Quick answer: to get value from heatmaps and session recordings, start with one important page and one question, such as "why do visitors leave before they reach the contact form?" Check the scroll map to see how far people get, the click map to see what they try to use, then watch 20 to 30 recordings of visitors who did not convert, filtered by device. Write down every repeated problem, such as rage clicks, dead clicks, missed buttons or confusion, and fix the ones that appear most often first.
What these tools show that analytics cannot
Google Analytics tells you what happened: how many people visited a page and how many converted. Heatmaps and recordings show why: where visitors clicked, what they ignored, where they hesitated and where they gave up.
Free tools like Microsoft Clarity record this automatically once the tracking code is installed. Hotjar and similar tools work the same way. If you have not set one up yet, our GA4 and Microsoft Clarity setup guide covers it in a few minutes.
The three types of heatmap
Scroll maps
A scroll map colors the page by how many visitors reached each point. It answers: do people see the important content?
Look for a sharp drop. If most visitors stop scrolling before your call to action, your reviews or your key information, either move that content up or give people a reason to keep scrolling. A gradual fade is normal; a cliff is a problem.
Click maps
A click map shows where people click or tap. It answers: what do visitors try to use?
Watch for:
- Clicks on things that are not links, such as images, icons or underlined text. Visitors expect them to do something.
- Few clicks on your main call to action. The button may be hard to see, unclear or in the wrong place.
- Lots of clicks on navigation from a landing page. Visitors may be looking for information the page does not give them.
Area or attention maps
These group clicks by section or show where people spend the most time. They help you compare sections: is anyone using the comparison table, or do they skip straight past it?
Session recordings: watch with a filter
Recordings are powerful but slow. Watching random sessions is the fastest way to waste an afternoon. Always filter first:
- One page or one funnel step, such as the checkout or the contact page.
- One device type. Phone and desktop behave very differently, so watch them separately.
- Visitors who did not convert. They show you the problems.
- Optional signals, such as rage clicks, dead clicks, quick backs or JavaScript errors, which Clarity flags automatically.
Watch 20 to 30 recordings per segment. By then, the repeated problems are obvious and new recordings stop teaching you anything new.

The warning signs to note
| Signal | What it looks like | Common cause |
|---|---|---|
| Rage clicks | Several fast clicks on the same spot | Something looks clickable but does nothing, or loads slowly |
| Dead clicks | A click with no response | Images or text styled like links |
| Excessive scrolling | Fast scrolling up and down | Visitors cannot find what they are looking for |
| Quick backs | Visitor leaves and returns almost at once | A link led somewhere unexpected |
| Form hesitation | Long pauses or retyping on one field | Confusing field, unclear format or error message |
| U-turns at checkout | Visitor reaches checkout and goes back | Surprise costs, missing payment options or trust concerns |
Turn observations into fixes
Use a simple log while you watch:
| Page | Device | What happened | How often | Likely fix |
|---|---|---|---|---|
| /contact | Phone | Error on phone field, user gives up | 6 of 25 | Accept any phone format |
| /services/seo | Desktop | Clicks on process icons, nothing happens | 9 of 30 | Link icons to the matching section |
Then prioritize by frequency and by how close the problem is to a sale. A bug in the checkout that affects one in five visitors comes before a confusing icon on the About page.
Once you have a likely fix, decide how to validate it. With enough traffic, run an A/B test. With less, make the change and compare the same metrics before and after. Our guide to A/B testing with low traffic explains both options.
Common mistakes
- Watching without a question. You will see everything and learn nothing.
- Mixing devices. A heatmap that combines phones and desktops shows clicks in places neither actually has.
- Trusting tiny samples. A heatmap based on 40 visits is a hint, not evidence.
- Ignoring page changes. If the page changed mid-period, the map blends two versions.
- Forgetting privacy. Make sure the tool masks form inputs and personal data, and that your cookie consent and privacy policy cover it.
A monthly routine
- Pick your two most important pages, usually a key landing page and the contact or checkout page.
- Check scroll and click maps for phone and desktop.
- Watch 20 filtered recordings of non-converting visitors per page.
- Log repeated problems and add the top three to your fix list.
- Next month, check whether the fixes changed behavior.
Frequently asked questions
Is Microsoft Clarity really free?
Yes. Clarity is free with no traffic limits, and includes heatmaps, recordings and automatic detection of rage and dead clicks.
How much traffic do I need for heatmaps?
A few hundred visits per page and device gives a reasonable picture. Below that, recordings are more useful than heatmaps, because each one shows a complete visit.
Are session recordings legal?
They are widely used, but you must follow privacy law where you and your visitors are. Mask personal data, mention the tool in your privacy policy and respect consent rules where they apply.
What should I fix first?
The problem that appears most often on the page closest to a conversion: checkout, booking or contact form issues usually come first.
Want someone to watch the recordings and hand you a prioritized fix list? Our conversion rate optimization service starts with exactly this kind of research, or you can start with a free audit.