Kristina ;D
How to Check Your Website's Usability With Five People
You can find many of your site's biggest usability problems by watching five people try to use it. Give each one a simple task, stay quiet, write down where they hesitate and then fix the problems that appear more than once. This small test costs little and often shows problems that guessing would miss.
What a usability test is
A usability test is an observation, not an interview. You are not asking people whether they like the site. You are watching whether they can do what a real visitor would want to do. Opinions are useful, but behavior is more honest: someone may say a page is fine and still be unable to find the contact link.
Why five people are enough to start
The same problems tend to appear again and again. After a handful of sessions, you will see the patterns, and further sessions repeat what you already know. It is more useful to test five people, fix the problems and test again than to run one big session with many people.
Who to ask
Choose people who resemble your real visitors: potential customers, not your closest friends who already know what you do. If you cannot find them, ask acquaintances who have never seen the site. Offer a small thank-you, such as a coffee or a gift card, and be clear that you are testing the site, not them.
Choose three or four tasks
Tasks should reflect what visitors really want. Write them as situations, not instructions:
- "You want to know whether this person can help with your project. Find out what services are offered."
- "You want to see examples of previous work. Show me where you would look."
- "You decide to get in touch. How would you do it?"
- "You want to know roughly how long a project takes. Where would you look?"
Avoid using the words that appear on your buttons, or you will test reading, not finding.
How to run a session
- Explain that you are testing the site and that nothing the person does can be wrong.
- Ask them to think aloud: say what they see, what they expect and what confuses them.
- Give one task at a time. Do not help, even when you want to.
- Write down what they do, what they say and where they stop.
- At the end, ask what they would change and what they remember about the site.
Use a phone for at least some of the sessions. Many people will visit from one, and problems often appear there first.
What to look for
- Hesitation: long pauses before a click.
- Wrong clicks: going to a page that does not contain what they expect.
- Repeated scrolling up and down without finding the answer.
- Misunderstanding: describing your offer differently from how you describe it.
- Giving up: saying "I would leave here".
Turn notes into a short list of fixes
After all sessions, group the notes by problem. Count how many people met each one. Problems that stopped several people are the first to fix. Write each fix as a small, specific change, such as "rename the menu item to Services" or "place the contact button at the top". Then change the site and test again with new people, if possible.
| Problem | People affected | Possible fix |
|---|---|---|
| Could not find contact | Several | Add a button in the header |
| Unclear offer on first screen | Several | Rewrite the headline in plain words |
| Small text on phone | One or two | Increase size and contrast |
Combine it with other signals
Observation tells you why something happens. Numbers show how often. If your analytics show that many visitors leave a page quickly, a usability test can help explain why. For more about the numbers, see website conversion rate, and for the principles behind the test, what user experience design is.
What to say, and what not to say
During the session, keep your comments short and neutral. Useful phrases are "What are you thinking now?" and "What would you expect to happen?" Avoid explaining, defending the design or hinting at the answer. If a participant asks "Should I click here?", answer "What would you do if I were not here?" It feels uncomfortable at first, but it is how you see the real behavior.
Do not take criticism personally. A participant who cannot find a button has given you a gift: a problem you can fix before it costs you customers.
A low-effort version for busy people
If you cannot arrange sessions, send the link to five people with one question: "Where would you click to contact me?" or "What do you think this page offers?" Their answers show whether the main message is clear. It is not as rich as a live observation, but it is much better than assuming that everything is obvious.
Common objections, answered
People often say that they have no time, that friends will not be honest or that the site is too small to test. A session often takes about half an hour, and honesty comes easily when you ask people to think aloud and tell you what is confusing. A small site needs testing as much as a big one, because every visitor matters more when there are few. Even one session will probably show something you did not expect.
FAQ
Can I test my site myself?
You can try, but you know the site too well. Other people will find problems you cannot see.
Do I need special software?
No. A phone, a computer, a notebook and some patience are enough. Recording the screen, with permission, can help.
How often should I test?
After any major change and at least once a year. Small tests are better than rare big ones.
What if people say different things?
Pay attention to what they do, and to problems that repeat across people.
Should I fix everything I find?
Start with problems that block the main goal, such as contacting you, and leave small details for later.