A minimum viable product (MVP) is a startup's very first usable product, built to test demand. This guide shows how to scope that first version, track learning, and decide to iterate or pivot. Most early risk is building something few people want.
CB Insights ranks lack of market need as the top failure cause, cited in roughly 35–42% of failures, in its startup failure research. That sample covers self-reported venture-backed post-mortems, not all small businesses. Treat the number as a warning about demand risk, not a prediction for your idea.
Table of Contents
- What counts as an MVP?
- What should you build first?
- When should you pivot or persevere?
- How do you measure demand and fit?
What counts as an MVP?
Y Combinator defines the MVP as a business's very first product whose launch tests demand and concludes validation. Y Combinator stresses an MVP is a usable product, not a prototype, in its practical design guide. Users must be able to complete the core job.
A prototype shows what might exist. An MVP lets a customer pay, book, subscribe, or finish a task. Even manual work behind a simple page can qualify when the user gets real value.
What should you build first?
Start from one demand test, not a feature list. Pick the smallest set that lets strangers try the core promise. Cut everything that does not change the buy, signup, or keep decision. Use these cuts to stay lean.
Each choice must aid learning. Do not add admin panels, themes, or scale work early. Extra scope slows learning and hides weak demand. If no stranger will use the rough version, polish will not fix it.
- One user, one problem, one action to complete.
- One channel to reach those users.
- One way to charge, reserve, or measure keep.
When should you pivot or persevere?
The Lean Startup methodology directs teams to build to begin learning as quickly as possible. The Lean Startup methodology then uses actionable metrics showing cause and effect to decide to pivot or persevere, as explained on its official methodology page. That decision is the point of the MVP. Favor metrics tied to behavior.
Signups mean little without activation and repeat use. Change one offer or flow at a time so cause stays clear. Lenny's Newsletter research across dozens of successful consumer products frames this stage as iterate until enough users stick around. Expect several rounds of use, feedback, and fixes. Stop adding features when held use grows without prompts.
How do you measure demand and fit?
Track activation and cohort retention from the start. Activation shows new users reached value. Retention shows they return without a push. MeasuringU reports the Sean Ellis test signals product-market fit when at least 40% of active users say they would be very disappointed without the product.
The threshold comes from benchmarking nearly 100 startups, described in MeasuringU's product-market fit analysis. Use it with behavior data, not alone. Shopify's validation guide operationalizes that test with a core survey question for active users. Ask active users: How would you feel if you could no longer use this product? Sort answers by very disappointed, somewhat disappointed, and not disappointed to gauge demand.