Best User Testing Services

What a good user testing service includes

User testing gigs are priced per tester or per session. Check that the service states:

• Number of testers and how they are chosen.
• Tester profile: age, experience, devices, location.
• Tasks and questions.
• Recordings: screen, voice and sometimes face.
• Summary of findings with time stamps.
• Devices: phone, tablet or desktop.
• Consent and privacy for recordings.

Ask for a short summary that ranks problems by how often they occurred and how serious they were. Watching every recording is valuable, but a ranked list tells you where to start.

How to set up user testing

1. What you want to learn: one or two main questions.
2. Tester profile matching your customers.
3. Realistic tasks, such as find and buy a gift under a budget.
4. A test link or prototype, with test accounts.
5. Follow-up questions after the tasks.
6. Devices to test on.

Write tasks as goals, not instructions. Asking testers to click the pricing button tells them where to go; asking them to find out how much it costs shows whether they can find it themselves.

Watch at least a few recordings yourself, ideally with your team. Seeing a real person struggle is more convincing than any report, and it often ends long internal debates about what to fix.

Test early and often rather than once at the end. A short round of testing on a rough prototype can change the direction of a design, while testing a finished product mostly finds problems that are expensive to fix. Plan two or three small rounds, fixing problems between each one. Each round costs little, and the improvements build on each other quickly. Keep the recordings, as they are useful for onboarding new team members too.

What drives the price

• number of testers and sessions
• specific tester profiles
• moderated live sessions versus unmoderated
• number of devices
• analysis and reporting depth
• turnaround time

Unmoderated tests, where testers record themselves, are usually cheaper and faster. Moderated sessions, with a facilitator asking follow-up questions, give deeper insight for complex products.

Red flags

• Testers who do not match your audience.
• Leading tasks that tell testers what to do.
• Reports of opinions with no recordings.
• No consent for recordings.
• Promises of statistical proof from a few testers.

Treat findings as problems to solve, not votes. One tester failing a key task is worth investigating, even if others succeeded.

For fixing what testing reveals, see our UX design guide and UX writing guide. For checking that features work, read our QA guide.

Quick pre-order checklist

• I have one or two clear questions to answer.
• Testers match my customers.
• Tasks are realistic goals, not instructions.
• I will receive recordings and a ranked summary.
• Recordings are made with consent.