This guide is for product owners, founders and product managers of web apps and SaaS products who have no dedicated UX researcher. It sits under Netbase's software product engineering work and complements our UI/UX design service, where short usability rounds are part of the design process.
In this guide
- Can a product team run usability tests without a researcher?
- How many participants does a small usability round need?
- How do you write usability test tasks and a session script?
- What should you measure in a small usability test?
- Which usability test format fits your question?
- A first usability round in seven steps
- Where AI fits
- What delivery record exists, and what does not
- Alternatives: other ways to find usability problems
- Limits of this guide
- Frequently asked questions
- Next step
Can a product team run usability tests without a researcher?
Yes, for the most common question: can people complete the key tasks, and where do they get stuck? That needs observation, not statistics, and a product manager, designer or developer can learn it quickly.
Nielsen Norman Group describes thinking aloud, where participants say their thoughts while using the interface, as cheap and robust: a poorly run study still produces reasonably good findings unless the facilitator puts words in the participant's mouth. That robustness is what makes a self-run round worth doing.
A team without a researcher should not claim what share of all users will fail, or compare designs statistically from a handful of sessions. Those questions need larger samples and a trained analyst.
How many participants does a small usability round need?
Five participants from one user group is the standard starting point. Nielsen Norman Group reports that testing with five users uncovers about 85% of usability problems, and that each extra participant mostly repeats what you have already seen.
The same source recommends several small rounds instead of one large one, so each round checks the fixes from the last. When a product serves clearly different groups, such as buyers and approvers, it suggests three to four people per group for two groups and three per group for three or more.
A practical round card for a team without a researcher:
-
Decision
One design question per round, written before recruiting
-
Participants
Five from one user group; three per group when there are several
-
Session
Under an hour: warm-up, four to six tasks, short wrap-up
-
Roles
One facilitator who talks, one note-taker who only writes
-
Debrief
Same day, while notes are fresh, with severity agreed together
-
Next round
After the top fixes ship to the prototype or build
Recruit real users, not colleagues: a colleague knows the product vocabulary and hides the confusion you are testing for.
How do you write usability test tasks and a session script?
Write tasks as situations the participant recognises, not as steps. Nielsen Norman Group's rules for task scenarios are to make tasks realistic, make them actionable (ask people to do the thing rather than describe how they would), and avoid clues such as the label of the button they should press.
A task-writing checklist:
- Goal, not steps. "Your team needs a second admin account; set one up" instead of "Open Settings and click Users."
- No interface words. If the menu says "Workspace", the task does not.
- A clear end state. The facilitator knows when the task is done and what counts as success before the session starts.
- One task, one question. If a task tests two screens, split it, or you will not know which one failed.
A session script keeps five sessions comparable:
- Welcome. Explain that the product is being tested, not the person.
- Consent. Ask permission to record the screen and voice.
- Warm-up. Two or three questions about how the participant does this job today.
- Thinking aloud. Ask them to say what they are looking for as they go.
- Tasks. Read each task, hand over the written card, then stay quiet. Answer questions with "What would you do if I weren't here?"
- After each task. Ask the Single Ease Question (below) and one open question about what was hardest.
- Wrap-up. Ask what they would change first, thank them and stop the recording.
What should you measure in a small usability test?
Measure what five sessions can show honestly: task completion, help needed, ease and problem severity. Keep counts as counts: "three of five" is honest; a percentage invites a precision the sample does not have.
-
Task outcome
- How to record it
- One of four levels: success, success with a minor issue, success with a major issue, failure
- Report it as
- Count per level, per task
- Do not
- Average the levels into one number
-
Assists
- How to record it
- Each time the facilitator had to help
- Report it as
- Count per task
- Do not
- Hide assisted completions inside "success"
-
Ease
- How to record it
- Single Ease Question, 1 to 7, after each task
- Report it as
- Each participant's score next to their notes
- Do not
- Compare five scores with a published average
-
Problem severity
- How to record it
- 0 to 4 per problem, agreed at the debrief
- Report it as
- A ranked problem list
- Do not
- Rank by how often the team discussed it
The four outcome levels follow Nielsen Norman Group's guide to success rate, which warns against averaging levels and recommends reporting how many users reached each level. The Single Ease Question asks "Overall, how difficult or easy was the task to complete?" on a seven-point scale immediately after the task; MeasuringU reports its average at around 5.5, which is useful context for large studies, not a pass mark for five sessions.
Severity combines frequency, impact and persistence. Nielsen Norman Group's severity scale runs from 0 (not a usability problem) to 4 (usability catastrophe, fix before release). Rate each problem together at the debrief, then fix the 4s and 3s before the next round.
The System Usability Scale is a ten-item questionnaire scored from 0 to 100, with an average of 68 in MeasuringU's data. With five people, compare your own rounds with each other, not with that benchmark.
Which usability test format fits your question?
| Criterion | Moderated remote | Unmoderated remote | In person |
|---|---|---|---|
| Best for | Complex workflows and B2B tools where follow-up questions matter | Short, clear tasks on a public product | Hardware, kiosks or devices you must hand over |
| Facilitator needed | Yes, live | No; the script must stand alone | Yes, in the room |
| Main risk | Facilitator leading the participant | Participants who rush or misread the task | Travel and scheduling |
| Choose it when | You need to understand why people struggle | You need to see whether a fix works | Context changes behaviour |
Do not run a usability round to answer questions it cannot answer. Whether people want the product, which price they accept, and which of two visual styles they prefer are demand and preference questions; use interviews, analytics or an A/B test on live traffic instead.
A first usability round in seven steps
-
Write the decision
One sentence: what will you change, keep or drop depending on the result?
-
Pick the user group and recruit five
Screen for the real role and experience level; book one spare.
-
Write four to six tasks
Apply the checklist above and define success for each before anyone sees them.
-
Pilot the script once
Run it with one person outside the team and fix unclear tasks.
-
Run the sessions
One facilitator, one note-taker, recordings on, no help unless the participant is fully stuck.
-
Debrief the same day
List problems, rate severity together and agree the top fixes.
-
Fix and test again
Change the prototype or build, then run the next round on the same tasks to check the fixes.
Where AI fits
AI tools can take clerical work out of a round: drafting task wording, transcribing recordings and grouping session notes into candidate problems. A person still watches every session, decides which notes are real problems and agrees severity, because a summary cannot see where a participant hesitated. Netbase works with the major commercial and open-source AI tools and models, chosen per project, under human control.
- In delivered work
Recommendation engine
A recommendation engine delivered for 4over4's online printing store; it is not a usability tool.
- Available capability
Session-note assistants
Transcripts, note grouping and task drafts reviewed by people; not yet tied to a published case.
What delivery record exists, and what does not
Netbase JSC was founded in 2012 and has served 500+ clients. It designs, builds, integrates and runs software across technologies and industries. Its delivery teams combine business analysts, project managers, solution architects, developers, QA and UI/UX designers, so the people who watch a usability round can also change the design and the code.
Netbase's delivery lifecycle starts with discovery and strategic alignment and runs agile execution with outcome-based milestones, which is where short test rounds fit. In custom development, the client owns the IP created for it, including prototypes and research notes made for the project. One registered record shows interface work on a live product: for a large-format printing company, Netbase redeveloped the website's UX, UI and most of the system in several phases.
Netbase does not publish a usability score, a task success rate or a count of test rounds. This guide explains a method; it is not a case study.
Alternatives: other ways to find usability problems
- Expert review against heuristics. Fits an existing interface when you need a fast first list; it finds likely problems but cannot show which ones real users hit.
- Analytics and session data. Fits a live product with traffic; it shows where people drop off, not why.
- A/B tests. Fit a live product with enough traffic to compare two versions on one outcome.
Our design system adoption guide covers the next step once tested patterns become shared components, and the guide to UI/UX design agencies lists questions to ask a partner about how research is done. If the product itself is not yet scoped, start with product discovery and roadmapping.
Most Netbase projects are delivered on fixed-price contracts; milestone-based and monthly team retainer models are also offered.
Limits of this guide
- It covers formative testing that finds problems; it does not cover benchmark studies or statistical comparisons.
- Accessibility testing with assistive-technology users needs its own recruiting and checks.
Plan the next step with a Netbase consultant
Frequently asked questions
Five people from one user group are enough to find the most common problems in a design, according to Nielsen Norman Group's research. They are not enough to say how often a problem occurs across all users. Run several small rounds and test the fixes rather than one large round.
A product manager, designer or developer who can stay neutral can facilitate, with a second person taking notes. Avoid the person who designed the screen being tested, because they tend to explain it.
Yes. A clickable prototype is often the cheapest place to find problems, because changing it costs less than changing code. Make sure the task paths work in the prototype.
Next step
If your team needs to know whether people can use what you are building, tell us about the product and the decision you face and we will reply with a proposal for a first test round.
Related services and solutions
AI-enabled software product engineering from discovery to operation Netbase provides software product engineering for companies that need a web, SaaS, mobile or custom platform built and kept running, with AI features and AI-assisted engineering where they pay off. We cover the full lifecycle, from discovery to operation, and the same team stays accountable for architecture, delivery and handover, so nothing important gets lost between phases.
Learn More
Product discovery that validates AI features and ends in a release roadmap Netbase provides product discovery workshops for founders and product leaders to turn a product idea into validated scope, architecture options and a release roadmap. We test the problem with real users, prototype the riskiest flows, check whether planned AI features have the data and accuracy they need, and agree what the first release must prove before committing a build budget.
Learn More
UI/UX design services: research, interfaces and design systems built in code Netbase designs the interfaces of web apps, SaaS products, mobile apps and online stores: user research, flows, wireframes, clickable prototypes and visual design, plus a design system shared by designers and developers. Because UI/UX designers sit inside the delivery team, the same people who design a screen see it built, tested and released.
Learn More
Digital product accelerators for SaaS, mobile and AI products Digital product accelerators are Netbase-built modules and delivery patterns that help founders and product teams launch SaaS, mobile and AI products faster. Instead of rebuilding accounts, billing, workflows and integrations from scratch, you reuse proven components, spend engineering time on what makes your product different, including the AI features buyers now expect, and own the custom code built for you.
Learn More
Work with Netbase remotely
Every Netbase engagement is delivered remotely, so clients in any country and any industry can work with us; Netbase has worked with clients in 80 countries. Contracts and invoices are issued by Netbase JSC, and payment is made by bank transfer to Netbase JSC's company account or PayPal Business. Talk to the CEO, Huy Nguyen (David), on WhatsApp or by email.
Discuss a project
Netbase JSC helps organizations design, build, modernize, and operate digital products and AI-enabled business systems.project@netbasejsc.com
+84 937 869 689
91 Nguyen Chi Thanh, Dong Da, Hanoi, Vietnam
Get in touch
Tell us what you want to build, modernize, or operate.