Maze / Lookback / UserTesting
Unmoderated and moderated research platforms that let UX teams test prototypes, gather usability data, and watch real users attempt real tasks.
What it is
About Maze / Lookback / UserTesting
Maze, Lookback, and UserTesting represent three different modes of user research. Maze specialises in unmoderated usability testing linked directly to Figma prototypes: you publish a prototype, write a task script, and Maze recruits participants (or you share a link to your own panel), then automatically collects task completion rates, time-on-task, misclick maps, and follow-up survey responses — all without a researcher needing to be present. The quantitative output makes it fast to compare two design variants or track usability improvements across iterations. Lookback is a moderated interview and usability-testing platform: researchers conduct live sessions via video, sharing screen control with participants, and the platform records, transcribes, and timestamps the session for later analysis and highlight-reel creation. UserTesting is the largest enterprise platform, offering both moderated and unmoderated testing with a pool of over one million vetted testers filtered by demographic, location, job role, and behaviour — making it the default for teams that need rapid, demographically targeted feedback on a tight deadline. For a UX Designer, these platforms replace the most time-consuming parts of the research cycle: recruiting participants, scheduling sessions, transcribing recordings, and compiling results into a shareable report. A Maze test linked to a Figma prototype can be set up in two hours and return quantitative results from twenty participants within twenty-four hours — compressing a research cycle that would previously take two weeks of recruitment and moderated sessions into a single day. The research output — task completion rates, heatmaps of misclicks, and verbatim participant feedback — goes directly into the design iteration cycle rather than sitting in a slide deck.
What you can do with it
Capabilities
Link a Figma prototype of a new checkout flow to Maze and write three tasks that mirror real user goals — "find a product in size M and add it to your basket", "apply a discount code", "complete the purchase" — then share the test link to 30 participants and analyse the resulting heatmaps of where users clicked versus where the design expected them to click
Run a five-second test in Maze showing participants the hero section of a redesigned homepage for five seconds, then asking them what they remember and what the company does — quantifying whether the value proposition is landing before any usability testing of the flow
Use Lookback to conduct a moderated remote session with a user attempting to navigate a new information architecture — sharing your screen to display a prototype, observing where they pause and express confusion, then asking follow-up questions in real time and timestamping the moments to build a highlight reel
Export Maze task completion data to a spreadsheet and calculate the System Usability Scale (SUS) score from the post-task questionnaire responses — producing a single headline number that tracks design quality across prototype iterations
Use UserTesting's panel filters to recruit ten participants matching a specific demographic — for example, UK-based adults over 55 who use online banking weekly — and run an unmoderated test of a financial services redesign to identify whether accessibility assumptions in the design hold up with the actual target audience
How to learn it
Learning Resources
Maze University (maze.co/resources) — free; short courses on unmoderated test design, writing good task scripts, and interpreting misclick heatmaps; the "How to run your first usability test" guide is the most practical starting point
Lookback's guide to moderated usability testing (lookback.com/resources) — free; covers session facilitation techniques, how to avoid leading participants, and how to tag and clip recordings into insight reels
Nielsen Norman Group's free articles on usability testing methodology (nngroup.com) — the gold standard of UX research practice; the articles on task writing and sample size selection are essential reading before running any test
Practise writing task scripts before booking participants — a poorly written task invalidates the data. The golden rule: write tasks as goal-oriented actions ("book a flight from London to Edinburgh next Friday") never as instructions that reveal the interface ("click the Search button and fill in the date fields")
Pro Tip
The most common mistake in unmoderated testing is writing tasks that are too prescriptive. If your task says "using the navigation menu, find the returns policy", you have eliminated the most valuable part of the test — discovering that users do not look in the navigation menu at all. Write tasks that describe the goal, not the path: "you want to return a jacket you bought last week. Find out how to do this." Observation of the path they take is the data.
Skills that use this tool
Roles that use this tool
Alternatives