On desktop, a consent banner is a strip of screen that most visitors barely notice. On mobile, the same banner can take a third of the viewport, cover the content the visitor came for, and sit directly under their thumb. Placement stops being a design preference and becomes the whole user experience of your consent flow.

The choice matters for compliance as well as conversion. Regulators object less to where a banner sits than to whether it coerces: blocking content until a choice is made, making reject harder to reach than accept, or covering the very controls the visitor needs. A placement that respects the content and offers equal choices will survive scrutiny that a clever modal will not.

The three placements, honestly assessed

Bottom anchored banners are the mobile default for a reason. They appear where thumbs already are, they leave the page title and first content visible, and they can persist unobtrusively while the visitor reads before deciding. Their weakness is banner blindness: frequent visitors learn to ignore the bottom strip entirely, which is fine for them and not a compliance problem as long as no non essential tracking fires before a choice.

Top banners push content down, which makes them impossible to miss and disruptive in a specific way: they shift the page as they load, moving the content the visitor was about to tap. Layout shift is both an annoyance and a measurable performance failure. Center modals are the most visible and the most resented. They interrupt, they read as a wall, and they invite the visitor to tap whatever makes them disappear, which is usually accept. That is not a choice. It is an exit door with one handle.

What regulators actually object to

Guidance from European regulators and several US state frameworks converges on a few points. Consent must be freely given, so blocking all access until a choice, sometimes called a cookie wall, is risky in many jurisdictions and flatly problematic in some. The reject option must be as easy as accept, in reach, in steps, and in visual weight. A placement that hides reject behind a scroll inside the banner while accept sits fixed fails that test.

Placement interacts with all of it. A modal where accept is a large button and reject is a small link in the corner is a dark pattern regardless of intent. A bottom banner where both options are full buttons, side by side, with a settings option equally reachable, is defensible even if a regulator dislikes banners in principle. The geometry of the choices matters more than the geometry of the container.

Design for the thumb and the exit

Mobile banners should be designed one handed. Both main actions within thumb reach at the bottom of the screen, touch targets at least forty four pixels, and text large enough to read without zooming. If your banner needs a scroll to show the full options on a common phone size, it is too much banner. Shorten the copy, move detail to a settings layer, and keep the first layer to the decision itself.

Respect the visitor who decides nothing. Many people neither accept nor reject immediately; they want to read the article first. A well behaved banner stays out of the content, does not follow the scroll aggressively or reappear on every page, and keeps non essential tracking off until a real choice is made. Visitors who are left alone usually choose eventually. Visitors who are chased choose accept just to end the chase, and everyone knows it, including auditors.

Test placement like any other conversion surface

Placement is testable within compliance limits. Measure decision rate, the split between accept, reject, and settings, time to decision, and how banner presence affects engagement metrics like scroll depth and return visits. A placement that doubles accept rates by making reject hard to find is not a win. It is evidence. Look for placements where overall decision rates rise because the choice is easy, not because one option is.

Test on real devices at real sizes, including small phones and landscape orientation, where banners behave worst. The banner that looks tasteful in a desktop design mock can cover most of a small phone's usable screen. If the team has not seen it on the smallest device your audience uses, the placement decision has not really been made yet.

Frequently asked questions

Where should a consent banner go on mobile?

Bottom anchored is the safest default for most sites. It keeps content visible, places both choices within thumb reach, and avoids the layout shift of a top banner and the interruption of a modal. Whatever placement you choose, accept and reject must be equally easy to reach.

Are modal consent popups a compliance problem?

They can be. Modals that block all content until a choice is made risk being treated as cookie walls, and modals where accept is prominent and reject is hidden are a classic dark pattern. A modal with equal, clear choices and a way to proceed is less risky, but a banner is usually safer and better liked.

How much of the mobile screen can a banner cover?

There is no official percentage, but if the banner needs scrolling to show its options or covers the content entirely, it is too big. Keep the first layer to the decision itself, with detail behind a settings control, and test on the smallest phones your audience uses.

Does banner placement affect consent rates?

Yes. Placement changes how many people decide at all and how they split. Measure the full picture, decision rate, accept and reject split, and return visits, because a placement that raises accepts by hiding reject is creating legal risk, not better consent.