What a cookie banner costs you

Consent banners are usually discussed as a legal question. On the six sites where I have been able to measure before and after, they are also an expensive design decision, and the cost is larger than teams expect.

The clearest effect is on the first scroll. A full-width banner covering the bottom third of a phone screen means the visitor's first interaction with the site is dismissing something. Median time to first meaningful scroll went from 1.9 seconds to 4.4 across the six sites. That is nearly three seconds of a visit that typically lasts under thirty.

The number nobody quotes is how many people click “accept all” purely to make it go away, and then leave. On two of the sites we could see this directly: about 11 per cent of sessions consisted of accepting the banner and closing the tab. Those visitors are counted as consenting, which flatters the analytics, and they read nothing at all.

Measured on six client sites: bounce, scroll depth, and the number people never mention.

Position matters more than wording. Moving the banner from a bottom sheet covering a third of the screen to a compact bar roughly 88 pixels tall recovered most of the scroll-depth loss on two sites — the same text, the same buttons, a smaller footprint.

The genuinely effective fix is upstream: use fewer things that need consent. A site running only first-party analytics with no cross-site identifiers has less to ask about, and a smaller banner with a single clear choice. That is a conversation with whoever chose the tracking stack, not with the designer, which is why it rarely happens.