MVP, prototype or proof of concept: which do you need?
What is the difference between a proof of concept, a prototype and an MVP?
A proof of concept answers 'can it be done?', a prototype answers 'how would it work and look?', and an MVP answers 'will people use it and pay for it?'. They are three tests for three different risks — technical, design and market — and most products need one or two of them, not all three. Choose by the question you cannot answer yet, not by the word an investor or a vendor happened to use.
In short:
- Proof of concept (PoC): a small technical experiment that shows whether the hardest part can be built. For you and your developers. Usually thrown away. Days, or a few weeks at most.
- Prototype: screens that show how the product would work and look, usually with made-up data and nothing real behind them. For users, stakeholders, investors and the people who will build it. Days to a few weeks, depending on how many screens and how real they must look.
- MVP (minimum viable product): the smallest working product that real users can sign into and use with real data, often paying for it. For your first customers. With us from $12k, and about ten weeks for a typical one.
What is a proof of concept?
A proof of concept tests technical feasibility: whether one specific, uncertain thing can be done at all, or done well enough. It has no design to speak of, often no screens, and only enough code to get an answer. The result is a short written verdict — yes, no, or yes if — with the evidence behind it.
Good PoC questions are narrow and genuinely open. Can we read the data this machine exports? Can this outside system's API do what we need, fast enough? Can an AI model pull the right fields out of our invoices often enough to save staff time? (Our article on what it costs to add AI features explains how to test that kind of feature.) If an experienced developer could answer the question from memory, it does not need a PoC.
A PoC is for you, your technical lead and sometimes an investor who needs to know the hard part is possible before funding the rest. Keep it small: days, or a couple of weeks of one developer's time. If it needs months, it has turned into a product without anyone deciding so.
What is a prototype?
A prototype shows how the product would work and what it would look like before the real thing is built. It ranges from sketches and wireframes — boxes and labels on paper or a screen — to a clickable set of screens that looks finished, made in a design tool or as a rough front end with made-up data. Behind the screens there is nothing: no accounts, no database, no payments.
It answers questions about flow and understanding. Can a first-time user find their way from sign-up to the thing they came for? Do the screens make sense to the staff who will use them every day? Does everyone — founders, investors, developers — picture the same product? Putting a prototype in front of a handful of target users and watching them try a task is one of the cheapest ways to find a confusing screen.
The effort grows with the number of screens and how real they must look: a few days for a handful of clickable screens, longer for a whole product with every state designed. What a prototype cannot tell you is whether anyone will use the product for real, or pay for it.
What is an MVP?
An MVP, or minimum viable product, is the smallest version of a product that real users can sign into and use for the job it promises, with real data — enough to learn whether they want it. The term was popularised by Eric Ries's book The Lean Startup in 2011, and both halves of it matter: minimum, so it ships soon and costs little, and viable, so the people using it judge the idea and not the bugs.
An MVP is working software. It has accounts, a database and hosting, and often payments, because 'will people pay?' is frequently the question it exists to answer. It is for your first customers, and for investors who want evidence of use rather than a deck. That is how we approach MVP development: a product real users can log into, not a prototype that only works in a demo, on a live URL from the first week.
What does each one cost, and how long does it take?
A proof of concept and a prototype are usually small pieces of work. Their cost depends almost entirely on the one question being tested or the number of screens being designed, and we do not publish prices for either. An MVP is a product and is priced like one; how much an MVP costs and how long it takes to build an app go into the detail.
These are the figures our MVP cost calculator gives, counted in developer-weeks with design and testing included:
- The smallest useful web product — one kind of user, sign-in, hosting and the core screens, about four developer-weeks: $12k–$16k, and about five weeks with one developer.
- A typical MVP — a web app for customers and staff with an admin panel, payments, email and push notifications and file uploads, connected to one or two other systems, on a proven design system — about 18 developer-weeks: $23k–$34k, and about ten weeks with two developers.
- The same typical MVP with a dedicated team instead of a fixed price: two developers at $12k a month for about two and a half months.
Which one do you need?
Start from the riskiest thing you do not know yet, and build the cheapest thing that answers it:
- You are not sure it can be built — an unusual integration, a connection to hardware, an AI feature that has to be accurate. Start with a proof of concept for that part alone.
- You know it can be built, but not how people will move through it, or you need to show it before anyone will fund it. Make a prototype.
- You know it can be built and how it should work, but not whether people will use it and pay. Build an MVP.
- You know all three — for example, software that replaces a spreadsheet your own staff already use every day. Skip the experiments and build the first release from a written scope.
How do a proof of concept, a prototype and an MVP lead into each other?
When a product needs all three, they come in that order: the proof of concept removes the technical risk, the prototype settles the flow, and the MVP tests the market. Each one hands something to the next, but rarely its code:
- From the PoC, the answer and what was learned: which approach works, where its limits are and what it will take to build properly. PoC code is written to answer a question fast, so treat it as a sketch unless it was written to be kept.
- From the prototype, the screens and the decisions behind them. A clickable prototype is one of the best inputs to a written scope, alongside a short software project brief.
- From the MVP, the product itself. Its code is the foundation of the second release, and the features you cut from it become the next scope. Built in short slices, it keeps doing part of a prototype's job too: with us the first slice is on a live URL in week one and there is a clickable demo every Friday, so changing your mind stays cheap.
What are the most common mistakes?
Each of these spends money without answering the question you actually have:
- Calling a clickable prototype an MVP. A prototype can show that people understand the idea and like the look of it. It cannot show that they will come back next week or pay, because there is nothing real to use. Clicks on made-up data are not usage.
- Building a proof of concept nobody needed. Sign-in, card payments, sending email, an app in both stores: these are solved problems, and a PoC for them proves what everyone already knew. Spend the time on the one part that is genuinely uncertain, or skip the PoC.
- Shipping the PoC as the product. Code written to answer a question in a week is rarely the code you want running a business. Decide before you build it whether it is meant to be kept.
- Polishing the prototype for months. Once it has answered the flow questions, every extra week of polish is a week without real users.
- Forgetting the 'minimum'. An MVP that tries to be version three takes version three's time and budget, and learns its lessons late.
- Forgetting the 'viable'. An MVP so thin or so buggy that people leave for reasons unrelated to the idea tells you nothing about the idea.
- Not writing the question down. Before you build any of the three, write down the one question it must answer, and what answer would change your plan.
Is a prototype the same as an MVP?
No. A prototype shows how a product would work, usually with made-up data and nothing behind the screens; an MVP is working software that real users sign into and use with real data. A prototype can tell you whether people understand the idea; only an MVP can tell you whether they use it and pay for it.
Do I need a proof of concept before building an MVP?
Only if part of the product might not be technically possible, or not possible well enough. Most business software is built from solved parts — accounts, payments, notifications, an admin panel — and needs no proof of concept at all. When there is a genuine unknown, test that part alone before committing the full budget.
Can a prototype be turned into an MVP?
A clickable prototype made in a design tool cannot, because there is no code behind it. Its screens and decisions carry straight over into the MVP's scope and design, which saves time. A coded prototype can be developed further only if it was written to be kept.
How long does it take to build an MVP?
Our typical MVP — a web app for customers and staff with an admin panel, payments, notifications and file uploads — reaches its first real users in about ten weeks with two developers. The smallest useful web product takes about five weeks with one developer. Either way, the first slice is on a live URL in week one.
How do I know which one my idea needs?
Write down what you do not know yet: whether it can be built, how people would use it, or whether they would pay. Then send us the brief. We reply within one working day, and for an MVP you get a written scope with what we would build first, what it costs and when it ships. Worst case, you get a free second opinion.