User testing
Real people from your audience try your site or app on camera while we note every point they get stuck.
GradaxEuropean companyOrder outright
Founded 2014·Built and operated in Europe·1 to 4 weeks
About this service
You know your site too well to see its problems. We recruit people who match your customers, give them realistic tasks such as finding a product, booking an appointment or completing checkout, and record what they do and say. The result is a short report of the problems that came up repeatedly, ranked by how many people they stopped, with clips as evidence.
What five people can show you
Watching five people try to complete the same task on your site is usually enough to reveal the biggest problems in that journey. The first person surprises you. By the third, you see the same stumble again. By the fifth, you are mostly seeing repeats. This is a long-established finding in usability research and the reason small tests are so useful.
What five people cannot give you is percentages. If two out of five fail checkout, that does not mean 40 percent of your customers do. It means checkout has a problem serious enough to stop real people, and here is the footage of exactly where. For finding and fixing problems, that is what you need.
Recruiting the right people matters more than the number
Testing a specialist B2B ordering site with general consumers produces misleading results. They struggle with terminology your real customers know and breeze through steps that confuse actual buyers. The value of a test depends heavily on whether the participants resemble your customers.
We screen participants against a profile we agree with you: their role or life situation, how familiar they are with products like yours, the device they would normally use. Tasks are written as realistic scenarios rather than instructions. "You need a gift for a friend under 50 euros, delivered by Friday" reveals far more than "click on the gifts category".
- Tasks written as real situations, not click instructions
- Participants asked to think aloud as they work
- No help given during the task, even when they get stuck
- Problems counted only when they recur across participants
What you get
Everything below is in the price. Nothing here is an upgrade.
- Test plan with tasks and participant criteria
- Recorded sessions
- Findings report with severity ranking
- Highlight clips of key problems
- Readout call
What changes
- Tasks completed or failed by real people from your audience, on video
- Problems ranked by how many participants hit them
- Clips showing each serious problem, for anyone who doubts it
- A fix list with the change most likely to remove each problem
How it runs
Plan
Tasks and participant profile agreed, based on what the business needs to know.
Recruit
Participants screened to match your real customers.
Observe
Sessions recorded as people think aloud through each task.
Report
Repeated problems ranked by severity, with clips and fixes.
Compare the sizes
The standard price for each size. Extras are priced beside it, not hidden behind a call.
| Included | Five unmoderated tests€2901 week | Eight tests with report€5002 weeks | Moderated study€1,0503 to 4 weeks |
|---|---|---|---|
| Five participants | Included | Included | Included |
| Up to three tasks | Included | Included | Included |
| Recorded sessions | Included | Included | Included |
| Findings summary | Included | Included | Included |
| Mix of desktop and mobile | Not included | Included | Included |
| Up to five tasks | Not included | Included | Included |
| Severity-ranked report with clips | Not included | Included | Included |
| Readout call | Not included | Included | Included |
| Six moderated live sessions added | Not included | Not included | Included |
| Follow-up questions on motivation and trust | Not included | Not included | Included |
| Retest of the top fix after you change it | Not included | Not included | Included |
| Order | Select | Select | Select |
Add what you need
Choose the size
Add what you need
Your total€500
ContinueThe standard price, plus whatever you add. Confirmed in writing before anything starts; nothing is charged until you accept it.
Questions
Are five users really enough?
For finding the biggest problems in one journey, five usually reveals most of them. It is not enough to measure percentages, which is a different kind of study.
Do participants get paid?
Yes, participant incentives are included in the package price.
Can you test a design before it is built?
Yes. A clickable Figma prototype works well, and fixing problems there is far cheaper than after launch.
Related searches
Related work
ScopedStore conversion reviewWhere the money leaks between the product page and the confirmation.from €890on scope
ScopedAccessibility remediationAn audit against WCAG, then the fixes, then the proof.from €890on scope
ScopedUI design sprintScreens designed and prototyped in two weeks, ready to build.from €1,800on scope
Order it
Standardised enough to order on its own. The price is confirmed in writing before anything starts.