An MVP isn't a throwaway prototype

MVP stands for minimum viable product: the smallest version of a product that already gives real users real value. The idea behind it is simple. Instead of spending a year building a complete product behind closed doors and then discovering that the market wants something different, you go out early with a first version and learn from the people using it.

In practice, the term is often misunderstood, in two directions. Some build a demo with buttons that don't do anything and call it an MVP. That isn't an MVP, it's a presentation: nobody can work with it, so nobody learns anything. Others build half a product with every feature, each one only half done. That isn't an MVP either. It's an unfinished product.

A good MVP is small but complete in what it does. It solves one problem properly instead of ten problems halfway.

The art of an MVP is in leaving things out, but what you leave out matters. Features can wait: convenience in the admin area, rare special cases that someone handles by hand at first, integrations temporarily replaced by a manual export. All of that can be added later without rebuilding anything that already exists.

Foundations can't wait. A data model that gets the core concepts of the business wrong, user management without proper permissions, data protection that's supposed to come “later” — those are the shortcuts that get expensive after a few months. Take them, and you're not building an MVP. You're building a prototype you'll have to throw away as soon as it succeeds. And that's exactly when there's no time to rebuild.

The real value of an MVP comes after launch. With us, you can log in and start working with the system a few weeks after the project begins. You watch it grow, you adjust the direction, and you see early on where the original idea looks different in reality than it did on paper.

At VERSIFY, a platform for insurance brokers, the MVP was built in a few months. Since then, we've been developing the platform further based on what the first pilot customers experience, rather than a list written before launch. They particularly value the analysis features and the reduced effort in case management.

That's the point of an MVP: the market decides what gets built, not the meeting.

AI-assisted development has made building first versions much faster. That's real progress, but it's also a new trap: an impressive demo can now be built in days. An MVP that feels like magic after three weeks and is unmaintainable after three months still hasn't helped you.

That's why we build MVPs for production, not for the demo. Most of our MVPs aren't throwaway prototypes but the first version of the later product, with an architecture that grows as the market pulls. That way, you invest little up front, learn fast and don't have to start over when it works.