Session replay and heatmap tools
What do session replay and heatmap tools do?
Two working definitions:
- Heatmap. A picture of one page that shows where visitors click, tap or scroll.
- Session replay. A recording of what one visitor did on the site, replayed afterwards.
Both are qualitative. They show what happened in some visits and leave you to work out why. A CRO-software vendor's audit guide, for example, recommends heatmaps and session replays for studying behaviour1. That tells you how the tools are sold. It does not tell you how well they work.
Why do ecommerce brands use them?
A conversion rate tells you that visitors left. It does not tell you where they got stuck. These tools put a face on the number.
Typical questions they help with:
- Does anyone see the delivery information on the product page?
- Do people tap something that is not a button?
- Where does the mobile checkout lose people?
- Does the landing page from an ad show what the ad promised?
The answers are leads. They need checking.
How do we implement it?
Check analytics first
Start with the numbers, then write the question you want recordings to answer. Without a question you collect footage and no decisions.
Agree masking and consent
Decide with your counsel what may be recorded. Mask form fields before you start.
Record chosen pages
Pick which pages to record and what to hide. Then someone watches and writes down what they see.
Treat findings as hypotheses
Judge any change on orders and margin, in a controlled test, not on how the heatmap looks afterwards.
When should you not use them?
Before running any tests, our conversion rate optimisation audit piece covers what to check.
What does it cost, in effort?
We do not give prices, and we do not rank vendors. In effort terms:
- Installing a script. Usually quick, but it adds another script to your site.
- Setting it up properly. Deciding which pages to record and what to hide.
- Watching. This is the real cost. Someone has to sit through recordings and write down what they see.
- Consent. Recording starts according to your consent set-up. Your counsel decides the rules.
What are the pitfalls?
Recording what people type
Check whether the tool can capture what visitors enter into forms. Mask those fields and agree the rules with your counsel before you start.
Mistaking a story for a finding
One striking recording is a hypothesis, not a result.
Using them to measure
To see whether a change worked, you need a controlled test, not a lower click count on a heatmap.
Watching with no question
Without a question, you collect hours of footage and no decisions.
Our take: Watching
This is our view, not a market fact. We rate these tools Watching.
Why:
- Qualitative research and quantitative research answer different questions. Nielsen Norman Group, writing about usability studies, describes qualitative work as finding problems and quantitative work as measuring how users perform across a wider population2. A few recordings can suggest a problem. They cannot tell you how common it is.
- We found little independent evidence. The material on these tools we found in our research was agency and vendor content, which sells the tools rather than tests them.
- Privacy is a real constraint. We do not give a general rule on what may be recorded. Your counsel decides what you may record and when.
Watching means: we are not against them. We would not put them in a plan as a core method, and we would not treat what they show as proof.
For ZFD: In a CRO audit we treat recordings and heatmaps as a source of hypotheses. A change is then judged on orders and margin, not on how the heatmap looks afterwards. Our CRO service covers audits and implementation, with no promised result.
FAQ
What is a heatmap?
It is a picture of a page that shows where visitors click, tap or scroll. It shows what they did, not why.
What is session replay?
It is a recording of one visitor's session on your site, played back afterwards.
Can session replay prove that a change worked?
No. It is qualitative. Nielsen Norman Group describes qualitative studies as a way to find problems, and quantitative studies as the way to measure them2. A controlled test does the measuring.
Do I need consent to record sessions?
Your counsel decides, and your consent set-up applies the decision. We do not give a general answer here.
Do these tools replace A/B tests?
No. They suggest what to test. A controlled test shows whether the change helped.
Let’s find what’s actually working.
Next step. Want to check whether recordings would help with your site? Talk to us about your setup. For the wider picture, see how to improve ecommerce conversion rate.