“Minimum viable product” is one of the most misunderstood phrases in startups. It isn’t a cheap, buggy version of your dream product. It’s the smallest thing that tests your riskiest assumption with real users.
Find your riskiest assumption
Every startup rests on assumptions:
- People have this problem (tested in Level 2).
- They’ll use our solution instead of their workaround.
- They’ll pay ₹X for it.
- We can deliver it at a cost that leaves a margin.
- We can acquire customers for less than they’re worth.
Which one, if wrong, kills the business? Your MVP should test that one first.
Six MVP types
- Concierge MVP. You deliver the service by hand. Example: a “meal planning service” that is really you on WhatsApp sending plans you wrote.
- Wizard of Oz. The customer sees an automated product; behind it, humans do the work.
- Landing page MVP. The product is described but not built. You measure sign-ups and pre-orders (Level 2).
- Single-feature MVP. Build just the one feature that delivers the core value. Most apps start here.
- Piecemeal MVP. Stitch together existing tools (Google Forms, Sheets, WhatsApp Business, Razorpay payment links, no-code builders) to deliver the experience.
- Pilot or prototype. For hardware or manufacturing, a small batch or a 3D-printed prototype tested with a few customers.
Scope it brutally
List every feature you want. Then cross out anything that doesn’t directly test your riskiest assumption. What remains is your MVP. If it takes more than 6–8 weeks to build, cut more.
Launch, then listen
- Get it in front of 10–50 real users.
- Watch them use it, in person if possible.
- Track whether they come back without being reminded. That’s the real signal.
Your takeaway
The goal of an MVP is learning per rupee. Next: how to actually get it built, whether with no-code, freelancers, an agency or a technical co-founder.