MVP is one of the most misused terms in startup culture. Some people think it's a wireframe. Others think it's a demo. This guide clarifies what an MVP actually is and why it matters.
The Definition That Actually Works
MVP = Minimum Viable Product = the smallest working product that tests your core hypothesis and gets feedback from real users. Key phrase: "working product." Not a demo, not a pitch deck, not a Figma mockup. Something users can actually use and give feedback on. Key phrase: "core hypothesis." Your assumption is: "Founders will pay for a tool that..." Your MVP tests that. Everything else is secondary. Key phrase: "real users." Not your friends. Not people who are being paid to test it. Actual potential customers who found your product organically. Example: You think "Developers will pay $100/month for a tool that automatically optimizes their database queries." Your MVP: A web app where developers paste their slow SQL, click "optimize," and see suggestions. No login, no payment, no UI polish. Just the core insight: Can we give good optimization suggestions? If developers use it and love the suggestions, you've validated the core. If they ignore it, you've learned to pivot before spending 6 months building.
The Three Things Your MVP Must Have
First: One core feature. Not 10 features. One. The one thing you're betting on. Second: Real users trying it. 50+ people, minimum. Ideally strangers who found your product without your arm-twisting. If you have to beg your mom to use it, that's a signal your positioning is weak. Third: Measurable feedback. Did they use it? How often? What did they ask about? What features were missing? You need data, not "it's cool." Set up Mixpanel or Posthog (free tier). Track usage. What do users actually do vs. what you expected? Without these three, you don't have an MVP. You have a prototype or a proof-of-concept. Prototypes are internal learning tools. MVPs are external learning tools.
What an MVP Is NOT
MVP is not: A beautiful product. Your MVP can be gray buttons and no animations. If it converts, it works. MVP is not: A complete product. It has 3–5 features, not 50. You'll add more after learning from users. MVP is not: A sellable product (usually). Some MVPs monetize from day one. Most don't. Your goal is learning, not revenue. If you charge, be transparent: "This is beta." Charge a low price or make it free. MVP is not: Your final product. You'll rewrite it. The codebase is throw-away code. Don't optimize prematurely. Ship fast, learn, then rebuild better. MVP is not: A marketing stunt. "We launched on Product Hunt" is nice, but it's not validation. Validation is: Users came back. Users asked about pricing. Users invited their friends.
The Cost of Waiting vs. Shipping
Scenario A: You spend 6 months building the "perfect" product before launching. You spend $80K and ship with 50 features. You get 100 signups, but only 5 stick around. You wasted $80K and 6 months learning what users actually want. Scenario B: You spend 4 weeks and $15K building an MVP with 3 features. You ship to 100 beta users. 20 stick around. You learn that users want feature X (which wasn't in your original plan) and don't care about features Y and Z. You pivot, build feature X in week 5, and 40 users stick around. You've validated the core in 5 weeks. Scenario B is 10x smarter because you're learning from users, not your imagination. Most founder failure is self-inflicted: You build in stealth for 6 months, launch with a flourish, and discover users don't care. The MVP prevents this.
How to Know Your MVP Is Done
Your MVP is done when: Users can accomplish the core action (sign up, create something, see a result). You can show it to strangers without explaining 10 minutes of context. You've gotten feedback from 50+ users. You're learning the same feedback twice (diminishing returns). Your MVP is NOT done when: The database schema is perfect. The code is beautiful. The UI is polished. The marketing site is complete. These are nice-to-haves after validation. Releasing a half-finished MVP beats a perfect product that never ships. Ship early and often.
From MVP to Product
Once you've validated the core (users want it, will pay for it), you enter the product phase. Now you: Build more features based on user feedback. Scale the infrastructure. Hire a team. Create a go-to-market strategy. This is different from MVP. You're building for scale, not learning. But it all starts with MVP. If you skip MVP and jump to "product," you'll build what you think users want, not what they actually want.