blog single

Usability testing on a small budget

Colleagues observing a usability session

Published on:

November 13, 2025

TAG:

UI/UX DESIGN

BY:

Edgeless Team

Read Time:

5 MIN READ

post detail

Usability testing has a reputation for being slow and expensive. In practice, five short sessions with the right people will find most of the problems in a flow. The cost is mostly scheduling and preparation, not equipment. A small team can run a useful test in a week and change the design before the next release.

Test one real task

Pick one task that matters and can be completed in fifteen minutes. Request a quote, find a document, book an appointment, update a payment method. A focused task produces specific findings. A general instruction to explore the site produces polite feedback and very little evidence.

Write the task as a goal, not as a set of steps. Ask the participant to find out how much a service costs and request a quote. Do not tell them to open the services menu and click pricing. The route they choose is part of what you are testing.

Choose participants who resemble the audience. Five people who have done the task in real life will expose more than twenty colleagues who know the product. Recruit through customers, a partner, a local network or a research panel, depending on the audience.

Run the session without leading

Ask the participant to think aloud and explain that you are testing the design, not them. Stay quiet while they work. When they ask whether they are doing it right, answer with a question such as what would you do next. The silence is uncomfortable and it is where the findings come from.

Record the screen and the audio if participants agree. Notes are useful, but a recording captures the exact moment someone hesitated, the words they used and the thing they tried first. Review the recording the same day while the details are fresh.

Have one person run the session and one person take notes. The facilitator focuses on the participant and the note taker records observations without interpreting them. Combining the roles splits attention and loses detail.

Designer taking notes during a user session — Five sessions, one task and a same day review will produce more actionable findings than a survey sent to a thousand people.

Look for patterns

When three of five participants make the same mistake, that is a pattern worth fixing. When one person dislikes a color, that is a preference and it can wait. Separate observations about behavior from opinions about taste, and report them differently.

Write findings as problems with evidence. Participants looked for pricing in the footer because the navigation label did not mention cost. That is a finding. The navigation is confusing is an interpretation without a fix.

Rank the problems by the cost of the mistake. A blocked task is more serious than a slow one. A confusing label on the main flow is more serious than a small issue in a settings page. Fix the blockers first and retest the same task.

Make the test a habit

Run a session every month or every release, not once a year. A small recurring test catches regressions and keeps the team close to the people who use the product. The cost is a few hours and the benefit compounds.

Keep a shared library of recordings and findings. New team members learn faster from five clips of real frustration than from a design document. The library also shows whether a problem was fixed or merely discussed. Review the clips together at the start of each design cycle so the evidence stays in the room.

Close the loop with participants. A short message explaining what changed because of the session builds goodwill and makes it easier to recruit the next time. People join research when they see that it leads somewhere.

Recruiting without a panel

A short message to recent customers is usually enough. Explain what the session involves, how long it takes and what they receive. A small thank you payment improves response rates and signals that their time has value.

Screen for the behavior you are testing, not for demographics. Someone who has requested a quote in the last six months is a better participant than someone who matches an audience profile but has never done the task. Ask two screening questions and stop there.

Book a sixth session as a backup. People cancel, and a no show in a five session plan leaves you with four. Confirm the day before and keep the calendar slot open for a replacement.

Remote sessions work

A screen share with a phone or a laptop is enough for most tests. The participant shares their screen, thinks aloud and uses the site as they normally would. You see the hesitations, the wrong turns and the exact words they search for.

Ask them to use their own device and their own network. A test on a fast office connection with a large monitor hides the problems that appear on a phone in a poor signal area. The device is part of the context.

Have a fallback for technical problems. A phone call with a shared link can replace a video session, and a short delay is better than a cancelled slot. Keep the session to thirty minutes so a problem does not consume the whole booking.

Report findings in a page

List the problems in order of severity, with the evidence and the suggested change. Add a clip for each one. A page that a product owner can read in five minutes is more likely to produce a fix than a forty page report with an executive summary.

Separate what you observed from what you recommend. The observation is the evidence. The recommendation is a judgement that the team can accept, modify or reject. Mixing them makes it hard to see which part is data.

Common mistakes

The most common mistake is explaining the interface during the session. The moment you tell someone where to click, the test stops measuring the design and starts measuring your instructions. Let them struggle, then ask what they expected.

The second is testing with colleagues. They know the product, they know the vocabulary and they are polite about problems. Use them for a pilot run of the script, then test with people who have never seen the site.

What you do not need

You do not need a lab, eye tracking or a statistically significant sample for most product decisions. You need a realistic task, a representative participant and a reviewer willing to watch without defending the design. Those three things are available to almost any team.

You also do not need a long report. A one page summary with three problems, the evidence and the proposed fix is more likely to be read and acted on. Attach the clips for anyone who wants the detail.

You do not need to test every audience in the first round either. Pick the group with the most to lose if the flow fails, learn from them and add a second group when the first set of fixes ships. Research compounds when it follows the release cycle.

Five people attempting one real task will teach you more than a survey with a thousand opinions.
Edgeless Team

We run usability sessions, turn the findings into a prioritized fix list and retest the flow after the changes ship.

Explore UI and UX design

Start with the flow that loses the most customers. Test it, fix the top three problems and test again. The routine is small enough to keep and specific enough to change the product.

Usability

Research

Usability testing on a small budget

Let's Talk About Your Project