Maze

What is Maze?

Maze is a user research platform founded in 2017 by Jonathan Widawski and Thomas Mary, with its product taking off the following year. It grew out of the observation that usability testing was slow, expensive and locked to lab settings, when most teams just needed quick answers from real users.

The core flow is unglamorous but powerful: you build a test from a Figma prototype, a live URL or a set of questions, share a link, and participants complete the test whenever they have five minutes. Maze records task success, misclicks, time on task and paths through the flow.

Modern Maze leans heavily on AI. Results get summarized automatically, patterns are highlighted, and you can even generate follow-up questions. The goal is to compress the loop from design to insight so teams test more often instead of shipping on vibes.

Key Features

• Prototype testing with Figma, Adobe XD and Sketch integrations
• Remote moderated and unmoderated tests
• Task-based metrics: success rate, misclick rate, time on task
• AI-powered result summaries and pattern detection
• Heatmaps and click paths for visual analysis
• Participant panel for recruiting testers at scale
• Question-based surveys and card sorting
• Dashboards and reports for sharing findings

How to get started

Create a free Maze account and connect your Figma workspace, or paste in a live URL. From the dashboard, create a new test, choose a prototype, and write a few tasks for participants to complete.

Share the test link with your panel, your team or Maze's built-in participant recruitment. Results stream in as people respond, and the AI summary gives you a first read before you dig into the raw data.

• Create a free account at maze.co
• Connect Figma or add a live URL to test
• Build a test with tasks, questions or card sorting
• Share the link or recruit participants through Maze
• Review the AI summary and click-path data
• Export findings and share reports with your team

Use cases

• Validating a Figma prototype before writing production code
• Catching confusing navigation in a live product
• Comparing two design versions with A/B style tasks
• Running quick surveys with existing customers
• Testing onboarding flows before launch
• Building a continuous research habit for small teams

Pricing and licensing

Maze uses a freemium model with a free plan for individuals that includes basic tests and a limited number of responses. Paid plans add unlimited tests, more participant responses, advanced AI features and team collaboration tools.

Exact plan limits and prices change frequently, so check the official website for the current breakdown before choosing a tier.

Pros and cons

Maze's strengths are speed and accessibility. A test can be live within minutes of finishing a prototype, participants respond on their own time, and the AI summaries make results digestible even for busy teams. It is genuinely the fastest path to feedback for most product teams.

The downsides are that unmoderated tests lack the depth of moderated sessions, and results depend on writing clear tasks. For exploratory research or complex journeys, you will still want interviews or moderated testing alongside Maze.

Alternatives

• UserTesting: moderated and unmoderated testing at enterprise scale
• Optimal Workshop: card sorting, tree testing and first-click tests
• Hotjar: session recordings and heatmaps for live sites
• Lookback: live moderated research with recording
• Attention Insight: AI attention prediction before you test with users