What Is an MVP? Minimum Viable Product Explained
A minimum viable product (MVP) is the simplest version of a new product that lets a team collect the maximum validated learning about customers with the least effort. It is not a cheap, half-finished first release — it is a deliberate experiment built to test your riskiest assumption before you invest further.
Quick Answer: An MVP is the smallest experiment that starts the Build-Measure-Learn loop. Its purpose is learning, not shipping. You build just enough to discover whether real customers want what you are offering — using the least time, money, and effort possible.
How an MVP works: the Build-Measure-Learn loop
An MVP works by turning your idea into a single experiment inside a repeating cycle that Eric Ries calls the Build-Measure-Learn loop in The Lean Startup. You build the smallest thing that can test one belief, measure how real people respond, and learn whether to keep going, change direction, or stop.
The loop has three stages, and the MVP is what you build to enter it:
- Build: Create only the minimum needed to test one assumption — not a full product, not every feature you eventually imagine.
- Measure: Put it in front of real customers and watch what they actually do, not what they say they might do.
- Learn: Use that evidence to decide whether to persevere with the plan or pivot to a new one.
The output you care about is what Ries calls validated learning — proof, backed by real customer behavior, that a belief about your business is true or false. Every MVP targets a leap-of-faith assumption: the riskiest thing that has to be true for the idea to work. If that assumption fails, nothing else matters, so you test it first.
This is the part non-technical founders most often miss. An MVP is a learning vehicle, not a rough draft of the finished app. The question is never "how little can we ship?" but "what is the fastest way to learn whether this idea is worth pursuing?"
A concrete example of an MVP done right
A classic MVP example is the early Dropbox demo video. Before building the full file-syncing engine, founder Drew Houston released a short video showing how the product would work — and demand surged before a single customer had touched real software.
Here is why it counts as an MVP rather than marketing. Houston's riskiest assumption was not "can we build this?" — the engineering was hard but possible. It was "do enough people actually want effortless file syncing to sign up?" The video tested exactly that belief with almost no product built, and the jump in sign-ups gave him validated learning that demand was real.
Contrast that with the common trap: spending a year building a polished app, launching it, and only then discovering nobody wants it. The Dropbox video produced the same learning in days. That is the difference between an MVP and a small first version — the MVP is designed around a question, not around a feature list.
MVP vs prototype vs pretotype: what is the difference?
An MVP, a prototype, and a pretotype are easy to confuse, but each answers a different question. A prototype tests whether you can build something; a pretotype tests whether you should; an MVP tests whether real customers will adopt and pay once it exists in the market.
The table below compares the three by the question each one answers and what it is built to prove.
| Approach | Core question | Shown to | Main goal |
|---|---|---|---|
| Prototype | "Can we build it, and how should it look?" | Internal team, sometimes users | Test feasibility and design |
| Pretotype | "Should we build it at all?" | A handful of target users | Test demand with almost no build |
| MVP | "Will customers adopt and keep using it?" | Real early customers in the market | Maximize validated learning |
Takeaway: A prototype and a pretotype rehearse the idea privately; only the MVP puts a working slice in front of real customers, which is what turns opinion into validated learning. Pretotyping was coined by Alberto Savoia — "pretend it's a prototype" — to pressure-test demand before any real build.
When an MVP matters most for validation
An MVP matters most when your idea rests on an unproven assumption about what customers want and discovering the truth late would be expensive. If you already have strong evidence of demand, you may not need one; if you are guessing, an MVP is the cheapest way to stop guessing.
Because an MVP is fundamentally a test, it is really one kind of validation experiment — a structured way to convert a risky assumption into evidence. Different assumptions call for different formats, from concierge MVPs to landing pages, and the guide to the types of MVP walks through which to pick for which question.
An MVP is not the whole job, though. It is one stage inside a larger process of testing an idea end to end, which the complete guide to startup idea validation lays out step by step. Used well, an MVP protects the scarcest resources a founder has — time and money — by killing weak ideas early and pouring fuel on the ones that earn it. Tools like Edmired exist to make that testing loop faster to run.
Key Takeaways
- An MVP is a learning tool, not a cheap product. Its job is to produce validated learning about customers, not to ship the smallest possible app.
- The goal is maximum learning for minimum effort. You build only enough to test one belief and no more.
- Every MVP targets the riskiest assumption. Test the leap-of-faith belief that everything else depends on, and test it first.
- It powers the Build-Measure-Learn loop. Build a test, measure real behavior, then learn whether to persevere or pivot.
- Real customer behavior beats opinion. An MVP measures what people do, which is why it outranks surveys and internal prototypes.
- The term has a clear lineage. Frank Robinson coined "minimum viable product"; Eric Ries and Steve Blank made it central to the lean startup movement.
Frequently Asked Questions
What does MVP stand for in business?
MVP stands for minimum viable product. In a startup context it means the version of a new product that lets a team learn the most about customers with the least effort. The emphasis falls on "viable" and on learning — the product must be real enough to generate genuine feedback, but no larger than that test requires.
Who came up with the minimum viable product concept?
The term "minimum viable product" was coined by Frank Robinson around 2001. It was later popularized by Eric Ries in The Lean Startup and by Steve Blank through his customer development work. Ries reframed the MVP specifically as a tool for validated learning, which is the definition most founders rely on today.
Is an MVP just a cheap version 1 of my product?
No — this is the most common misconception. An MVP is not a low-budget first release; it is an experiment designed to answer a specific question about customer demand. A version 1 aims to serve customers, while an MVP aims to teach you whether those customers exist. An MVP can even be a video or a manual service with no finished software at all.
Does an MVP have to be a working product?
Not necessarily. An MVP only has to be real enough to test your riskiest assumption honestly. Sometimes that means functioning software; often it means a landing page, a demo video, or a service you deliver manually behind the scenes. What matters is that real customers make a real decision, producing evidence you can trust.