Maze

What is Maze?

The traditional product development cycle in India is fundamentally flawed: Design in Figma -> Hand off to Engineering -> Code for 4 weeks -> Launch -> Find out the users hate it. This results in massive technical debt and wasted capital. Founded in 2018 by Jonathan Widawski and Thomas Meier, Maze addresses this exact issue. Backed by a $40 million Series B funding round led by Felicis Ventures (bringing their total funding to $55 million), the platform introduces "Continuous Product Discovery" before a single line of code is written. Maze is an unmoderated testing platform. You paste your Figma prototype link into Maze, define a task (e.g., "Find the settings menu and change your password"), and Maze generates a unique URL. You WhatsApp that URL to 50 of your actual users.

The users click through the prototype on their own time. Maze quietly runs in the background, generating heatmaps of where they clicked, calculating the average time it took them to complete the task, and measuring the "Misclick Rate." When a PM opens the Maze dashboard the next morning, they have quantitative proof of whether the UX works or fails.

6 Key Features That Matter

1. Figma / Adobe XD Integration: Maze does not require you to rebuild your UI. It seamlessly imports the interactive click-paths directly from Figma. If the designer updates the Figma file, the Maze test updates automatically.
2. Task-Based Heatmaps: The most powerful visualization. When looking at the results for a task, you see a glowing heatmap over your Figma design. If the goal was to click the "Next" button, but the heatmap shows 80% of users clicking a non-interactive image, you instantly know your UI hierarchy is broken.
3. Misclick Rate & Success Metrics: Maze provides hard numbers. It tells you that 65% completed the task via the expected "Direct Path," 20% took an "Indirect Path" (getting lost but eventually finding it), and 15% completely gave up.
4. Post-Task Surveys: After the user completes (or fails) the UI task, Maze can immediately prompt them with a multiple-choice question or open text field (e.g., "On a scale of 1-5, how difficult was that to find?"). This marries quantitative data with qualitative sentiment.
5. Card Sorting & Tree Testing: Crucial for Information Architecture. If you are building a massive e-commerce menu, you can use Maze to run card sorting exercises to see where users naturally expect to find the "Electronics" vs "Home Appliances" categories.
6. Maze Panels: If you don't have an existing user base to send links to, you can pay Maze to recruit testers for you. You can filter by demographics, though the Indian panel specificity can sometimes be broad compared to US panels.

Pricing Breakdown (INR Context)

Maze's pricing is structured around "Responses" (a completed test). Note: Converted at 1 USD = ₹84. Excludes 18% GST.

• Free Tier: ₹0. A very restrictive tier meant for playing around. Allows 1 project and up to 100 responses per month, but severely limits the types of questions and blocks complex logic paths.
• Starter: ~₹8,300/month ($99/mo). Includes 300 responses per month and unblocks critical features like CSV exports and advanced block types. Perfect for a small design pod.
• Team: ~₹25,000/month ($299/mo). Unlocks Card Sorting, Tree Testing, conditional logic branching, and up to 1,200 responses. This is the sweet spot for an established Indian startup.

Who Should Use Maze?

Product Managers and UX/UI Designers at any stage. If you are building a B2B SaaS tool with complex onboarding, or a Fintech app where regulatory flows must be flawless, running a Maze test before committing engineering hours is non-negotiable.

It is especially valuable for remote-first teams. Because the testing is "unmoderated," you don't have to schedule 20 Zoom calls. You just blast the link via email and review the data asynchronously.

Who should NOT use it: Teams looking for deep, qualitative emotional insights. If you need to see a user's facial expression when they see a price, or if you need to ask "Why did you do that?" dynamically based on their actions, you need moderated testing tools (like UserTesting or Dovetail via Zoom calls).

First 5 Setup Steps for Product Teams

A good Maze test requires strict preparation.

1. Prepare the Figma File: Maze only works if your Figma file has interactive "noodles" (prototyping links) connected. Ensure all paths to the goal are clickable.
2. Define the Mission: In Maze, create a new Project. Write a clear, scenario-based mission prompt. (Bad: "Click the profile button." Good: "You want to update your billing address. Show me how you would do that.")
3. Set the Expected Path: Click through your own prototype within Maze to define the "happy path." Maze will measure user deviations against this baseline.
4. Add Follow-up Questions: Add an Opinion Scale block immediately after the task to capture subjective difficulty.
5. Distribute the Link: Generate the URL. For Indian markets, the highest conversion strategy is dropping the link into a targeted WhatsApp group or sending an in-app Intercom message to a specific cohort offering a small Amazon voucher for completion.

Top Alternatives in the Indian Market

• UserTesting: The heavyweight champion of moderated and unmoderated video testing. It records the user's screen and their voice while they speak their thoughts aloud. It yields incredibly deep qualitative insights but is drastically more expensive than Maze.
• Lyssna (formerly UsabilityHub): A direct competitor to Maze. Lyssna is excellent for quick "Five Second Tests" (flashing a design on screen and asking what the user remembers) and first-click testing, often slightly cheaper than Maze for basic needs.
• Hotjar: While Hotjar also provides heatmaps, it is fundamentally different. Hotjar provides heatmaps on your live coded website. Maze provides heatmaps on your un-coded Figma prototypes. You should ideally use both.