MVP, prototype, POC — what you actually need first

Every kickoff call, a founder says “let’s build an MVP.” Half the time they mean prototype. A quarter of the time they mean proof of concept. The rest of the time they actually mean MVP. Confusing these three costs money.

Proof of Concept (POC)

What it is: A technical experiment to answer one question. Can this idea even work? Can Claude write brand-safe captions? Can we run realtime multiplayer on Supabase? Does OCR work on Indian-format invoices?

What it looks like: A working script or a bare-bones demo. Ugly, unbranded, no login, one happy path. Built for the founder, not for users.

Cost: ₹40k-₹1.5L Time: 1-3 weeks Ship as: Loom video, private staging URL, or throwaway CLI tool What comes next: Kill it or graduate to a prototype

Prototype

What it is: A high-fidelity clickable version that shows what the product feels like — before you build the backend. For selling the idea to co-founders, investors, or your first ten customers.

What it looks like: Figma prototype, or a fake-front-end web app where every button “works” but the data underneath is hardcoded. Real screens, real interactions, real brand.

Cost: ₹1-4L Time: 2-4 weeks Ship as: Shareable link, private beta What comes next: Validate demand, then decide on MVP scope

Minimum Viable Product (MVP)

What it is: The smallest complete product that solves the core problem for real users, with real backend, real payments, real accounts. Not fancy. Not full. Not the whole roadmap. Just: does the core loop deliver value?

What it looks like: A live app with login, one or two core flows working end to end, real database, real payments if needed. Usually 20% of the eventual feature set, doing 80% of the work.

Cost: ₹5-15L Time: 4-8 weeks Ship as: Public app on the Play Store, live website, or invite-only beta What comes next: Watch usage. Ship the next 10 features based on what users actually do, not what you planned.

Which one do you actually need?

• If you don’t know if the tech will work — POC
• If you need to raise money or sell the vision — Prototype
• If you have paying interest and need to serve real users — MVP

The common mistake

Founders start with MVP scope and MVP budget when they actually need a prototype. They burn ₹8L on a full build that gets used by 20 people, learn something obvious, and pivot. If they had spent ₹2L on a prototype first, they’d have learned the same thing for a quarter of the cost.

The other common mistake

Founders who need MVP but build POC-quality — no login, no payments, no polish — then wonder why users don’t come back. If it’s live for real customers, it’s an MVP. Budget accordingly.

Rules of thumb

• POC: nobody outside your team should ever use it
• Prototype: users can click through but can’t hurt themselves
• MVP: users can pay you and you can lose them if you break it

Pick the smallest one that answers your current question. Ship it. Then decide what’s next based on evidence.