ComparisonsNovember 4, 2026 · 6 min read

Maze vs DASHUND: two different ways to test a Figma prototype

Maze runs studies with missions and questions; DASHUND watches free, unguided use of a Figma prototype. Two useful approaches that work well together.

By Thomas Cescousse, product designer and maker of DASHUND

Maze and DASHUND both test Figma prototypes, but they ask different questions. Maze runs a study: missions, questions, often a participant panel, and it measures whether people complete what you asked them to do. DASHUND gives no instructions at all: you share a link, people use the prototype like a real product, and you see their paths. The first validates a specific hypothesis; the second shows what happens when nobody is guiding. Both are useful, and they fit together nicely.

Why compare Maze and DASHUND?

I'm a product designer, and I built DASHUND because I wanted to know what people actually did with my prototypes. When someone says "test a Figma prototype," Maze is usually the first name that comes up, and for good reason: it's the reference in the category. Designers often ask me, "So it's like Maze?" The honest answer is "No, it's a different approach." This post explains the difference, without caricature, using facts I was able to check.

How does Maze work?

Maze is a continuous user research platform. As of November 4, 2026, it offers:

• unmoderated tests with missions ("find a way to…") and questions, plus automatically generated reports;
• interviews, with an AI moderator on the Enterprise plan, and card sorting;
• a recruitment panel advertised at more than 5 million participants;
• one-click Figma prototype import, including interactive components and password-protected prototypes.

Results cover mission success, usability scores, path analysis and heatmaps. Maze also displays SOC 2 Type II certification. It's a complete tool, built for product and research teams that run studies on a regular basis.

How does DASHUND work?

DASHUND is analytics for Figma prototypes. You publish your prototype from the DASHUND plugin for Figma, and that version is frozen at publication. Then you create one link per audience (your teammates, a group of customers, a LinkedIn post) and share it.

Whoever opens the link sees the prototype and nothing else. No account, no install, no task, no question, no webcam or mic. They use it the way they'd use the product.

On your side, you see every session step by step, the paths people took (the Flow, as a Sankey diagram, and the Sitemap), the completion rate toward the screen you picked as a goal, drop-offs, and heatmaps that also show clicks outside the clickable areas. You can compare two versions or two audiences, and run an A/B test from a single link. The full list is on the Features page.

DASHUND doesn't record video, run surveys, host interviews or recruit testers. That's a choice, not a gap.

Missions or free use: what's the difference?

People behave differently when you give them a mission. If the prompt says "change your shipping address," it already tells them where to go and what matters. The test is great at measuring whether people can find the feature. It says less about whether someone, left alone, would have gone looking for it.

Free use, on the other hand, shows what people do on their own: where they start, what they look at, where they loop, where they stop. It's weaker at answering a very narrow question, because nothing guarantees testers will reach the screen you care about.

Neither method is "the right one." One is closer to an exam, the other to an observation. The right pick depends on what you're trying to learn.

Maze and DASHUND side by side

DASHUND's Free plan includes 1 workspace, 2 live prototypes, 100 sessions a month and 30 days of history, and it doesn't expire. The other plans are on the Pricing page. Maze's pricing can change, so check its official page before you decide.

When should you pick Maze?

Maze is the right call when:

• you have a specific hypothesis to validate with a task ("can people find the export button?");
• you want to ask questions before, during or after the test;
• you need to recruit participants you don't know, with screening criteria;
• you run several methods (tests, interviews, card sorting) and want them in one place;
• your team has a research budget and formal security requirements.

When is DASHUND a better fit?

DASHUND fits better when:

• you want natural behavior, without a prompt shaping the path;
• you already have audiences (teammates, customers, a community) and want lots of sessions rather than a study;
• you want to compare two versions of your design, or two audiences, or run an A/B test from one link;
• you work solo or on a small team and want a low, public price;
• you don't want to collect any personal data about your testers.

How do you use Maze and DASHUND together?

Often, that's the best answer. A sequence that works well:

1. Publish the prototype with DASHUND and share the link widely. Look at the Flow and the heatmaps: where do people get lost,

where do they drop off?

1. For the two or three screens you spotted, write a Maze study with targeted tasks and questions to understand why things

go wrong.

1. Fix the design, publish the new version in DASHUND, and compare it with the first one.

DASHUND shows the what, at volume; Maze helps you dig into the why. If you're just getting started, the guide How to test a Figma prototype walks through these steps in detail.

FAQ

Maze or DASHUND: which one should I choose?

Choose Maze to validate specific tasks, ask questions or recruit a panel. Choose DASHUND to observe free use across many sessions, compare versions and run A/B tests, with no personal data collected. Many teams get the most out of using both.

Does DASHUND give testers tasks?

No. Testers only see the prototype, with no task, no question and no account. You pick a goal screen to measure completion, but the tester never knows what it is.

Can I test the same Figma prototype with Maze and DASHUND?

Yes. Maze imports the prototype from Figma, and DASHUND publishes it from its plugin. Both read the same file, each with its own method.

Does DASHUND recruit testers like Maze?

No, DASHUND has no panel. You share your links with your own audiences: teammates, customers, your community, social networks.

How much do Maze and DASHUND cost?

As of November 4, 2026, Maze's pricing page lists a Free plan and an Enterprise plan with a custom quote. DASHUND has a permanent Free plan, then Solo at $14, Pro at $29 and Business at $79 per month.