What is a usability test?

A usability test is a session where a real person from your target audience tries to complete realistic tasks with your website, app or prototype while you watch. The aim is to find where they struggle, not to ask whether they like the design.

Can I run usability tests on a prototype instead of a live product?

Yes, and it is often the cheapest time to test. A clickable Figma prototype lets you find confusing steps before any code is written. Test the live product later to catch problems with speed, real data and devices.

Should I pay usability test participants?

A small thank-you, such as a voucher or a discount, is common and helps people turn up. Keep the amount modest so it does not attract people who only want the reward, and give it regardless of how the session goes.

How long should a usability test session last?

Around 30 to 45 minutes is enough for a small business test: a few minutes of introduction, three to five tasks and a short wrap-up. Longer sessions tire participants and produce less useful behaviour at the end.

What is the difference between moderated and unmoderated testing?

In moderated testing, a facilitator runs the session live, can ask follow-up questions and watches body language. In unmoderated testing, participants complete tasks alone with a tool that records them. Moderated gives more depth; unmoderated reaches more people faster.

Can friends and colleagues be test participants?

Avoid it for your main tests. They know too much, want to be helpful and rarely match your real customers. They are fine for a quick pilot session to check that your tasks and setup work.

How to Run Usability Testing With 5 Users on a Budget

You can spend weeks debating a checkout page in meetings, or you can watch five customers try to use it and know the answer by Friday. Learning how to run usability testing is one of the cheapest, highest-value skills a business owner or product team can pick up. You do not need a lab, special software or a big research budget, just a clear plan, a handful of the right people and the discipline to watch rather than explain.

This guide walks through a small-budget method from planning to fixing, including a sample task script and the mistakes that spoil results. It is the same basic approach we use within our UX audit services when a project includes research with real users.

Quick answer: Pick one or two key journeys, recruit about five people who match your real customers, and give each a realistic task to complete while thinking aloud on a video call or in person. Do not help or explain. Note where they hesitate, misunderstand or fail, group the problems you see repeated, fix the worst ones and test again with another small group.

Why five users is usually enough

Jakob Nielsen’s widely cited article Why You Only Need to Test with 5 Users argues that small, repeated rounds of testing find most of the serious usability problems more efficiently than one large study. After a few sessions, you start seeing the same problems again, and each extra session teaches you less.

Two caveats matter:

Five per user type. If your product has distinct groups, such as buyers and sellers or patients and clinic staff, test a few people from each.
Several small rounds beat one big one. Test, fix, then test again. The second round checks that your fixes worked and reveals problems the first ones were hiding.

Usability testing finds problems; it does not measure how common they are across all your customers. For that, use your analytics alongside it.

How to run usability testing: step by step

Step 1: Choose what to test and why

Pick one or two journeys that matter to the business, such as “find a service and send an enquiry” or “sign up and create the first invoice”. Write down what you want to learn: “Can new visitors find pricing?” is a useful goal; “Do people like the new design?” is not.

Your analytics can point you at the right journey. If one page has an unusually high exit rate or one form loses many people, start there. Our UX audit checklist helps you spot those problem areas first.

Step 2: Decide what people will use

A live website or app shows real speed and real content.
A clickable prototype lets you test before building. A well-linked Figma prototype is realistic enough for most tasks.
A competitor’s product can be tested too, to learn what your customers expect.

Make sure everything works from start to finish for the tasks you will set. Dead ends ruin sessions.

Step 3: Recruit the right five people

The participants should match your real customers, not your team. Low-cost sources:

Existing customers who agree to a short call
Your email list or WhatsApp business contacts
Social media followers or community groups related to your field
Visitors invited through a short pop-up or form on your site
People in your customers’ profession reached through professional networks

Use a two- or three-question screener to check they fit: their role, whether they have bought something similar, which device they use. Book one extra person in case someone does not turn up.

Step 4: Write realistic tasks

Tasks should describe a goal, not the steps. Avoid words that appear on your buttons, or you are testing word-matching rather than understanding.

Weak task	Better task
“Click on Services and find the pricing page.”	“You are thinking of hiring a firm for this. Find out roughly what it would cost.”
“Use the booking form.”	“Book a haircut for this Saturday afternoon with any available stylist.”
“Sign up for an account.”	“You have just heard about this app from a friend. Get started and set it up for your shop.”

Three to five tasks fill a 30 to 45 minute session comfortably.

Step 5: Prepare a simple script

A script keeps sessions consistent. A basic outline:

Welcome (2 minutes). “We are testing the website, not you. There are no wrong answers. Please think aloud as you go.”
Background questions (3 minutes). How they currently solve this problem, which devices they use.
Tasks (20 to 30 minutes). Read one task at a time and let them work.
Wrap-up (5 minutes). “What was the most confusing part? What, if anything, would stop you using this?”

Ask for permission to record the screen and audio before you begin.

Step 6: Run the sessions

A video call with screen sharing works well for websites and desktop software. For mobile apps, ask participants to join the call from a laptop and share their phone screen, or meet in person and record the screen.

The single most important rule: do not help. When a participant asks “Should I click this?”, reply “What would you do if I were not here?” The think-aloud method works because you hear their reasoning in real time. Gentle prompts such as “What are you looking for now?” keep them talking.

If you can, have a second person take notes so the facilitator can focus on the participant.

Step 7: Capture what you see

For each task, note:

Did they complete it? Easily, with difficulty, or not at all?
Where did they hesitate or backtrack?
What did they say they expected to happen?
Any quotes that capture the problem in their words

Write notes immediately after each session, while it is fresh.

Turning findings into fixes

After five sessions you will have many notes. Turn them into decisions:

List every problem on a sticky note or spreadsheet row.
Group duplicates. If three people missed the same button, that is one problem seen three times.
Rate severity. Did it stop people completing the task, slow them down or merely irritate them?
Prioritise severe problems on journeys that bring revenue, then quick wins such as unclear labels.
Decide the fix for each, and who owns it.
Retest the changed journey with a fresh small group.

Keep the summary short: the top five problems, evidence for each and the agreed fix. A short highlight reel of clips is far more persuasive to a sceptical team than a long report.

Low-cost tools that help

A video-call tool you already use, for moderated sessions with screen sharing
Figma for prototypes participants can tap through without an account
Microsoft Clarity or similar recordings to see how real visitors behave between test rounds
A spreadsheet for the task log and problem list
Maze or similar tools if you later want unmoderated tests with more people

Common usability testing mistakes

Testing with colleagues or family instead of real customers
Asking for opinions (“Do you like it?”) instead of watching behaviour
Leading tasks that use the exact words on your buttons
Helping participants when they struggle, which hides the very problem you need to see
Testing too much at once, so no journey is covered properly
Running one round only and never checking whether fixes worked
Treating one person’s comment as a rule, rather than looking for repeated problems

Usability testing checklist

Clear goal and one or two journeys chosen
Prototype or product works end to end for each task
Five participants per user type, plus a spare
Screener questions to confirm fit
Three to five realistic, goal-based tasks
Script with welcome, tasks and wrap-up
Recording permission requested
Notes template for each task
Problems grouped, rated and prioritised
Fixes made and retested with a new small group

How Omebly can help

Some teams prefer to run tests themselves; others want an outside view and a structured report. Our team can review your site or app against usability principles, analyse your analytics and recordings, run task-based sessions with your real users and turn the findings into a prioritised fix list with redesign concepts in Figma. If the fixes involve a new layout, our website UI design team can redesign the affected pages. Projects are led by founder Nazmul Hussain, who has 10+ years of experience building websites, apps and digital products.

Explore our wider UI/UX design services, or see what is included in our UX audit services, from expert review to user research.

Want to see your product through your customers’ eyes? Contact us or message us on WhatsApp at +91 70444 33991 for a free consultation.

How Do You Run Usability Tests With Five Users on a Small Budget?