Usero Journal
User Testing Tool: How to Pick One (and When You Do Not Need One)
Before you buy a user testing tool, answer two questions: do you need a researcher on the call, and do you need the tool to find you participants? Your answers pick the tool, and getting them wrong is how teams end up paying enterprise prices for a feature they will use twice a year.
A user testing tool records a real person using your product so you can watch where they struggle. Almost every tool does that core job competently. They diverge on two expensive axes: whether the session is moderated (a researcher live on the call) or unmoderated (the participant works through a task list alone), and whether the tool recruits participants for you from a paid panel or expects you to bring your own.
Those two axes are where the money is. The recruited panel and the moderated live mode are the two features that push a plan from tens of dollars a month into enterprise contracts. If you do not need either, paying for both is how a small team ends up with a research platform it opens once a quarter.
The two questions that pick the tool
Walk these in order and most of the shortlist falls away.
- Do you need a researcher on the call? If you are exploring a fuzzy problem and need to ask follow-up questions in the moment, you want moderated. That is a live interview, and the tool's job is scheduling, recording, and a notetaking surface. Lookback and the big platforms do this well. The cost is time: moderated sessions are slow to schedule and do not scale past a handful.
- Do you already have users to ask? If yes, you do not need to pay for a panel, and that single fact removes the most expensive line on most quotes. You write a task list, send a link to your own users, and read the recordings. If you do not have users yet, or you need a specific demographic of strangers, a recruited panel earns its keep and you should pay for one.
Most small product teams answer "no, I can watch the recording" to the first and "yes, I have users" to the second. That combination, unmoderated and bring-your-own-participants, is the cheapest corner of the category and the one the big platforms serve worst, because their pricing is built around the panel and the live mode you just opted out of.
A replay shows you the clicks. A user test shows you the clicks and the sentence the person said while they made them. The sentence is usually the bug.
User Testing Tools Compared
Starting prices are monthly USD where published, checked early 2026. Panel recruitment is usually billed per participant on top of the plan. Confirm on each vendor site, these change.
| Tool | Starts at | Moderated | Built-in panel | The job it fits |
|---|---|---|---|---|
| UserTesting | Custom (enterprise) | Yes | Yes (large) | Enterprise research at scale |
| Maze | ~$99 | Limited | Add-on | Prototype + survey testing |
| Lookback | ~$25/seat | Yes | No (bring your own) | Moderated interviews |
| Useberry | ~$67 | Limited | Add-on | Unmoderated prototype tests |
| Usero | $0, paid from $19 | No (unmoderated) | No (bring your own) | Unmoderated test tied to your feedback |
The tools worth shortlisting
UserTesting
custom (enterprise)The panel heavyweight
The category's biggest name, built around a large recruited panel and a research-ops workflow. If your job is to test with strangers who match a target audience and you need dozens of sessions a month, this is what it is for.
Pricing is quote-only and lands in the five figures a year. For a two-person team that wants to watch five of its own users, it is the wrong shape and the wrong budget.
Research and CX teams that need recruited strangers in specific demographics, on demand, at volume.
Maze
from ~$99/moThe prototype-testing pick
Maze leans into testing prototypes and structured tasks, with quantitative metrics (paths, misclicks, time on task) rather than long qualitative recordings. Strong when the thing you are testing is a clickable mockup.
Recruitment is an add-on, and the quantitative bias means it is less suited to "watch someone talk through a live product and tell me what confused them."
Designers validating Figma prototypes and surveys before a line of code is written.
Lookback
from ~$25/seatThe moderated-interview pick
Lookback is built for the live call: a researcher, a participant, screen sharing, and a notetaking timeline, with recordings kept afterward. If question one was "yes, I need a researcher on the call," this is the lane.
It does not recruit for you, which is a feature if you have users and a gap if you do not. Moderated by design, so it does not help if you want a participant to run a task list unattended.
Teams running live, researcher-present interviews who bring their own participants.
Usero
free, paid from $19/moUnmoderated test tied to your feedback
The tool I work on, so weigh that. Usero user testing is unmoderated and link-shared. You write a task list, set an optional reward, and get an unguessable link you send to your own users. The participant consents, then Usero records the screen with rrweb and the microphone in parallel while they work, and stitches the replay, audio, transcript, and task completions together in the dashboard.
Two honest limits. There is no moderated mode, the participant runs the tasks alone. And the reward is a manual ledger, not a payout rail: you set an amount and mark a session paid, but Usero does not move the money, you settle up off-platform. So it is not a UserTesting replacement and it does not recruit for you.
The reason it sits in a feedback tool at all is the wedge. A test session is feedback like any other in Usero, so a usability problem clusters with the same complaint arriving from your widget and inbox, and from a cluster you can open a draft GitHub pull request with a first pass at the fix. You review it and merge it yourself, nothing auto-merges. The finding does not die in a research folder, it takes the same path to a diff as a bug report.
Small product teams with real users who want screen-and-voice tests without an enterprise contract, and want the findings to feed their roadmap.
When you do not need a tool at all
The honest answer for a lot of early teams is to run the test by hand first. A user testing tool earns its keep when you are running rounds often enough that scheduling, recording, and storage become the bottleneck. Below that, the tool is overhead.
- You have never watched a user once. Get five real users on a call, share your screen, give them three tasks, and shut up while they try. You will learn more in an hour than a tool's onboarding will teach you, and you will know what you actually want to record next time.
- You are testing a static prototype, not a live product. A clickable Figma and a quick path through it may answer your question without recording anyone, especially for a single screen.
- You need legal sign-off on recording and have not got it. Screen and voice recording of users is regulated. If you cannot get consent handled and sensitive fields masked, do not record yet.
If you have users and want the findings to ship
If you answered "I can watch the recording" and "I have my own users," and you want a usability finding to flow into the same place as your bug reports rather than a research silo, that is the corner Usero is built for. User testing is on the paid plan, recording is screen plus voice, and a finding can cluster and open a GitHub pull request you review and merge. Spin up a workspace, or read the user testing feature page and the setup docs for how a test is built and shared.
Related reading
- Session Replay Tool: How to Pick the Right OneRead
- How to Collect User Feedback In-App: 7 Methods ComparedRead
- The Loud 1% Are Not Your UsersRead
- How to Turn User Feedback Into Shipped CodeRead
Frequently Asked Questions
What is a user testing tool?
A user testing tool records a real person using your product so you can see where they struggle. The category splits two ways: moderated tools run a live interview with a researcher present, and unmoderated tools hand the participant a task list and record them working through it alone. Most also offer recruitment (a paid panel of participants) and screen plus voice recording, though the cheaper tools strip one or both out.
How much does a user testing tool cost?
It ranges widely as of early 2026, and the recruited panel is what drives the price. UserTesting is enterprise and quote-only, usually five figures a year. Maze starts around 99 dollars a month for the team plan. Lookback is around 25 dollars a month per seat. Useberry and PlaybookUX sit in the tens-of-dollars range. Recruiting participants through a panel is billed per session on top, often 30 to 120 dollars a participant. Bringing your own participants removes that line entirely. Confirm current pricing on each vendor site, these move.
What is the difference between moderated and unmoderated user testing?
In moderated testing a researcher is on a live call with the participant, asking follow-up questions and steering the session. It is richer but slow to schedule and expensive to run at volume. In unmoderated testing the participant gets a task list and works through it on their own while their screen and voice record, and you watch the recording afterward. Unmoderated is faster and cheaper per session, at the cost of not being able to ask a follow-up in the moment.
Do I need to pay for a participant panel?
Only if you do not already have users to ask. A panel is worth paying for when you need strangers in a specific demographic and you need them today. If you have a product with real users, recruiting from your own base (an email, a DM, a message to a power user) is free, and the participants already know your domain, which usually produces sharper feedback than a paid stranger seeing it cold.
Can I run a user test for free?
Yes, if you bring your own participants. Get five real users on a call, give them a short task list, and watch them work, that is a usability test and it costs nothing but your time. The Nielsen Norman rule of thumb is that five users surface most of the findings a single round will give you. A tool helps you do it at a distance and keep the recording, but the method does not require one.
How many users do I need to test with?
Five per round is the long-standing rule of thumb from Nielsen Norman, because the first handful of participants hit most of the same problems and each additional one finds less that is new. Run five, fix what you found, then run another five against the fix. More than five in a single round is usually better spent on a second round after changes.
Continue reading
Feedback API: Collect Feedback From Any Platform With One HTTP POST
A feedback API lets you collect feedback from Swift, Python, a server, or a CLI without a JavaScript widget. How a good feedback REST API works, with copy-paste curl, Swift, and Python examples, plus how reports cluster and can become a GitHub PR.
7 min read
How to Import GitHub Issues Into a Feedback Tool (and Why)
GitHub issues pile up and never dedupe. Import them into a feedback tool that clusters by meaning across every source, counts what recurs, and can open a GitHub PR you review and merge. A practical guide.
7 min read
Session Replay Tool: How to Pick the Right One (2026)
Picking a session replay tool comes down to one question: what do you actually need to watch? Compare FullStory, LogRocket, PostHog, Hotjar, Clarity, and Usero by the job they fit, with real pricing and an honest when-to-skip-it.
9 min read
Build a feedback loop your team actually uses
Usero collects, clusters, and turns user feedback into shipped fixes.
Get started free