Figma prototypes meet Maze: test designs before you write code

Figma and Maze have launched a new product integration that lets you import a Figma prototype into Maze with a simple copy and paste of the prototype link. The goal is to move user testing earlier in the design process, so teams can gather feedback on flows before committing to development work.

Maze is a user research platform that provides UX insights from real users, supporting both 1:1 interviews and large-scale tests with up to 50,000 users. Teams at companies like Ring, Revolut, and IBM use it to run quantitative research across multiple flows without adding heavy research overhead. With the Figma integration, prototypes can be tested directly, skipping the usual wait for a development environment or A/B test setup.

For Figma users who want to try the integration, Maze is offering 30% off any paid plan for the first three months with the promo code MAZE_LOVES_FIGMA.

Five research habits that improve usability tests

Regardless of whether your team has dedicated research staff or relies on occasional testing, these practices can help you get more meaningful results.

Define your success criteria first

Set a specific, quantifiable target before building a prototype. For example, a test might be considered successful only if 80% of participants complete the task and fewer than 2% misclick. Goals will vary by test, but they should always be defined up front.

Run a pilot before the real session

Usability tests benefit from the same caution that product teams apply to beta programs or gradual rollouts. Share the test with internal teammates, friends, or family first. Running them through the test as if it were the real thing helps ensure tasks are clear, the prototype behaves correctly, and the goals are meaningful, especially when you plan to test with a large population.

Include both experienced users and newcomers

Existing users are often eager participants. They know the product, are likely to adopt new features early, and can provide vocal, detailed feedback. Target the customer segments most likely to use the new feature to find engaged subjects.

But don't rely solely on seasoned users. Their familiarity with your product can bias the results. Testing with non-users brings a fresh perspective that helps you spot assumptions your existing audience may overlook.

Keep tasks free of leading language

Avoid phrasing that tells users where to click or what to do next. Words like "click here," "go to," and "navigate to" coach participants instead of testing whether the interface is intuitive. A task like "Click the Signup button to create an account and go to the Feed page to upload images" gives away the answer. Instead, phrase tasks around the end goal the user should achieve, without mentioning internal feature names.

Analyze results together as a team

The hardest part of synthesizing test results is often not extracting insights but gaining organization-wide buy-in on findings and next steps. Collaborating on analysis, much like collaborating on a design file, lets the whole product team discuss interpretations and decide on actions together. Maze supports this with instantly generated reports and a custom URL for sharing results with anyone.

For further details on the integration, refer to the getting started documentation. Feedback on the integration can be directed to the Maze team on Twitter at @mazedesignHQ.