What is an MVP — and why it isn't a 'stripped-down app'

MVP stands for Minimum Viable Product. It sounds like "a cheap, half-finished thing," and that's how almost everyone reads it. But an MVP isn't about "making it smaller." It's an experiment: the smallest thing that answers one scary question — does anyone actually want this?

What an MVP means, in plain words

An MVP is the first version of a product with exactly one core function worth opening it for. Everything else comes later. Not "an app missing half its features," but "an app that does one thing and tests whether people will use it."

The classic example: if you're building a food-delivery service, the MVP isn't an app with maps, chat, and payments. It's a landing page with a menu and an "order" button, where you deliver the first orders yourself. Cheap, fast, and you immediately see whether people order or not.

The big mistake: an MVP is not a "cheap app"

Here's the twist. The "Minimum" in MVP is not about quality or size. It's about how many guesses you're testing. You pick your riskiest guess ("people want a water tracker with reminders") and build just enough to test it — not one line more.

A half-baked app with ten unfinished features isn't an MVP. It's just a bad product. A real MVP does little, does it well, and answers a specific question.

How to narrow an MVP down to one question

A practical trick. Before you write any code, finish this sentence: "I believe that ___, and I'll test it if ___." For example: "I believe students will pay for notes from lectures, and I'll test it if 10 people click 'buy.'"

Then throw out everything that doesn't help get those 10 clicks. Sign-up? Not needed for the test. Dark mode? Later. Profile settings? Later. What's left is a tiny thing that's cheap to discard if the guess doesn't hold. That's the whole point: an MVP should be cheap enough that you won't mind dropping it.

An MVP isn't always an app

And here's the freeing part: an MVP doesn't have to be code at all. Often the most honest MVP is:

• A landing page describing the product with an "I want this" button — you count the clicks.
• A Google Form instead of a database — you collect requests by hand.
• You instead of an algorithm — you do the work manually until you're sure people are even asking for it.

Once demand is confirmed, you automate and turn the MVP into a real product. And before you charge money, it's worth thinking about what your product is worth.

FAQ: how is an MVP different from a prototype?

Short version: a prototype tests how it works, an MVP tests whether it's needed. You show a prototype to yourself and a couple of friends to feel out the interface; you hand an MVP to real users to watch their behavior. We broke it down in detail in MVP vs prototype.

FAQ: how long should an MVP take?

Measure it in days, not months. If your MVP takes more than a couple of weeks, extra stuff probably crept in — go back and cut everything that doesn't test the core guess. With vibe coding, a landing-page MVP is genuinely a one-evening build, which means you can test the idea this week instead of "someday."