On phones, the consent banner cannot sit politely in a corner. There is no corner. The bottom sheet, a panel sliding up from the base of the screen, has become the default answer, used by major platforms and copied by everyone else.
Like any default, it is done well and badly in roughly equal measure. The difference is measurable in the same terms as any interface: how much of the page it steals, how fast a visitor can decide, and whether the decision sticks.
Size is the whole design
A good mobile sheet occupies a quarter to a third of the viewport: room for a short explanation and two clear actions, small enough that the page behind it still reads as a real website the visitor chose to visit.
Sheets that swell past half the screen read as a takeover and trigger the same reflex as a bad pop up. If your text needs that much room, the text is the problem. Consent explanations should be short, with detail behind a settings or learn more option.
Equal actions, honest layout
Reject and accept must sit side by side, same size, same visual weight. Designs that make accept a bright button and reject a faint text link are the mobile version of a dark pattern, and regulators increasingly describe them in exactly those terms.
A settings option for granular choices belongs on the same layer, not buried under two taps. Visitors who want control and cannot find it quickly either accept angrily or leave, and both outcomes are losses.
Behavior after the choice
Once chosen, the sheet must go away and stay away: across pages, across sessions, for the period your policy states. Reappearing sheets are the single most common consent complaint on mobile, and they usually trace to records stored in a way the browser or the app webview keeps clearing.
Test inside the environments your visitors actually use, including in app browsers from social platforms. A banner that behaves on your office phone and loops endlessly inside an app webview is broken where it counts.
Timing and motion
Let the page render and settle before the sheet rises, a beat of about a second. A banner that arrives while the visitor is already reaching for a link turns taps into accidents, and motion should respect reduced motion settings by simply appearing.
Also plan for the keyboard moment: if your banner includes any input or a settings search, the sheet must adapt when the keyboard opens rather than hiding its own buttons beneath it.
Related reading
These earlier guides go deeper on connected questions: Consent in Single Page Apps: Tracking That Starts Before the Banner Loads Single page apps load trackers with the bundle, often before any consent interface appears. Sequencing consent in an SPA takes deliberate architecture, not a script tag., Consent Banner Frequency: How Often Is Too Often to Ask Re-asking for consent too often trains visitors to click anything to make the banner go away. Setting a sensible re-ask rhythm protects both data quality and trust., and Exporting Consent Records: What an Auditor Actually Wants to See When a regulator or partner asks for consent evidence, a raw data dump will not do. How to export consent records that answer audit questions..
Frequently asked questions
How much of the screen should a mobile consent banner cover?
About a quarter to a third of the viewport is the working range. Enough for a short explanation and clear equal actions, small enough that the underlying page remains visible and the banner does not feel like a takeover.
Is a bottom sheet better than a full screen consent wall?
For most sites, yes. A sheet lets the visitor see the site exists and decide in context, while a wall blocks everything until a choice is made. Walls can raise consent rates in the short term and raise complaints, exits, and regulatory attention with them.
Why does the banner keep coming back on my phone?
Usually because the consent record is not being stored or read correctly: cleared cookies, storage limits in the browser, or an in app webview with separate storage. Test the exact browser or app where it happens, since it may work fine elsewhere.
Should the reject button look the same as accept?
Yes. Equal size and prominence for accept and reject is both good practice and, increasingly, an explicit regulatory expectation. Making reject harder to find or weaker visually is treated as a manipulative design.