Research Beyond What Users Say: Learn to Observe What Users Actually Do
Interviews tell you what people say. Analytics show you what they do.
We often think of user research as interviews, surveys, and usability testing. These methods are incredibly valuable—but they capture only part of the story.
A user might tell you that a feature is easy to use, yet repeatedly hesitate before clicking. They might say they love a product, but abandon a key flow halfway through.
The gap between what people say and what they do is where some of the most valuable UX insights live.
From Findings to Behaviour
Modern products generate enormous amounts of behavioural data:
Where users click
How far they scroll
Where they drop off
Which screens they revisit
Where they hesitate
Which features they ignore
What paths they take through a product
Tools such as Hotjar can help visualise some of this behaviour through heatmaps, session recordings, and feedback.
But the tool itself isn't the insight.
The real skill is knowing what to look for and what questions to ask.
A Heatmap Doesn't Tell You Why
Imagine an e-commerce checkout where 60% of users stop at the payment screen.
The data tells you where something is happening.
It doesn't automatically tell you why.
Is the payment form confusing?
Is the shipping cost unexpected?
Is the button difficult to find?
Are users concerned about security?
Are they simply not ready to purchase?
This is where behavioural data needs to work alongside human-centred research.
Quantitative data can reveal the pattern. Qualitative research can help explain it.
The Research Stack
Instead of choosing one research method, think of research as a system.
01 — Ask
Use interviews and surveys to understand what users say, think, and feel.
02 — Observe
Use usability testing and session recordings to see what users actually do.
03 — Analyse
Use analytics and heatmaps to identify patterns at scale.
04 — Map
Use tools such as Affinity Mapping, Empathy Mapping, and Journey Mapping to organise what you've discovered.
05 — Decide
Turn patterns into insights, opportunities, and design decisions.
The goal isn't to collect more research.
The goal is to make better decisions.
What Behaviour Can Reveal
A useful research question isn't simply:
“What happened?”
Instead, ask:
“What might this behaviour be telling us?”
For example:
Observation: Users repeatedly click an element that isn't interactive.
Possible insight: The visual design creates an expectation of interaction.
Design opportunity: Make the element interactive—or change its visual language.
Another example:
Observation: Users scroll back and forth between two sections.
Possible insight: They may be comparing information.
Design opportunity: Make comparison easier by bringing the information together.
This is the difference between data and insight.
But Don't Let the Tool Think for You
One of the biggest mistakes in UX research is treating analytics as the answer.
A heatmap can show where users clicked.
It cannot tell you what they were thinking.
A session recording can show hesitation.
It cannot reliably tell you why they hesitated.
AI can summarise thousands of research findings.
It still needs human judgement to determine which findings actually matter.
That's why the future of UX research isn't about replacing researchers with tools.
It's about giving researchers better evidence and more time to think.
Go Deeper
🛠 Tool to Try
Hotjar — explore heatmaps, session recordings, and user feedback to understand real behaviour.
📖 Article
“User Behavior: How to Track and Analyze it on Your Website” — Hotjar
A practical starting point for understanding how behavioural data can complement traditional research.
📑 Research Paper
“Usage and Application of Heatmap Visualizations on Usability User Testing: A Systematic Literature Review”
Explore how heatmap visualisations can complement usability research.
📑 Go Deeper
“Eye Tracking, Usability, and User Experience: A Systematic Review”
A useful exploration of behavioural evidence and how it can contribute to understanding user experience.
📚 Book
Don’t Make Me Think, Revisited — Steve Krug
A classic introduction to usability and understanding how people actually interact with digital products.
Try This Yourself
Pick one digital product you use regularly.
Don't redesign it yet.
Instead, ask:
What do users say?
What do users do?
Where is there a mismatch?
What could that mismatch teach us?
Then create three columns:
Observation → Interpretation → Opportunity
You'll quickly see how much richer research becomes when you move beyond simply collecting findings.
The Takeaway
Research isn't about gathering more information. It's about seeing what the information is telling you.
AI can give you more findings.
Tools can give you more data.
But human judgement turns evidence into insight.
Learn. Practise. Apply it on The Design Ethos.
Comments