Minimum Viable Product (MVP)
A minimum viable product (MVP) is the smallest thing you can build to test your riskiest assumption with real customers. It is not a cheap, broken version of your dream product; it is a learning tool designed to get answers quickly and with as little waste as possible.
The Lean Startup and the MVP
Eric Ries popularised the MVP in The Lean Startup. He defines it as the version of a new product that lets a team collect the maximum amount of validated learning about customers with the least effort. Validated learning means evidence from real customer behaviour, not opinions or guesses.
The MVP sits inside Ries's build–measure–learn feedback loop:
Although you run the loop as build, measure, learn, Ries suggests you plan it in reverse: decide what you need to learn, then what to measure, and only then what to build.
Types of MVP
| MVP type | How it works | Tests |
|---|---|---|
| Landing page / smoke test | A simple page or social media post describing the product with a "Sign up" or "Pre-order" button | Whether people are interested enough to act |
| Explainer video or demo | A short video showing how the product would work. Dropbox is a well-known example of testing demand this way before the full product was ready. | Demand for a product that is hard to build |
| Concierge | You deliver the service personally and openly, by hand, to a few customers | Whether the outcome is valuable and what customers really need |
| Wizard of Oz | Customers see what looks like an automated service, but people do the work behind the scenes | Whether customers use it as you expect, before you automate |
| Single-feature MVP | A real product that does one core thing well | Whether the core feature solves the problem |
| Pre-sale | Taking deposits or orders before production | Willingness to pay, the strongest signal |
Measure the right things
Ries warns against vanity metrics: numbers that look impressive but do not guide decisions, such as total downloads or page likes. Prefer actionable metrics that show cause and effect, such as the percentage of sign-ups who place a second order within two weeks. Before launching an MVP, write down your success criterion: "If at least 20 of the 100 people who visit the page pay a KES 200 deposit, we continue."
Persevere or pivot?
After each loop, decide whether to persevere (keep improving the current strategy) or pivot. Ries describes a pivot as a structured course correction to test a new fundamental hypothesis. Examples include a zoom-in pivot (one feature becomes the whole product), a customer segment pivot (same product, different customers) and a channel pivot (selling through a different route).
Mercy and Sam, both 23, want to sell healthy lunch boxes to office workers in Westlands, Nairobi. Their riskiest assumption: workers will pre-order lunch the day before. Instead of building an app and renting a kitchen, they ran a concierge MVP: a WhatsApp catalogue shared with two office buildings, orders by 6 p.m., payment by M-Pesa and lunches cooked in Mercy's kitchen (after checking what food-handling permits they would need to continue). Success criterion: 30 paid orders in the first week. They got 41, but repeat orders came almost entirely from one building with no canteen. They zoomed in on offices without canteens before investing in anything bigger.
Spending months adding features before launching because the MVP "isn't ready". If you are not slightly embarrassed by your first version, you probably waited too long. The opposite mistake is releasing something so poor that it cannot test anything meaningful. "Viable" means it must still deliver enough value for customers to respond honestly.
Key points
- An MVP, as Eric Ries describes it, maximises validated learning for the least effort.
- Use the build–measure–learn loop, and plan it in reverse: learn, measure, build.
- MVP types include landing pages, demos, concierge, Wizard of Oz, single-feature products and pre-sales.
- Set a success criterion in advance, avoid vanity metrics, and decide to persevere or pivot.
Tip: use the ← and → keys to move between pages.
