Minimum viable product (MVP): what it is, with examples

A minimum viable product (MVP) is the first version of your product that real customers can try, built with as little effort as possible, because you want to know whether the idea deserves months of work and savings before you put them in. Sometimes that’s a landing page with a pre-order button. Sometimes a three-minute video. Or it’s you, delivering a service by hand to ten customers while the software still lives on a whiteboard.

First, what Eric Ries actually wrote in 2009, and why ‘minimum’ can mislead. Then the cases (a Milan shoe brand, Dropbox, Zappos) and a three-week plan for padel courts you could start on Monday. Italy adds a twist: prototyping can count as R&D. Right at the start? Our map from zero to your first round shows where the MVP fits.

What is a minimum viable product, exactly?

The definition most people quote is Eric Ries’s, written in August 2009, two years before his book The Lean Startup. An MVP, he said, is ‘that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort’. Then a warning that still trips founders up: ‘MVP, despite the name, is not about creating minimal products.’

So minimum describes your effort, not your ambition. Viable means a real customer can use the thing while you watch what they do with their money or their time. A prototype on your laptop doesn’t qualify. A clunky service that ten strangers pay for does.

	Question it answers	Who sees it	Typical form
Proof of concept	Can this be built at all?	The team, maybe a technical partner	A lab test or a script that proves one function
Prototype	How will it look and work?	Test users, designers, sometimes investors	A clickable mock-up, a 3D print, a sketch
MVP	Will customers use it, or pay for it?	Real early customers	A landing page, a manual service, a one-feature app
Beta	Is it ready for everyone?	A wider group, just before launch	A nearly finished product with known bugs

Types of MVP, with real examples

Apps get most of the attention, but the logic works just as well for shoes. Take Velasca, the Milan brand of handmade men’s shoes. It launched in 2013 to sell online and nowhere else, and in 2014 the founders tried a temporary store. Think of it as an MVP for a sales channel rather than a product: a short experiment, and it showed them what a physical channel could do. Eight years after launch, Forbes Italia was counting 12 Velasca shops (in Italy, France and the United States), all from a company that started out selling only online.

• A landing page, or smoke test, describes a product that doesn’t exist yet as if you could order it today. You pay for a little search advertising to bring visitors in and ask each of them for an email, a pre-order or a deposit, so the people who say yes are strangers who care before you’ve built a thing.
• A demo video is the Dropbox route from the top of this page. Drew Houston recorded and narrated the three minutes himself, and he pitched them at tech early adopters.
• With a concierge MVP you do the work by hand for a few customers, and they know there’s a person behind it. It’s slow, but you hear every complaint first-hand.
• Wizard of Oz is the concierge with the curtain closed: the customer sees what looks like a finished product, and you’re backstage doing the work. Zappos started this way in 1999 at Footwear Etc., a Sunnyvale shoe store, where Nick Swinmurn photographed the shoes. He put the photos online and made the shop a promise: any pair that sold, he’d buy at full price. That promise grew into the business Amazon bought in 2009, for $1.2 billion.
• A single feature can often be built in days with no-code tools, even if you can’t program, and our guide to building an MVP without coding has a two-week plan for it.

How to run an MVP in three weeks: a worked example

Say you want to sell Milan’s padel clubs a tool that fills courts left empty at the last minute. This first version is a concierge MVP, run by hand without a line of code, and it lives or dies on one number: the share of last-minute slots that get booked. The pass mark goes on paper on day one: at least 30% of the slots the clubs hand you, plus two clubs out of three willing to pay €3 per filled slot afterwards.

Week	What you do	Cost (example)	What you measure
1	Interview 12 club managers; three agree to a free trial	Your time	How many name empty courts as a top-three problem
2	Every afternoon, post the three clubs’ free slots in a WhatsApp group of 150 players you recruited	€80 of Instagram ads to build the group	Slots offered and slots booked
3	Keep posting, then ask each club to pay €3 per filled slot from week 4	Your time	Fill rate (pass: 30%) and clubs that say yes (pass: 2 of 3)

Suppose you filled 41 of the 120 slots offered (34%) and two clubs said yes. Now you know what to build first: the alert that goes out the moment a court frees up. Filled only nine? Then part of the plan is wrong, the channel or the customer, and that’s where a startup pivot begins. For the budget, our breakdown of how much money it takes to start a startup lists real costs, item by item.

MVPs in Italy: prototyping counts as R&D

Planning to register as a startup innovativa, Italy’s innovative startup status? Then your MVP work can do double duty. Article 25 of Decree-Law 179/2012 offers three ways to qualify, and you only need one. The R&D route asks for spending of at least 15% of the higher of your costs and your total value of production.

The same article widens what counts. On top of the accounting standards it lists ‘sperimentazione, prototipazione e sviluppo del business plan’ (experimentation, prototyping and business plan development), certified incubator services, and the gross cost of staff and outside consultants working on R&D, founders and directors included. The spending has to show up in the latest approved accounts. In year one, with no accounts yet, the legal representative signs a declaration instead. So keep your MVP invoices and timesheets. The other two routes are in our guide to the innovative startup in Italy.

MVP mistakes to avoid

• Building before talking. A dozen customer conversations cost nothing, so have them first. Our guide to validating a startup idea in seven days shows how.
• No metric. Or one picked after launch, which is worse, because then every result counts as a win.
• Too many features. At IMVU, Ries’s team once spent two weeks on a feature ‘that absolutely nobody wanted’. A simple AdWords smoke test, he wrote, would have shown how bad the concept was.
• Counting sign-ups as demand. An email address costs the visitor nothing. A deposit or a second order costs them something, which is why it tells you more.
• Polishing too early. Early adopters forgive rough edges when the problem goes away.

An MVP with paying users is solid evidence to bring to a first pre-seed round. Before you raise, our investor readiness test asks 12 questions and sends you a score and an action plan by email.

Related terms

• What you’re really testing for is product-market fit, and the first hint of it is MVP users who keep coming back.
• If the MVP has done its job, the next document is a go-to-market plan for selling at scale.
• Need the two-line version? See the MVP entry in our startup glossary, next to 49 other terms you’ll hear in your first year.

This page is general information, not legal or tax advice. Rules change: check the official source or ask a professional before you act.