Conversion and UX

Usability Testing

Also called User testing

Watching real people attempt a task on your site to find where they hesitate, misread or give up.

Quick facts: Usability Testing

Category
Conversion and UX
Also called
User testing
Level
Intermediate
Affects
Enquiry rate, navigation, wording of services
Where to see it
A phone, a task list, screen recording, remote testing platforms
In this article4
  1. How usability testing works
  2. Why usability testing matters
  3. Where usability testing goes wrong
  4. How to act on it

How usability testing works

You give somebody a realistic task — find the price for this service, book an appointment for next week, work out whether you deliver to their town — and you watch them try. You do not explain the site, you do not hint, and you do not defend it. You ask them to think out loud, and you write down every place they pause, misread something or take a wrong turn.

It can be moderated, with you present and asking questions, or unmoderated, with the person recording themselves while working through instructions. Either way the value comes from the same source: watching the attempt rather than asking for an opinion. What people say about a website and what they do on it are frequently different, and only one of the two costs you enquiries.

Why usability testing matters

Everybody who built the site knows where everything is, which makes them the worst possible judge of whether it can be found. Testing replaces that blind spot with evidence, and it does so before money is spent driving traffic to a page that does not work.

It explains things numbers only report. Analytics can tell you that people leave the booking page; a person talking through their confusion tells you it was the date field, or the word you used for a service that customers call something else. That kind of finding is usually cheap to fix and immediately useful.

It needs very little traffic, which makes it the right method for sites that cannot support formal testing. A new business with few visitors can still learn a great deal from a small handful of honest attempts.

Where usability testing goes wrong

Testing on colleagues is the usual shortcut, and it fails because they know the offer, the jargon and the layout. The next fault is leading: asking whether the button was easy to find teaches you nothing, because most people will say yes to be polite. Set the task and stay quiet.

Testing on the wrong device matters more than people expect. If your customers arrive on mid-range phones over mobile data, a test conducted on a laptop over office broadband will miss the problems that actually cost you enquiries. And a session that turns into a demonstration, where you rescue the participant every time they stall, produces a pleasant conversation and no findings at all.

How to act on it

Start with one task that matters commercially, recruit people who resemble your customers rather than your team, and run the session on the device they would really use. Record what they do if they agree to it, note the exact moment of each hesitation, and resist explaining until the task is over.

Fix the problems that stopped somebody completing the task first, then the ones that merely slowed them down. Re-test after the changes, because fixes create new confusion often enough to be worth checking. Pair the findings with session recordings to see whether real visitors behave the same way, and feed the conclusions into the build rather than the backlog — this is the cheapest input available to web design and development work, and the easiest to skip.

Do and do not

Do

  • Set a real task and then stay quiet
  • Test on the device your customers actually use
  • Fix blockers first, then re-test with new people

Do not

  • Test with colleagues who already know the site
  • Ask whether something was easy to find
  • Rescue the participant the moment they stall

Questions people ask about this

How many people do I need to test with?

Fewer than most people assume. The serious problems tend to appear with the first few participants, because they trip everybody in the same place. Rather than recruiting a large group at once, test with a small number, fix what you found, then test again with fresh people. Repeated small rounds beat one large session.

Can I test with friends and family?

Only if they genuinely resemble your customers and do not already know your business. Someone who knows what you sell will navigate on knowledge rather than on what the page says, which hides the exact problem you are looking for. If you must use people you know, pick ones who have never seen the site.

What is the difference between usability testing and A/B testing?

Usability testing tells you why something fails, by watching a person struggle with it. A/B testing tells you which of two versions performs better across a lot of traffic, without explaining the reason. They answer different questions and work well together: watch people to find the problem, then test a fix if your traffic allows.

Related terms

Found this useful?

Share it, or ask an AI to summarise it

Back to the glossary

Knowing the term is the easy part

Applying it to your own site and budget is the work. Book a call and I will tell you what actually applies to you.