Why MVPs are too expensive — and what to do instead

The Minimum Viable Product was supposed to be the cheap way to test an idea. Build the smallest thing that delivers value, ship it, learn. In theory, lean. In practice, most MVPs are neither minimum nor cheap.

The MVP is still a product

Look at what an MVP actually requires. You need to design it, build it, host it, and make it work well enough that a stranger's first impression isn't "this is broken." That means real engineering: authentication, a database, payments, error handling, a design that doesn't scare people off. Even a "minimum" version is weeks of work — often months, and often thousands of euros in developer time or your own nights and weekends.

And here's the part that stings: you pay all of that before you learn the one thing you actually needed to know — does anyone want this?

The real cost is the wrong question

An MVP answers "can we build it?" and "does it work?" Those are real questions — but they're the wrong ones to ask first. The question that kills most startups isn't "can we build it?" It's "does anyone care?" You can build a flawless product nobody wants. The graveyard is full of them.

When you build an MVP to answer a demand question, you've spent your most expensive resource — time — answering it in the slowest possible way. Worse, sunk cost sets in. After three months of building, "nobody's signing up" doesn't feel like a clean signal to stop. It feels like a reason to add one more feature.

Test the demand, not the product

There's an older, cheaper idea than the MVP: pretotyping — coined by Alberto Savoia. Instead of building the product to see if it works, you fake just enough of it to see if anyone wants it. You test desirability first, and you spend days, not months.

A landing page that describes the offer as if it already exists. A real ad campaign driving real strangers to it. A button that says "Buy" and measures how many people click. You're not lying — you're measuring intent honestly, and you tell people the moment they act. What you get back is data: not opinions from friends, but behaviour from strangers who don't care about your feelings.

If nobody clicks, you've learned that in a week, for the price of an ad budget — not in six months, for the price of a product. That's not failure. That's the cheapest, fastest "no" you'll ever get, and it just saved you a year.

Build only what earns a build

The point isn't to never build. It's to earn the right to build. Test the demand first; if the signal is a clear yes, then pour months into an MVP — now with evidence that someone is waiting for it. If the signal is no, you kill it cheaply and move to the next idea with your runway intact.

That's the whole game: fail fast, fail cheap, and only spend real money on ideas that have already earned it.

Two ways to start

If you want to run this yourself, our Academy teaches the whole loop — from finding a problem worth solving to reading the signals honestly. If you'd rather have it done for you, a Validation Sprint puts your idea in front of a real audience in two weeks and hands you a clear go/kill. Either way: test before you build.