top of page
Business & Enterprise
MarTech
Advertising Technology Platform
Maze
Democratized product research by turning design prototypes into self-serve, measurable user tests, despite established research agencies and moderated-testing tools’ slower, expert-led model.
1
MODEL
BUSINESS MODEL
SaaS
- Founded in 2018 by Jonathan Widawski and Thomas Mary, software developers who saw product research as slow and risky. [95][96]
- Launched on Product Hunt in 2018 as a prototype-testing tool for InVision designs. [92][94]
- Built integrations with Figma, Sketch, InVision, Marvel and Adobe XD so teams could turn existing designs into tests. [92][93]
- Added Maze Discovery in 2021, enabling research before a prototype existed. [94]
- Expanded into participant recruitment, moderated interviews, surveys, automated reports and AI-powered analysis. [97][96]
1. Identify a specialized expert workflow that blocks a broader team from making faster decisions.
2. Build the smallest self-serve tool that turns an existing team artifact into a decision-ready output.
3. Integrate with the tools where the input already lives, rather than asking users to recreate work.
4. Let non-experts run the first workflow, then add governance and deeper research methods as usage matures.
5. Expand from validation into the full research lifecycle only after the core workflow becomes habitual.
6. Avoid replacing expert judgment entirely; package expertise as templates, metrics and interpretation.
DISTRIBUTION MODEL
Maze first built a tool that converted Figma, InVision, Sketch, Marvel and Adobe XD prototypes into browser-based unmoderated tests. It delayed broader research capabilities until that core workflow proved demand: by 2021 it reached $1.5 million ARR, 40,000 companies and 600% year-over-year MRR growth. [92][93][94]
HOW THEY BUILT IT
HOW TO ARCHITECT IT
Product-Led Growth, Self-Serve / Website, Community-Led, Product Integrations
HOW THEY OPERATIONALIZED
HOW TO REPLICATE WHAT WORKED
dm
- Product Hunt launch reached designers and product builders already seeking faster validation tools. [92][94]
- Self-serve signup let teams convert a prototype into a test without engaging a research agency. [92][98]
- Figma, Sketch, InVision, Marvel and Adobe XD integrations placed Maze directly in the designer’s existing workflow. [92][93]
- Shareable test links reached participants and stakeholders, creating exposure among product, design and marketing teams. [93][98]
What worked: integrating with the customer’s existing creation tool removed workflow switching costs, while self-serve testing made a specialist service accessible to non-specialists.
Trap if copied blindly: community-led launches create attention but not retention if the tool solves only a one-off validation task; without recurring research cadence or team collaboration, activation does not become revenue.
| PATTERNS OF THIS MODEL
PATTERNS IN EXPERT-WORKFLOW DEMOCRATIZATION SAAS:
1. THE INPUT ARTIFACT IS THE DISTRIBUTION HOOK. When the product consumes a file or design the customer already has, adoption begins without data migration or setup.
2. SELF-SERVE TESTING SHIFTS THE BUYER FROM SPECIALIST TO TEAM. Designers, product managers and marketers can run tests, expanding usage beyond a central research function.
3. AUTOMATED OUTPUT CREATES A REPEATABLE DECISION LOOP. Metrics, drop-off analysis and reports make research consumable by non-researchers.
THE WEAKNESS: If the platform only automates unmoderated tests, it can be perceived as a shallow validation tool rather than a system of record for customer understanding.
What companies with this model reveal
| OPPORTUNITY INTELLIGENCE
GOLDMINE 1 — RESEARCH AS A PRODUCT-DECISION SYSTEM. The opportunity is not more tests but a connected record of hypotheses, evidence, decisions and outcomes across product cycles. Maze’s expansion from prototype testing into discovery and AI analysis points toward this gap. [97][94][96]
GOLDMINE 2 — PARTICIPANT SUPPLY AS A STRUCTURAL ADVANTAGE. Recruiting the right users is often harder than building the test; a reliable, segmented participant network can become a durable differentiator. Maze added recruitment alongside testing. [97][96]
GOLDMINE 3 — AI-ASSISTED RESEARCH GOVERNANCE. AI can summarize sessions and flag patterns, but enterprises need consent, bias controls, data security and traceable evidence. [97][96]
THE PIT — UNMODERATED TESTING WITHOUT INTERPRETATION. Click paths and completion rates can mislead teams if they are not paired with context, qualitative evidence and research design.
THE SECOND PIT — PROTOTYPE DEPENDENCY. A tool that only tests polished designs arrives too late to influence foundational product decisions. Maze’s Discovery launch addressed this risk. [94]
MOVE WITH CAUTION — Do not position solely as “Figma testing”; design-platform bundling or API changes can erode the wedge. Build value in research operations, participant quality and decision memory. [92][94]
Untapped Business Model / Gaps / Goldmines / Pits
Patterns & Insights
2
MARKET
mkt mt es
Maze reframed fragmented usability testing and surveys into “rapid testing” and later broader product research. Instead of asking teams to compare moderated-research agencies with survey tools, it made continuous, self-serve user testing part of the product-development workflow. [98][99][94]
MARKET TYPE
Emerging Market
ENTRY STRATEGY
Platform Entry
WHY THEY WON
Product research was transitioning from occasional, agency-led studies to continuous, in-team experimentation. The structural opening was the rise of collaborative design tools such as Figma, which made prototypes easy to produce but left validation slow and expert-dependent. Maze exploited that gap by making testing autonomous, remote and measurable. Established research providers could add unmoderated testing, but their service-heavy economics and expert-led positioning made rapid self-serve adoption less central. The lesson is that category creation often begins by attaching a new workflow to a tool that has already changed how work is produced.
EXECUTION
Maze entered through existing design platforms—initially InVision, then Figma, Sketch, Marvel and Adobe XD—rather than asking teams to abandon their design tools. It converted prototypes into shareable browser-based tests, using the designer’s current workflow as the entry point. [92][93][94]
FOOTHOLD STRATEGY
Wedge Strategy
The initial wedge was designers and product teams who needed to validate a prototype before development. They were reachable through Product Hunt and design communities, could adopt through self-serve signup, and gained immediate value by turning an existing prototype into a test link. [92][94][100] The expansion path was product managers, marketers and research teams needing discovery, recruitment, moderated interviews and enterprise research operations. [97][96]
GROWTH CAMPAIGN
CAMPAIGNS THAT WORKED
SO WHAT: The Figma integration and Product Hunt launch connected Maze to an active creator community at the exact moment a prototype existed.
KEY LEARNING: Category creation is easier when the new workflow attaches to an existing high-frequency artifact rather than asking users to change tools.
HOW TO USE: Identify the artifact your target user already creates, then design the fastest possible path from that artifact to a decision-ready output.
KEY LEARNING
THE STANDARD: Emerging markets reward vendors that define the workflow before buyers have fixed comparison criteria.
RULE 1 — BUYERS NEED A NEW DECISION HABIT: The vendor must show when and why to test, not merely provide a testing interface.
RULE 2 — PRICING STARTS SIMPLE AND EXPANDS WITH SCOPE: Self-serve plans fit teams, while recruitment, security and enterprise research require higher-tier pricing. [101][102]
RULE 3 — SUCCESSFUL ENTRANTS ATTACH TO AN EXISTING WORKFLOW: Design platforms, collaboration tools or developer systems can provide the initial distribution surface.
RULE 4 — FAILURE OCCURS WHEN THE CATEGORY REMAINS A FEATURE: If the workflow is easily added by a design or survey platform, standalone differentiation erodes.
MARKET TYPE: Emerging Market (category).
gc
Market Context
| MARKET INTELLIGENCE
PATTERNS IN COMMUNITY-LAUNCHED PLG:
1. COMMUNITY LAUNCHES CREATE EARLY ADVOCATES. Product Hunt and practitioner communities reach users who are actively seeking better workflows. [92][94]
2. INTEGRATION LED GROWTH REDUCES ACTIVATION FRICTION. The product becomes useful because it consumes work the user has already produced. [92][93]
3. SHAREABLE OUTPUTS EXTEND REACH. Test links expose the product to participants and stakeholders beyond the initial account. [93][98]
4. THE WEAKNESS: Community attention decays quickly unless the product creates a recurring reason to return, such as continuous research operations or team collaboration.
| MARKET ENTRY PLAYBOOK
THE STANDARD: Platform entry is appropriate when a host platform already contains the target user, the input artifact and a natural moment to trigger the new workflow.
RULE 1 — CHOOSE A HOST WITH A RECURRING CREATION LOOP: The platform should generate the artifact repeatedly, not occasionally.
RULE 2 — MINIMIZE INPUT REWORK: The product should consume the existing artifact directly.
RULE 3 — BUILD INDEPENDENCE BEFORE PLATFORM RISK MATERIALIZES: Expand the workflow beyond one host and create direct demand.
RULE 4 — REJECT THE APPROACH when the host can bundle the capability or restrict access without a compensating data, network or workflow advantage. Loom entered through the browser; Maze entered through design tools; both faced the same platform-dependency constraint.
How to enter
| FOOTHOLD STRATEGY PLAYBOOK
THE STANDARD: A wedge strategy succeeds when one narrow validation job creates immediate value and exposes adjacent research needs.
RULE 1 — SEGMENT SELECTION: Target teams with frequent design or product decisions and an existing artifact ready to test.
RULE 2 — EXPANSION: Move upstream to discovery and downstream to analysis only after prototype validation becomes habitual.
RULE 3 — FAILURE: A wedge fails when testing is episodic and does not create a recurring research cadence.
RULE 4 — MARKET-SPECIFIC VARIATION: In product research, the wedge should connect to the product-development calendar—discovery, prototype, launch and iteration—not merely to one design review.
How to get the first strong position
MARKET PATTERNS & PLAYBOOK
3
MONEY
money rev pri
Maze used a free entry path and per-user subscription plans, converting individual designers and product teams into paid accounts as testing volume, collaboration and research needs grew. By 2021 it had reached $1.5 million ARR and 40,000 companies; by 2026 it reported 60,000+ teams. [93][102][103]
REVENUE MODEL
Freemium, Subscription
PRICING MODEL
Freemium, Tiered Pricing, Custom Enterprise Pricing
WHY THEY WON
Customers pay recurring per-user subscriptions for testing capacity, collaboration, participant recruitment, advanced analysis and security. Revenue expands as research becomes a recurring team workflow rather than an occasional design validation; recruitment and enterprise requirements increase account value but also cost-to-serve. [101][102][96]
Maze offers a free version, trial access and paid plans priced per user; enterprise requirements are handled through custom pricing. The free tier enables prototype testing, while paid plans add capacity, collaboration, recruitment and advanced research capabilities. [102][104]
TARGET AUDIENCE
Designers and product teams (free/lower tiers; rapid prototype validation); product managers and researchers (paid tiers; recurring discovery and usability research); marketing teams (paid tiers; message and campaign testing); enterprises with compliance needs (custom; security, recruitment and research governance). [98][97][96]
CUSTOMER BUYING BEHAVIOUR
A designer or product manager has a prototype or product question, creates a Maze test from an existing design file, and shares it with users. When testing becomes recurring or requires participant recruitment, collaboration and reporting, the team evaluates paid plans and may involve procurement for enterprise security. [92][93][96]
| PRICING INTELLIGENCE
What makes this model effective & make customers pay
Customers are not paying for surveys; they are paying to reduce the risk of building the wrong product.
RULE 1 — PRICE THE DECISION, NOT THE TEST. A test is valuable when it prevents wasted development, poor conversion or a failed launch. [98][95]
RULE 2 — LET NON-EXPERTS START. Self-serve testing expands the buyer base beyond trained researchers. [98][99]
RULE 3 — CHARGE WHEN RESEARCH BECOMES OPERATIONAL. Recruitment, collaboration, reporting and governance justify recurring spend. [97][96]
RULE 4 — WATCH THE VENDOR COST TRAP. Participant incentives, recruitment quality and enterprise compliance can raise variable cost faster than subscription revenue.
The buyer is purchasing evidence for product decisions, measured in tests run, participants reached and decisions informed. [93][98][96]
PRICE & REVENUE
| Revenue Risk - The biggest threat to revenue stability
Per-user pricing can contract when design or product teams reduce headcount or consolidate tools. [102]
Figma, survey platforms and product-analytics vendors can bundle basic testing or research workflows, commoditizing simple prototype validation. [92][93]
Participant recruitment creates variable cost and quality risk; poor participant matching undermines trust in results. [97][96]
Enterprise research requires GDPR compliance, security and consent management, increasing product and operational burden. [96]
Research demand can be episodic, tied to product launches or redesigns, creating churn if the platform does not become part of continuous discovery. [98][94]
Funding/valuation: Maze raised $60 million by 2022 and later reported $93 million total raised; standalone ARR beyond the disclosed $1.5 million in 2021 is not publicly verified. [93][95][97]
Where the model can break
4
MOTION
motion ge cs
Maze built a repeatable product-led and community-led acquisition engine: Product Hunt launch, design-tool integrations and self-serve testing helped it grow from a prototype-testing tool to 40,000 companies by 2021 and 60,000+ teams by 2026. [92][93][94][103]
GROWTH EXPANSION MODEL
Product Line Expansion, Ecosystem Expansion
HOW THEY EXPAND
Maze expanded from prototype testing into Maze Discovery, participant recruitment, moderated interviews, surveys, automated reporting and AI insights. It expanded its ecosystem through design-tool integrations and strategic investors including Atlassian, Zoom and HubSpot. [95][97][94][96]
COMPETITIVE STRATEGY
Differentiation
HOW THEY COMPETE
Against moderated research platforms and research agencies, Maze differentiated on speed, self-service and quantitative usability metrics. Against survey tools, it differentiated through prototype-based behavioral testing and product-development workflow integration. [92][98][99]
GROWTH ENGINE
GTM
Content Flywheel
A designer imports a prototype from Figma or another design tool; Maze converts it into a test; the team shares the test with participants; automated metrics inform a product decision; recurring product cycles bring the team back to test again. The repeat trigger is each design iteration or product question. The engine weakens if teams test only before major launches, if participant supply is unreliable, or if insights do not visibly influence decisions.
Maze uses a primarily product-led, community-supported GTM. Design and product teams discover Maze through Product Hunt, communities and design-tool ecosystems; self-serve signup lets them convert an existing prototype into a test without sales involvement. Integrations reduce activation friction, while shareable test links extend exposure to participants and stakeholders. As research becomes continuous and enterprise requirements emerge, sales supports larger accounts needing recruitment, security and governance. The primary motion is PLG; the supporting motion is sales-assisted expansion. This works when the input artifact is digital and recurring; it is less effective where research requires complex fieldwork or highly regulated participant recruitment. [92][98][97][96]
SUSTAINING MOATS
High Customer Lock-In
moat
The moat is the accumulated research library: hypotheses, tests, participant insights, decisions and product-learning history. As teams run continuous research in Maze, the platform becomes the memory of what the organization has already learned about its users.
| MOAT INTELLIGENCE
THE STANDARD: A product ecosystem lasts when accumulated research context makes future decisions faster than starting elsewhere.
RULE 1 — RESEARCH HISTORY CREATES ORGANIZATIONAL MEMORY. Prior tests, segments and insights reduce repeated learning and make the platform harder to replace. [97][96]
RULE 2 — INTEGRATIONS CREATE WORKFLOW DEPENDENCE. Connections to Figma and other design tools embed Maze in the product-development process. [92][93]
RULE 3 — PARTICIPANT AND INSIGHT LAYERS ADD STRUCTURE. Recruitment, moderated research and AI analysis extend value beyond one-off prototype tests. [97][96]
THE SIGNAL: Maze’s growth from 40,000 companies in 2021 to 60,000+ teams by 2026 indicates recurring adoption; the moat does not protect against a design platform bundling adequate testing or against low-quality participant supply. [93][103]
Why this company remains defensible
ARR & TAKEAWAY
ARR Journey - what to do at each stage
PRE-$1M ARR — VALIDATE PROTOTYPE TESTING. In 2018, Maze launched on Product Hunt as a tool for testing InVision prototypes, proving demand among designers seeking faster feedback. [92][94]
$1–5M ARR — EXPAND THE CORE WORKFLOW. In 2020, it raised $2 million, added Figma support, and extended testing across major design tools; by 2021 it reached $1.5 million ARR and 40,000 companies. [92][93]
$5–10M ARR — MOVE UPSTREAM IN THE PRODUCT CYCLE. After a $15 million Series A in 2021, Maze launched Discovery to support research before prototypes existed. [93][99][94]
$10–50M ARR — BUILD AN END-TO-END RESEARCH PLATFORM. A $40 million Series B in 2022 funded expansion into new product-development use cases, recruitment and broader research capabilities. [95][101]
$50–100M ARR — NOT PUBLICLY DISCLOSED as a standalone ARR milestone; Maze reported 60,000+ teams and $60 million total funding in 2026. [103]
$100M+ ARR — NOT REACHED / NOT PUBLICLY DISCLOSED.
Rule: Attach to an existing creative workflow, prove one repeatable validation loop, then expand across the research lifecycle before enterprise requirements emerge.
COPY PLAYBOOK : What Worked → What Failed → What to Replicate → What to Avoid
THE STANDARD: Turn an expert-led workflow into a self-serve decision loop by consuming an artifact the customer already creates, then expand into the surrounding research lifecycle.
SEQUENCE:
1. Identify a recurring decision that currently waits on a specialist or agency.
2. Build the smallest tool that converts an existing artifact into a decision-ready output.
3. Integrate with the platform where that artifact is created before building a standalone destination.
4. Add participant supply, collaboration and governance only after the core workflow proves recurring usage.
WORKED: Maze’s design-tool integrations and Product Hunt launch created a low-friction entry path, while Discovery expanded its role earlier in product development. [92][93][94]
CAUTION: Do not rely on one design platform for distribution; do not sell automated metrics as insight without research design, participant quality and decision context.
bottom of page