Prototype — tests the design
A prototype is a clickable, interactive artefact that simulates the intended user experience. It exists to answer one question: does the design work? Do users navigate it the way we expect? Do they find the primary action? Do they get stuck?
A prototype does not have a real backend. There's no database, no auth, no persistence between sessions. Data is faked, buttons trigger simulated responses, and everything resets when you refresh. This is a feature, not a bug — the point is to iterate on design at ten times the speed of iterating on real software.
Typical tools: Figma with prototyping enabled, ProtoPie, Framer. Typical time: hours to a few days. Typical cost: free-to-low.
Proof of concept (POC) — tests a technical bet
A proof of concept exists to answer a specific technical question that could kill the project. Common examples: Can this AI model actually classify what I need it to? Can this third-party API deliver responses fast enough? Can our stack handle the load pattern we expect?
A POC is real code, usually messy, focused entirely on the risky bet. It has no user-facing polish — the audience is you and your co-founder, not future users. Once you have the answer, you throw the code away and start the MVP on a clean foundation.
Typical time: days to a couple of weeks depending on the technical question. Typical cost: developer time for the specific investigation, nothing more.
MVP — tests the market
A Minimum Viable Product is real, running software that a real user can sign up for and complete the core value action inside. It exists to answer the biggest question of them all: does anyone actually want this?
Unlike a prototype, an MVP has a real backend, real data persistence, real auth. Unlike a POC, it has enough polish that a stranger can use it without a tutorial. It is not a finished product — it is deliberately narrow, doing one thing that delivers the core value, and nothing else.
Typical time: one to four weeks with AI-assisted delivery, up to six months with a traditional agency. Typical cost: see our MVP development cost guide.
Quick decision framework
- Not sure about the design? Build a prototype first. Cheap and fast to iterate.
- Not sure the tech will work? Build a proof of concept before the MVP. Answer the technical question in isolation.
- Sure about both, but not sure users care? Skip straight to the MVP. Everything else is procrastination.
- Sure of everything? Congratulations — you don't need to be reading this. Ship it.
What FastPrototype builds
Despite the name, we build MVPs — real, running software you can put in front of real users. The "prototype" in our name is a nod to the speed, not the fidelity. If you're at the point where you need a clickable prototype, a tool like Figma will serve you better than us. If you're past that point and need actual working software, start a scope.