The most common question I get from founders is "how much does an MVP cost?" And every developer gives the same answer: "it depends."
That's technically true and completely unhelpful. So here's what I actually tell people when I'm having an honest conversation.
Start with validation, not a product
Before you build an MVP, prove someone wants it. A landing page, a waitlist, maybe a simple version of one core feature. Think Typeform plus Stripe plus a Notion database on the backend. It works. It's not pretty. It proves whether anyone will pay for the thing.
What you get: validation that the idea has legs. What you don't get: anything resembling a real product.
This is right for people who aren't sure anyone wants what they're building. If you're still in the "is this a real problem?" phase, building more than this is premature.
The real MVP
This is where most real MVPs live. A proper app — authentication, a core workflow, maybe two or three features that work together, basic admin capabilities. It's built on real infrastructure and can handle your first few hundred users.
It won't have every feature you want. You'll have to say no to things. That's the point. The discipline of cutting scope is what makes the difference between an MVP that ships and one that lives in development purgatory forever.
When you're past the MVP
At some point you're not building an MVP anymore. You're building version one of a product. Multiple user roles, integrations with external services, a polished design system, maybe a mobile app alongside a web app.
This is right if you've already validated the idea — through a cheaper build, through existing customers, through domain expertise that makes you confident in the market. It's wrong if you're guessing.
The mistake I see constantly
Founders who should be validating cheap spend big because they're afraid of looking unfinished. They want the product to feel "real" before anyone sees it.
But the first version of everything is embarrassing. Instagram launched as a simple photo app with filters. Nothing else. The founders who win are the ones who get comfortable with "good enough for now."
Start cheaper than you think you should. Learn faster than you think you need to. Spend the bigger money once you know what to build.