 ms |
| best4u.nl (external site) | CookieYes | 32 KB | 9 | 149 ms |
| latevaweb.com (external site) | CookieYes | 83 KB | 11 | 91 ms |
| codems.com (external site) | CookieYes | 33 KB | 10 | 67 ms |
| acm.org (external site) | Cookiebot | 158 KB | 5 | 641 ms |
| kartenwelt.rewe.de (external site) | Cookiebot | 157 KB | 5 | 294 ms |
| generaxion.com (external site) | Cookiebot | 145 KB | 5 | 425 ms |
How we measured
- Google PageSpeed Insights (Lighthouse 13.5.0) in mobile mode: an emulated mid-range phone on a slow 4G connection with a slowed-down processor. Each website was measured 3 times.
- We count only the requests to the banner vendor's own servers (script, settings, texts, consent logging) and add up the data they transfer.
- Processor time is the main-thread time of the vendor's scripts as reported by Lighthouse, which lists a script only from 50 ms on; a smaller cost is shown as "under 50 ms".
- Consentl is measured on consentl.com, which uses the same banner as our customers; only the banner's requests count, not the website itself.
- CookieYes and Cookiebot are measured on websites of their customers that load their banner, not on their own websites, where every file comes from the vendor's servers.
- We do not compare the page load times of the measured websites: they depend mostly on the rest of each website, not on the banner.
Check it yourself
Results vary a little from run to run and change with each vendor's updates. Open any website from the table in Google's free tool and look at the requests to the banner's servers: PageSpeed Insights (external site)
CookieYes and Cookiebot are trademarks of their respective owners. Consentl is not affiliated with them; we name them only to compare measured facts.