Quick summary
Prototype vs MVP: a prototype shows how your product would look and work so you can test the idea and the design with people, usually with no real code behind it. An MVP is the smallest working version that real customers can use and pay for. A prototype asks "does this make sense?"; an MVP asks "will people use it and pay?"
Where a POC fits: a proof of concept answers a third question, "can this be built at all?", and only matters when the technology itself is the risk.
Which to build first: for most SaaS ideas, a clickable prototype tested with five people in a week or two, then a lean MVP built in weeks, not months.
Prototype vs MVP at a Glance
| Proof of concept (POC) | Prototype | MVP | |
|---|---|---|---|
| The question it answers | Can this be built? | Do people understand and want it? | Will people use it and pay for it? |
| Who sees it | You, your developer, maybe an investor | A handful of target customers | Real customers |
| What it is | A small technical test of one risky piece | A model of the product: sketches, a clickable design or rough code | The smallest working product |
| Real data and logins | Usually not | No (design) or partly (coded) | Yes |
| Typical time | Depends on the risky piece, often days to a few weeks | 1 to 2 weeks for a clickable design | 4 to 8 weeks for a simple web MVP |
| Typical cost | Depends on the risky piece | About $2,000 to $5,000 for a clickable design | About $25,000 to $55,000 for a simple web MVP |
| What happens to it next | Thrown away once the answer is known | Usually thrown away or rebuilt | Becomes version one of your product |
The prototype and MVP figures come from published vendor guides (sources below); the POC row is our own rule of thumb.
Before you choose, answer these three questions:
- Question 1, "What is the riskiest thing I do not know yet?" Whether it can be built points to a POC. Whether people get it points to a prototype. Whether they will pay points to an MVP.
- Question 2, "Who do I need to convince?" Yourself or a technical cofounder, a POC. Early users or an investor in a meeting, a prototype. Paying customers, only an MVP.
- Question 3, "What will I do with the answer?" If no result would change your plan, skip that step and build the next one.
Ready to turn your prototype into a working MVP?
A senior developer on subscription: $2,995 a month, as many requests as you like, each shipped within 2 business days, and your money back if the first 14 days don’t prove it.
In this article
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
This guide is for founders deciding what to build first for a new SaaS product or web app, and what to pay for it. It is written in plain English for the person paying, not the person coding. If you have already decided to build an MVP and want to know who should build it, our guide to MVP development covers the options.
What Is a Prototype?
A prototype is a model of your product that people can look at or click through. Its job is to show the idea, so you can find out whether people understand it and whether the flow makes sense, before anyone pays for production code.
Prototypes come in three levels of detail:
Sketches and wireframes
Boxes and labels on paper or in a design tool. Fast to change, good for agreeing the screens and the order of steps with a cofounder or a developer.
Clickable design prototype
Realistic screens in a design tool such as Figma, linked together so you can tap through the main flow. It looks like an app but has no backend, so nothing is saved and nobody can sign in for real.
Coded prototype
Rough working code, often made quickly with an AI app builder, that does one or two things for real. It feels closer to the product, but it was built to demo, not to hold customers' data.
You learn the most from a prototype by watching people use it. Jakob Nielsen's research at Nielsen Norman Group found that a study with five users "has found 85% of the usability problems", and recommends "running as many small tests as you can afford" rather than one big one. Five target customers, one hour each, is enough to find the screens that confuse people.
What Is an MVP?
An MVP, or minimum viable product, is the smallest version of your product that real customers can use. Eric Ries, who made the term popular in The Lean Startup, defines it as the version of a new product "which allows a team to collect the maximum amount of validated learning about customers with the least effort." He adds that it "requires judgment to figure out, for any given context, what MVP makes sense."
The key difference from a prototype is that an MVP is real. People sign up, put in their own data, come back next week and, ideally, pay. That is the only way to learn whether the product is worth building further.
Speed matters more than polish. Y Combinator's Michael Seibel tells founders they "should be able to build it fast in weeks, not months", to "first time box your spec", and to "launch something bad quickly." The point is not to ship something broken. It is to ship the one thing that proves the idea, and nothing else.
Prototype vs MVP: The Key Differences
What it tests
A prototype tests understanding and design: can people work out what to do? An MVP tests behaviour and demand: do people use it again, and will they pay?
Who uses it
A prototype is shown to a few people, usually with you in the room or on a call. An MVP is used by customers on their own, without you explaining it.
What is real
A prototype can fake almost everything behind the screens. An MVP needs real logins, a real database, and data that is kept safe and private.
Cost and time
A clickable prototype takes days to a couple of weeks. An MVP takes weeks to a few months, because the parts nobody sees (accounts, payments, security, hosting) take real work.
What happens afterwards
A prototype is usually thrown away once it has answered its question. An MVP becomes version one, so the code needs to be good enough to build on.
The line between them can blur. Dropbox's founders famously tested demand with a demo video before the product was ready for the public. TechCrunch reported in 2011 that within 24 hours Dropbox "had 75,000 people signup for the wait-list," when the founders "were expecting 15,000, tops." The video was a prototype in form, but it was used the way an MVP is: to test whether people wanted it. Our MVP development guide covers that story and others like it.
POC vs MVP vs Prototype: Where a Proof of Concept Fits
A proof of concept (POC) is a small technical test that answers one question: can this be built? It is usually rough, rarely shown to customers, and thrown away once you have the answer.
Most SaaS products do not need one. If your product is logins, forms, dashboards, payments and integrations with well-known tools, the technology is not the risk: developers build those every day. A POC earns its cost when one piece is genuinely uncertain, such as:
- An AI feature that has to be accurate enough to trust, like pulling the right figures out of messy documents.
- An integration with a system that has no public API, or a poorly documented one.
- A performance target, such as processing millions of records overnight or answering in under a second.
- A hardware or device dependency, like reading data from a specific scanner or sensor.
Proof of concept vs prototype
A POC proves the technology works; a prototype proves people understand the product. A POC can have no screens at all, just a script that shows the hard part can be done. A prototype can have no working technology at all, just screens. If you are not sure the hard part is possible, do the POC first: there is no point testing screens for a product that cannot be built.
POC vs MVP
A POC is for you and your developer; an MVP is for customers. A successful POC tells you the risky piece is possible. It does not tell you anyone will pay for it, and its code is rarely fit to build the product on. Plan to build the MVP properly once the POC has answered its question.
What a Prototype and an MVP Cost
There is no official price list, so the useful figures come from firms that publish their own estimates. Treat them as planning ranges from sellers, not quotes.
| What you build | Published range | Typical timeline | Source |
|---|---|---|---|
| Clickable design prototype (no backend) | $2,000 to $5,000 | 1 to 2 weeks | EarlyVersion.ai |
| Coded prototype with 2 or 3 key features, AI-assisted | $5,000 to $15,000 | 2 to 4 weeks | EarlyVersion.ai |
| Production-ready prototype with logins and security | $15,000 to $50,000+ | 4 to 8 weeks | EarlyVersion.ai |
| Simple validation MVP | $30,000 to $55,000 | 6 to 8 weeks | Emerline |
| Web or PWA MVP | $25,000 to $55,000 | 4 to 8 weeks | Emerline |
| Mid-level transactional MVP | $60,000 to $120,000 | Not stated | Emerline |
| Native mobile MVP | $120,000 to $250,000+ | 16 to 24+ weeks | Emerline |
DEV.co, which sells both, puts its prototype work in fixed 1 to 3 week sprints (from one user flow and 5 to 7 screens up to 3 to 5 flows and 20+ screens) and its MVPs at 4 to 8 weeks.
Two things stand out. A clickable prototype costs a small fraction of an MVP, so it is cheap insurance against building the wrong screens. And the gap between a "coded prototype" and a "production-ready prototype" is almost all the invisible work (logins, permissions, security, hosting) that separates a demo from something customers can trust. For more on what drives MVP prices, see MVP development services, and for timelines by route, how long it takes to build an app.
Which Should You Build First?
Test only the risky piece, in days or a few weeks, before you spend on design or a full build. Skip it if your product uses well-understood building blocks.
A clickable prototype tested with five target customers. It is the cheapest way to fix confusing screens before they are built, and it doubles as the brief for your developer.
A prototype cannot tell you whether people will pay. If that is the open question, build the smallest working version, time-box it to weeks, and put it in front of paying customers.
For most first-time SaaS founders the order is: prototype the main flow, test it with five people, cut everything that is not needed to prove the idea, then build the MVP. Write down what you learned from the prototype as the brief; our guide to writing software requirements has a template.
When an AI-Built Prototype Becomes Your MVP
AI app builders such as Lovable, Bolt and Replit have changed the prototype step. A founder can now make a coded prototype in an afternoon that signs people in and saves data. That is useful, and it creates a new risk: the prototype quietly becomes the product, with real customers' data in it, without anyone checking it.
The risk is measurable. In Veracode's 2025 tests of more than 100 AI models, the generated code introduced security flaws in 45% of cases. A prototype does not need to be secure, because nobody's real data is in it. An MVP does.
Mistakes That Waste the First Version
Polishing the prototype
Weeks spent on pixel-perfect screens that will be rebuilt anyway. A prototype only needs to be real enough for people to react to it.
Asking people whether they like it
People are polite. Give them a task to do with the prototype and watch where they get stuck instead of asking for opinions.
An MVP with every feature
If the MVP takes six months, it is not minimum. Time-box it, list every feature, and cut anything the first ten customers can live without.
Treating the MVP as throwaway
Unlike a prototype, the MVP becomes version one. Cheap, messy code here is paid for in every feature after it.
Not owning the result
Whatever you build, keep the designs, code, hosting and data in your company's accounts, so you can change developers without starting again.
How StackTie Fits
StackTie is a development subscription for founders building a web app or SaaS product. We pick up where the prototype stops: you share your Figma file, your AI-built prototype or a plain-English description, and we turn it into a working MVP, one request at a time. Each request goes on your own Trello board, is worked in the order you set, and comes back within 2 business days as a pull request with a plain-English summary of what changed and why. A whole MVP is split into pieces that each ship on that schedule, so you can put the first screens in front of customers while the rest is built. If you already have an AI-built prototype, we can start by checking and fixing it rather than rebuilding it. One senior developer, Michael, leads every build. Everything we build belongs to your company, there is no contract, and there is a money-back guarantee. We do not build native iOS or Android apps, and we are not a whole team working in parallel. The flat monthly price and what is included are on the homepage.
