Interpreting heatmaps & refreshing screenshots

What to look for

• High clicks on non-links — visitors expect something to be clickable. Consider making it so, or removing the false affordance. The dead click overlay highlights these.
• Rage clicks — repeated frantic clicks signal frustration: a broken control, slow response, or a misleading element. Follow them into Rage clicks for element-level detail and the triage workflow.
• Important content below the scroll fold — if the scroll map shows few people reach your CTA, move it up.
• Ignored CTAs — low clicks on a primary button suggests a visibility or copy problem — a good A/B test candidate.
• A hotspot that appears on only one device — usually a layout bug rather than a behavioural difference. Switch the device filter to confirm.

Check the sample size first

A heatmap built from forty sessions is a picture of forty people, and it will look convincingly like a pattern. Check the interaction count before drawing conclusions — and be especially careful after filtering, since narrowing to Mobile + Safari + one variation can leave very little data behind a very confident-looking gradient.

Refreshing the screenshot

Heatmaps overlay data on a captured screenshot of your page. If you redesign the page, click Refresh Screenshot to capture the current version. Capture runs in the background and can take a couple of minutes; the viewer updates when it's ready.

If the screenshot comes back with a cookie banner or chat widget covering the page, add a rule under Screenshot exclusions and capture again.

After a redesign, the older interaction data was recorded against the old layout. Refreshing the screenshot does not re-position it — treat a significant redesign as a fresh start and read the heatmap from that date forward.

From heatmap to replay

A heatmap shows what happens in aggregate; session replay shows why for an individual. When a hotspot puzzles you, watch two or three sessions that produced it — it's almost always faster than theorising.