The takeaway: You do not need a lab or a research budget to test your product's usability. Pick one task your users must complete, watch a handful of people try it, note exactly where they stall, and fix the biggest stall. Then test again.
What a usability test actually is
A usability test is simple: you ask someone to complete a task in your product while you watch and listen. Nielsen Norman Group, a UX research consultancy that has studied this methodology for decades, defines it the same way: a facilitator gives a participant a task, then observes their behavior and listens for their feedback as they work through it.
You are not asking whether they like your product. You are watching whether they can use it. That difference matters, because opinions are easy to collect and behavior is what shows you what to fix.
Finding people to test with is easier when real users already know you exist. You can launch your startup on Maidensail to get in front of founders and early adopters who may become your first testers.
How many people you need
Jakob Nielsen's widely cited finding is that testing with 5 users uncovers the large majority of usability problems in a design. His reasoning: after the first few participants, you start seeing the same problems repeat. His research, originally published with Tom Landauer in 1993, models this as a curve that flattens sharply after a handful of users.
Two caveats from the same source are worth keeping in mind:
- The point is iteration, not one big study. Nielsen recommends running several small rounds of about 5 users instead of one large study, because the goal is to fix the design between rounds, not to document every possible flaw in one pass.
- Distinct user groups need separate coverage. If your product serves very different kinds of users (say, admins and end customers), Nielsen suggests testing 3-4 people from each group rather than 5 from one.
For an early-stage product, that means your first test can be 3-5 people and an afternoon of your time.
Before the test: pick one task
Choose a single, concrete task that a real user would want to complete - ideally the action that matters most to your product. Good examples:
- Sign up and complete the first meaningful action
- Find a specific setting and change it
- Complete a purchase or invite a teammate
Write the task as a realistic scenario, not an instruction. "You just heard about this tool from a friend - sign up and set up your first project" works better than "click the sign-up button," because it leaves the participant to figure out the path the way a real user would.
Running the session
You can do this over a video call with screen sharing, or in person. The mechanics:
- Ask them to think aloud. A running commentary ("I'm looking for the settings... I expected this to be in the menu") is often the most useful data you will collect.
- Do not help. When they struggle, let them struggle a bit. Where they stall is the finding. If they ask "should I click this?", a good response is "what do you think?"
- Take notes on behavior, not conclusions. Write down what happened ("hesitated 20 seconds on the pricing page, clicked FAQ twice") rather than your interpretation. Interpretation comes after.
- Ask follow-up questions at the end. "What did you expect to happen there?" is a useful one.
If you have never talked to users this way before, our guide on how to talk to your users and hear the truth covers the interviewing side of this skill.
A common first task to test is your onboarding, since every new user goes through it. If that is what you test, our post on onboarding that turns signups into users can help you decide what "working" looks like.
What to do with the results
After 3-5 sessions, patterns usually surface on their own. A practical way to process them:
- List every stall point. Each place a participant hesitated, backtracked, or failed.
- Rank by severity, not frequency alone. A problem one participant hit that completely blocked the task can matter more than a minor annoyance three people mentioned.
- Fix the top one or two, then test again. This is the iterative loop Nielsen's research points to: small test, fix, small test again. Each round tends to surface a new, smaller set of problems.
Avoid the temptation to fix everything at once. A short list you act on beats a long report nobody reads.
The takeaway
Usability testing is a small, repeatable habit, not a project: one task, a few people, careful watching, one fix at a time. The first session usually pays for itself within the hour, because watching a real person use your product almost always shows you something you could not have guessed from inside your own head.
Your launch deserves users who stick around
A product people can actually use is worth telling the world about. When you are ready, launch your startup on Maidensail and put it in front of founders and early adopters who are looking for what is next.
Sources: Nielsen Norman Group - Why You Only Need to Test with 5 Users (Jakob Nielsen, 2000, based on Nielsen & Landauer, 1993); Nielsen Norman Group - Usability Testing 101 (Kate Moran, 2019, reviewed 2026).

