Website usability testing means watching real people try to complete tasks on your site, like requesting a quote or buying a product, and noting where they get stuck. Testing with about 5 people per round is usually enough to uncover most major problems, and you can run it in a few days without expensive tools.
If your site gets traffic but few leads or sales, the cause is often the experience, not the traffic. This guide shows how to run a usability test from start to finish, what to measure, and when to bring in a professional.
What website usability testing is and why it matters
Usability testing is observing people use your site to reach real goals. You don't ask "do you like it?" You ask "find the price for service X and request a quote," then watch.
It answers practical questions:
- Can people find what they need without help?
- Where do they hesitate, go back or give up?
- Is the main button obvious?
- Is the form easy to fill out on a phone?
Analytics tell you that something is wrong, such as a high bounce rate. Testing shows you why. They work best together. Teams who understand the difference between UI and UX know that ease of use matters as much as looks.
When to test your website
You don't have to wait until the site is finished. Testing early is usually cheaper than fixing things later.
- Before building: test a prototype (a clickable draft) to avoid rebuilding finished pages.
- Before launch: catch obvious problems your team no longer sees.
- When conversions are low: lots of visits and few inquiries is a classic sign.
- Before a redesign: learn what already works and should stay.
- After big changes: new navigation, new checkout, new services page.
How many people to test
The Nielsen Norman Group, a leading UX research firm, recommends testing 5 users in a qualitative study: you find nearly as many problems as with a much larger group, at the best return on investment. They suggest roughly 3 to 4 per group when your audiences differ a lot.
In practice, run one round with 5 people, fix what you find, then repeat. That beats one giant round.
Step-by-step: how to run a website usability test
1. Define the goal
Pick one clear question. Weak: "see if the site is good." Strong: "can a new visitor request a quote in under 2 minutes?"
2. Choose 3 to 5 realistic tasks
Describe scenarios, not clicks. Instead of "click Services," say: "You need a website for your clinic. Find out whether this company does that and how to contact them."
Good tasks for service businesses: find what the company does, find a work sample, understand how to hire them, request a quote.
3. Recruit the right people
Invite people similar to your ideal customer. Avoid close friends and your own team, since they already know the business. Recent customers and people in your target group work well. A small thank-you, like a coffee or a discount, helps.
4. Pick a format
- In person, moderated: you sit beside them and ask questions. Richest detail.
- Remote, moderated: same thing over a video call with screen sharing.
- Remote, unmoderated: the person completes tasks alone while the screen is recorded. Cheaper and faster, but you can't ask follow-ups.
For a first test, remote moderated is a good balance.
5. Run the session without leading
- Say you're testing the site, not the person.
- Ask them to think out loud.
- Don't help, even when it's hard to watch. Every mistake is a finding.
- If they ask "should I click here?", answer "what would you expect?"
- Note where they hesitate, go back, or say "I don't get this."
Get permission before recording and explain how the footage will be used.
6. Track simple measures
- Did they complete the task? (yes, with help, or no)
- How long did it take?
- How many errors or backtracks?
- Where did they give up?
- What did they expect to find?
7. Prioritize and fix
Combine notes from all 5 sessions. If 3 or more people struggled at the same point, fix it first. Rank issues by how much they block inquiries or sales and by the effort to fix. Tackle severe and easy first.
8. Test again
After fixing, run another round. Usability testing is a short loop: test, adjust, repeat.
Common usability testing mistakes
- Testing with your own team: people who know the site can't see its problems.
- Leading questions: "this menu is clear, right?" guarantees false answers.
- Vague tasks: without a goal, people just wander.
- Ignoring mobile: many visitors use phones. Test there too.
- Fixing everything at once: you won't know what worked.
- Relying only on heatmaps: they show where people click, not why.
Signs your website needs a usability test now
- Plenty of visits, few inquiries.
- Customers ask by message things that are already on the site.
- The form is started but rarely submitted.
- Sales staff say prospects "couldn't find" something.
- The site was built years ago and never observed from the outside.
If several sound familiar, a professional review of the structure helps. That is part of UI/UX design work.
What to ask a vendor when building a new site
Ask whether the project includes validation with real users before launch. Ask how they test prototypes and how findings become changes. Also plan the schedule: our guide on how long it takes to build a website helps you fit a testing stage without delaying launch.
Frequently asked questions
How much does website usability testing cost?
It depends on the format. A DIY test with 5 people over video can cost just your time and a thank-you gift. Professional testing with recruiting, a script and a report costs more. Price varies with the number of participants, site complexity and depth of analysis.
How many users do I need to test my website?
About 5 users per round surfaces most usability problems, according to the Nielsen Norman Group. If your audiences are very different, test 3 to 4 from each group. After fixing issues, repeat with a new small group.
Is usability testing the same as A/B testing?
No. Usability testing watches a few people to understand why they struggle. A/B testing compares two page versions with many visitors to see which converts better. The first finds problems; the second helps choose between solutions.
Can I run a usability test myself?
Yes, and it's a good start. Prepare clear tasks, invite people from your audience and don't help during the session. The risk is bias: whoever built the site tends to defend their choices. If possible, have a neutral person run or analyze it.
How often should I test my website?
Test before launch, before major changes, and whenever conversions drop. For sites that drive sales, a small round every few months keeps the site aligned with real customer behavior.
How DevzDesign can help
DevzDesign builds websites, landing pages and digital products focused on clarity and results. If you suspect your site is losing leads to navigation problems, we can help plan a test, interpret the findings and turn them into practical UX and UI improvements.
Want to talk? Contact us and tell us about your situation.
