The MVP guide: launch the smallest version that proves something

Write the hypothesis in one sentence: “SME managers will pay $X per month to stop doing Y by hand.” The MVP exists to confirm or refute it.

Every proposed feature passes one test: does it help answer that question? If not, it waits.

Mixing them up means spending at the wrong time.

• aA prototype tests a flow with a few people, often without production code.
• bAn MVP is a real product online, used by real customers, solid enough to grow.
• cThe product comes after: it widens based on what the MVP learned.

Modall publishes $15,000 to $30,000 for a prototype or micro-SaaS over 4 to 8 weeks, and $40,000 to $80,000 for a market-ready MVP over 3 to 5 months. Our costs page gathers these figures and other sources.

An MVP without measurement proves nothing. Before launch, decide what you will track:

• aactivation: how many sign-ups do the core action at least once;
• bretention: how many come back the following week;
• cconversion: how many pay, or agree to pay;
• dthe decision threshold: the number that will make you continue, pivot or stop.

The most frequent, and the most expensive:

• aBuilding the admin dashboard before the first customer.
• bAdding features to feel safe instead of measuring.
• cLaunching without knowing which number would change your mind.
• dThrowing the MVP away after validation because its foundations did not hold.