Prototype, proof of concept or MVP? What each tests, UK costs (prototype free, PoC £2k–£10k+, MVP from £10k), timelines and which to build first.
Short answer. A prototype tests whether people understand and want the idea; a proof of concept (PoC) tests whether the technology can do the job; an MVP tests whether the market will use and pay for a real product. Most founders should start with a prototype, add a PoC only when there is real technical risk, and then build the MVP. At Magora a working prototype is free (5–10 working days), a PoC costs £2,000–£5,000 (or from £10,000 for an AI PoC on real data), and an MVP starts from £10,000 in 6–8 weeks.
This guide is for founders and product owners deciding how to spend the first part of their budget. It compares the three, explains which question each one answers, and gives the order, costs and mistakes to avoid.
Last checked: 2 November 2026.
Prototype, PoC and MVP compared
| Prototype | Proof of concept | MVP | |
|---|---|---|---|
| Question it answers | Do people understand it and want it? | Can the technology do what the product needs? | Will people use it and pay for it? |
| What it is | A clickable model of the key screens, no real backend | A narrow technical spike: a model, integration or data source tested | The smallest real product users can sign up to and use, in production |
| Who it convinces | Users, co-founders, early investors | You, your CTO, technical due diligence | The market, and investors who want traction |
| Typical time | 5–10 working days | 2–4 weeks; AI PoC 4–6 weeks | 6–8 weeks after Discovery |
| Magora price | Free | £2,000–£5,000; AI PoC from £10,000 | From £10,000, most £10,000–£25,000 |
| What happens to it | Guides design; not production code | Informs architecture; often thrown away | Becomes version one of the product |
Magora 2026 prices excluding VAT, as published on our MVP development and AI proof of concept pages.
Which one do you need?
- Is the idea still being shaped? Build a prototype and put it in front of ten target users. It is the cheapest way to find out what is wrong with the idea.
- Is there a technical question nobody can answer confidently? For example, whether a model can read your documents accurately enough, or whether a third-party API can deliver the data in time. Run a PoC on that question only.
- Do you know what to build and who for? Build the MVP: the smallest product that real users can use and pay for, with the analytics to see what they do.
Not every project needs all three. A booking or marketplace product with standard technology rarely needs a PoC. An AI product almost always does, because accuracy on your own data is the main risk.
The usual sequence and timeline
| Stage | Time | Output |
|---|---|---|
| Working prototype | 5–10 working days | A clickable core flow to test with users and show investors |
| Discovery | 2 weeks, from £3,000 | Prioritised feature list, architecture, the one metric the MVP has to move, and a fixed-price quote |
| Proof of concept, if needed | 2–6 weeks | Measured answer to the technical question, and a go or no-go |
| MVP build and launch | 6–8 weeks | A production product with real users and analytics |
| Measure and iterate | Ongoing | The next release planned from real usage, not the original wish list |
Roughly two to three months from first call to live users, with something you can click at every stage. More detail in how to build an MVP.
What investors want to see
A prototype explains the product and the user journey, which helps at the idea stage. A proof of concept answers a technical due diligence question. Usage data from a live MVP is the strongest evidence of demand, because it shows what people do rather than what they say. If you are raising, ask your investors which question they need answered before you choose what to build. UK founders can also look at Innovate UK grants for software projects.
We help founders scope, build and launch an MVP real users can try, starting with a free working prototype in 5–10 days. First release in 6–8 weeks.
Examples from our work
- LYFE (health, 2023): a direct-to-consumer web app for home sexual-health test kits, built as a no-code MVP on Bubble in the standard £10,000–£25,000 band.
- BodySync (healthcare AI, 2025): an AI MVP that drafts physiotherapy treatment plans, deployed in UK clinics; above the standard band, from £40,000, because of the AI.
- Retrofit Navigator (energy grants, 2026): a phased MVP of about £19,000 over 12 weeks, with the full scope at about £31,000 over 20 weeks (indicative, excluding VAT).
Mistakes to avoid
- Building the MVP before testing the idea. A free prototype finds the obvious problems for nothing.
- Calling a demo an MVP. If users cannot sign up and use it for real, it does not test the market.
- A PoC that tests everything. Keep it to the one technical question that could kill the product.
- A throwaway MVP. Build it production-grade so a good result does not trigger a rewrite.
- Too many features. The MVP should answer one question; everything else waits for the second release.
Have an idea to test? Send us a short brief: we will build a free working prototype of the core flow in 5–10 working days, yours to keep whether or not you build with us. More about our MVP development.
Frequently asked questions
What is the difference between a prototype, a proof of concept and an MVP?
A prototype is a clickable model that tests whether people understand and want the idea. A proof of concept tests whether a piece of technology can do the job. An MVP is the smallest real product users can sign up to, use and pay for; it runs in production and becomes version one of the product.
Which should I build first?
Usually a prototype, because it is the cheapest way to test the idea with users and investors. Add a proof of concept only if there is real technical risk, such as whether an AI model is accurate enough on your data. Build the MVP when you know what to build and need real usage data.
How much does a prototype, a PoC and an MVP cost?
At Magora a working prototype is free and takes 5–10 working days. A proof of concept costs £2,000–£5,000 over 2–4 weeks, or from £10,000 over 4–6 weeks for an AI proof of concept on real data. An MVP starts from £10,000 and takes 6–8 weeks; most cost £10,000–£25,000.
Do investors want a prototype or an MVP?
It depends on the stage and the investor. A prototype helps explain the idea and the user journey; an MVP produces real usage data, which is the strongest evidence of demand. A proof of concept answers the technical question, if there is one.
What is the difference between an MVP and an MMP?
An MVP is the smallest product that tests whether people will use and pay for it. A minimum marketable product (MMP) is the first version you are ready to market widely, usually the MVP plus what early users showed was missing.
Can a prototype become the MVP?
A clickable prototype usually cannot: it has no real backend. A production-grade MVP, on the other hand, should become version one of the product without a rewrite.
Pavel Shevtsov
Pavel Shevtsov is a seasoned professional in the field of software development, currently serving as the Managing Director at Magora, a leading software development company known for its innovative solutions. With a passion for leveraging technology to solve complex business challenges, Pavel has played a pivotal role in Magora's growth and success.
LinkedIn