Building an MVP without a big budget

An MVP (minimum viable product) is not a smaller version of your dream product. It's the cheapest experiment that tells you whether customers will use and pay for the core of it. If you're not slightly embarrassed by the first version, you probably waited too long.

Start from the riskiest assumption

Ask: what must be true for this to work? Usually it's 'customers will pay for X' or 'we can deliver X reliably'. Build only what tests that.

Cheap ways to build it

1. 1
   ConciergeDeliver the service by hand to the first customers. You learn exactly what they need before automating anything.
2. 2
   Wizard of OzCustomers see a simple front end; behind it, you do the work manually.
3. 3
   No-code toolsForms, spreadsheets, website builders and automation tools can cover a surprising amount of a first version.
4. 4
   Single featureCode only the one thing nobody else does, and use existing services for login, payments and email.
5. 5
   Pre-saleSell before you build: a demo, a mock-up and a payment link. If nobody pays, you saved months.

Keep it small

• One type of customer, one main use case.
• A fixed deadline: 4–8 weeks, not 6 months.
• Cut every feature that doesn't help test the main assumption.
• Plan how you'll measure success before you launch.