Analyze heatmaps and session recordings to improve a web page

Help me turn heatmap and session-recording data into specific improvements for a web page. Page and its goal: [URL or description; e.g. "pricing page: get free trial sign-ups"] Traffic context: [visitors in the period, main sources, mobile vs. desktop split] Heatmap findings: [describe or paste: click maps, scroll depth (% reaching each section), move/attention maps, rage clicks, dead clicks] Session recording notes: [patterns from watched sessions: hesitations, back-and-forth, form errors, quick exits] Conversion data: [rate and any funnel step drop-offs] Analyze: 1. Separate devices: compare mobile and desktop; never combine them in one heatmap reading. 2. Attention: where most visitors stop scrolling and whether the key message and call to action appear before that point. 3. Clicks: clicks on non-clickable elements (expectation mismatches), rage clicks (broken or slow elements), and ignored calls to action. 4. Friction: form fields or steps where recordings show hesitation or errors. 5. Sample size: whether there's enough data to trust each finding, and which findings are only hypotheses. Return a table: observation, likely cause, recommended change, expected impact, confidence (high/medium/low). Then the top 3 changes to try first, written as A/B test hypotheses ("If we ___, then ___, because ___"), and what to measure. Remind me to mask personal data in session recordings and respect consent settings.