What is usability testing and what it tells you

On-Off Group

September 2026

Usability testing is watching real customers try to complete a real task on your product while you stay quiet and record what happens. It tells you where people get stuck, how long the task takes and how many finish, so you have a before number to compare against later.

What is usability testing

Usability testing means giving a real person a real task on your product, then watching them try to do it while you keep quiet. No demo, no explanation, no helping. You record the screen, you record what they say, and you count whether they finished.

This is for anyone who has to explain usability testing to a boss before there is budget for it. It covers what it is, what it will and will not tell you, and the number to write down before you change anything.

Surveys and focus groups ask people what they think. This is different. You watch someone fail at a step your team assumed was obvious, and nobody asks them whether they liked it.

What it looks like on the onboarding flow

Say the onboarding flow has been live for a year. Everyone on the team has an opinion about which step loses people. Branch staff say customers call them about the ID upload. Marketing says the form is too long. Nobody has watched a customer do it end to end.

So you book five people who match the customers you are worried about. Each one gets about 30 minutes. You give them one instruction, something like "open an account using this link". Then you stop talking.

What comes back is usually blunter than any internal review. People read the ID instructions three times and still upload the wrong side. They stop at a field that asks for something they do not have on them. One person closes the app and says they will do it in a branch instead.

You now know which step to fix, and you can show your boss the clip of it happening rather than describing it. If you want the mechanics of running that session, we wrote them up in How to run a usability test with five real users.

What usability testing tells you, and what it does not

It tells you where people get stuck, what they expected to happen instead, how long the task took, and how many finished without help.

It does not tell you how many customers in total have this problem. Five people cannot give you that. It does not tell you what customers will pay for, and it does not tell you whether the product is a good idea. Those are different questions with different methods.

Be careful how you word the findings when you present them. If you say five users prove there is a market, someone in the room will tear that apart, and your real findings get thrown out with it. Say the smaller true thing instead: five out of five failed at the ID step, here is the recording. That is enough to get the ID step fixed.

Big vague claims get picked apart in the meeting. A small honest number tends to get the work approved.

The before number, and why it usually does not exist

Most teams we work with have no baseline. The work gets done, the screens change, and when the boss asks whether it worked the honest answer is a shrug.

Usability testing is the cheapest way to get a before number, because you are counting finishes and timings anyway while you watch. Write down two things before you change anything. How many of the five finished the task without help. How long it took them, on average, from first tap to done.

After the fix, run the same task with five different people and compare. Now you have something to answer with when your boss asks if it worked.

On a series of five-day design sprints with a large insurer, the customer-facing journeys that used to take six months or more to get out were designed and tested with customers in two weeks and live in four. Completion rates on the new journeys went up by 80 percent against the old ones, and the old drop-off points disappeared. The client measured it. They could only do that because someone had written down how the old journeys performed first.

How to explain it before you have funding

Keep it short when you take it to your boss. A few sentences is enough.

What we will do: watch five customers try to complete the onboarding flow, on video, over two weeks. What we will come back with: the step that loses people, a clip of it happening, and the number who finished. What you need approved: five participants, a facilitator, and about a day of someone's time to cut the clips. Price those against your own usability testing rates.

Someone will say the team already knows what is wrong. Answer that with the recording. In an internal review the most senior opinion usually wins. Play a clip of a customer giving up on the ID upload and the argument stops.

If you cannot get participants approved this month, going through the screens against usability standards will still show you the obvious problems. We set out how we do that in How to review a website or app against usability standards in 60 minutes. It is weaker evidence than watching a real customer, and you should say so when you present it.

What to do this week

Pick one task, not a whole product. "Open an account", "pay a bill", "submit a claim".
Write down the before number for that task today: how many people finish, and how long it takes. If nobody knows, that is your finding.
Book five people who match the customers you are worried about. Not colleagues.
Run the sessions with one instruction and no help, and record the screen.
Take your boss a couple of minutes of clips and the number who finished. Leave the slides out.

Questions people ask

How many people do you need for a usability test?
Five people who match the customers you are worried about is enough to find the big blockers. After five, the same problems keep repeating rather than new ones appearing.

Is usability testing the same as journey mapping?
No. Journey mapping is your team drawing what it believes happens. Usability testing is watching a customer actually do it, which is why the two often disagree.

What number should we record?
Pick the plainest one for the task you are testing: how many of five people finished without help, and how long it took them. Record it before any changes so you can compare after.

Does usability testing work for internal tools too?
Yes. Branch staff and back office users struggle with screens in the same way customers do, and turnaround time is usually easier to measure internally.

On-Off Group trains teams, tests products with real customers, finds where a transformation has stalled and builds what gets it moving, for banks, insurers and enterprises in the Philippines, since 2015. How we help with usability testing.