Most UK founders mix up MVP, prototype and proof of concept, then build the wrong one first. This guide explains what each stage proves, what it costs in the UK, where R&D tax relief applies, and how to sequence them so a year of runway is not wasted.
Masum Shamjad
Founder & CEO
May 21, 2026
Three engineers in a room, a founder with a deck, and a board that wants to see something working before Christmas. Someone says the word MVP. Someone else says a proof of concept would be safer.
By the end of the meeting the budget is approved for the wrong one. Most UK scale-ups make this scoping mistake at least once. The three approaches sound similar, but each answers a different question and belongs at a different point in the product timeline.
Pick the wrong one and you spend six months proving the wrong thing. This guide compares MVP vs Prototype vs PoC across what each proves, who you show it to, what it costs in the UK, and how R&D tax relief treats the spend. By the end you will know which to commission, in which order, for the question you actually need answered.
The MVP vs prototype vs PoC mistake we see most often is the same one. A founder skips the proof of concept and builds an MVP straight from a slide deck. Twelve weeks in, the team discovers the third-party API they assumed would work cannot do the one thing the product depends on.
The MVP gets rebuilt. The runway gets cut in half. The investors stop returning calls.
We see this pattern across UK fintech, healthtech and logistics teams almost every quarter. The technical risk was real, and a proof of concept would have surfaced it in two weeks. The MVP took six months to surface the same thing.
The mirror-image failure is just as common. A team builds a proof of concept, demos it to the board, and then ships it as if it were a product. The technology worked but no one wanted to pay for it.
CB Insights tracks this directly. Their analysis of post-mortem startup write-ups finds that 35% of failed startups cite "no market need" as the primary cause (CB Insights, 2021). Proving the technology works does not prove the market wants it.
So the choice between MVP, prototype and PoC is not a matter of taste. It is a matter of which question your business is most exposed to right now. That is what the next section helps you answer.
Most teams pick by instinct or by the loudest voice in the room. A structured choice between MVP, prototype and PoC needs five questions answered before the brief is written. Every project we have run starts with this set.
A proof of concept proves a specific technology can do a specific thing. A prototype proves a design is usable. An MVP proves customers will pay to use it.
A PoC is shown to a CTO or head of architecture inside the building. A prototype goes to design reviewers and a handful of target users. An MVP is shipped to the open market through paid acquisition or organic launch.
PoC budgets sit between £5,000 and £15,000 for a two to four week effort. Prototypes run £8,000 to £25,000 over three to six weeks. An MVP at the UK regional agency rate of £350 to £550 per day lands at £30,000 to £80,000 across three to six months.
A PoC almost never touches live users. A prototype is tested with a small group under controlled conditions. An MVP is exposed to live customers, real payments, and every compliance rule that comes with that exposure.
A PoC failure tells you the technology will not work and saves you the MVP build. A prototype failure tells you the experience needs another design pass. An MVP failure tells you the market does not want the product, which is the most expensive failure of the three.
Hold these five answers steady before any spend is committed. They map directly to the three stages, and the next section sets each one against the others.
The fastest read of MVP vs prototype vs PoC is the side-by-side comparison below. Use it as a reference; the deeper sections that follow profile each one against the same set of criteria.
Proof of Concept (PoC): proves the technology, is shown to internal engineers and CTOs, costs £5,000 to £15,000 in the UK over two to four weeks, and rarely touches real users.
Prototype: proves the design, is shown to design reviewers and target users in a test setting, costs £8,000 to £25,000 over three to six weeks, and touches real users only in controlled tests.
MVP: proves the market, ships to paying customers in the open market, costs £30,000 to £80,000 for a standard UK build over three to six months, and exposes the product to every real user from launch day.
The next three sections take each one in turn and explain what it actually contains.
A proof of concept is the lowest-cost answer to a single question: will this technology do the thing we need it to do? It ignores how the product looks. It exists to remove technical risk before anyone designs a screen.
A PoC is usually a script, a notebook, or a stripped-down service endpoint. It runs against the real data source, the real API, or the real hardware where the risk lives. The output is a yes-or-no demonstration to a technical reviewer.
Commission a PoC when a single technical assumption underpins the whole product. Examples we have built include classifying scanned warehouse labels at 99% accuracy with a vision model, running an open-source LLM inside an NHS firewall on the available GPU, and settling multi-currency payments within the latency a fintech market-making feature needs. If one technical answer can kill the project, prove it before anything else is built.
Expect £5,000 to £15,000 for two to four weeks of work from a UK agency. The brief should fit on a single page and name the one question being answered, the data or system the PoC will use, and the accuracy or performance threshold that counts as a pass. Anything broader than that turns the PoC into a small prototype, which doubles the cost without adding rigour.
Once the technology question is settled, the next risk to surface is whether real people can use what you intend to build. That is the prototype.
A prototype answers a different question from a PoC. It shows what the product will feel like to use, before you commit to a real build. The prototype vs MVP question often hinges on this difference: a prototype simulates the experience, while an MVP delivers it.
Prototypes are clickable design files, usually in Figma. Fidelity ranges from low (greyscale wireframes) through mid (styled but static screens) to high (interactive flows with real-feel transitions). High-fidelity prototypes can be indistinguishable from a real app inside a 90-second user test.
Commission a prototype when the core risk is whether users will understand or trust the experience. The signals are complex workflows, regulated industries where users are nervous (health, legal, finance), or radical changes to a familiar pattern. If three target users cannot complete the main task in a prototype, the MVP will not save them.
Expect £8,000 to £25,000 over three to six weeks. The brief should name the three user journeys you most want to test, the user profile you want feedback from, and the success criteria for each test (completion rate, time-to-task, qualitative confidence). Skip the brief and you receive a beautiful Figma file that no one has held against a real user.
A prototype that earns its keep tells you the experience works. It does not tell you anyone will pay for it. That is the MVP's job.
An MVP is the smallest live product that can answer one question: will real customers pay for this? A PoC proves capability; an MVP proves demand. Everything else flows from that one difference.
An MVP is a real product, on a real domain, with real payments. It contains only the features needed to deliver the core value proposition once. Every feature the team argues for must be cuttable; if it is not, the build is not minimum.
Commission an MVP when both the technology risk and the experience risk are low or retired. The remaining question is whether anyone will sign up, log in, and pay. Most UK founders we work with try to skip to this stage too fast and pay for it later.
A standard UK MVP lands at £30,000 to £80,000 over three to six months, based on the median UK developer day rate of £500 (ITJobsWatch). Complex multi-role MVPs in regulated industries reach £70,000 to £150,000. The brief should name the single customer outcome the MVP delivers, the acquisition channel the launch will use, and the metric that defines product-market fit for this product.
An MVP is not a beta or a soft launch. It is the leanest version of the real product, designed from day one to be paid for. Treat it as such or call it something else.
With each stage profiled, the next question is how they connect into a sequence for your business.
Sequencing is where most decision frameworks fail. The textbook order is PoC, then prototype, then MVP, then full product. Real projects rarely follow it that cleanly.
The order depends on which risk dominates. If the dominant risk is technical, start with a PoC; if it is experiential, start with a prototype; if it is commercial, build the smallest MVP that can charge a card. Pick the stage that retires the largest remaining unknown.
Some teams need all three. A deep-tech fintech building an LLM-backed compliance product would run a PoC on model accuracy first. Then a prototype to test how compliance officers want to interact with it, and finally an MVP for two paying customers.
Some teams need only one. A SaaS founder rebuilding a known workflow in a new niche does not need a PoC because the technology is settled. They may not need a prototype either if the workflow is well-understood, and an MVP at £35,000 can validate the market in eight weeks.
Sequencing aside, the cost of each stage is shaped by UK-specific factors that competitors writing on this topic almost never address.
Few comparisons of MVP vs prototype vs PoC look at what the UK tax code and the regulator do with each stage. They should. The answers change the real cost of every decision.
UK R&D tax relief under the merged RDEC scheme (from April 2024) offers a 20% taxable expenditure credit on qualifying spend (British Business Bank). PoC work qualifies most cleanly because it directly addresses scientific or technological uncertainty. MVP work qualifies in parts (the novel technical components, not the commercial polish), and pure design prototypes typically do not qualify on their own.
UK GDPR applies the moment personal data is processed (ICO). A PoC against synthetic data has no exposure, and a prototype tested with paid participants under an information notice has limited exposure. An MVP processing real customer data is fully in scope, which means a Data Protection Impact Assessment for any reasonable risk level.
In fintech, an MVP that handles regulated activities (payments, lending, insurance distribution) needs FCA authorisation or a sanctioned route such as the FCA Regulatory Sandbox. The same logic applies to healthcare products under MHRA classification, and to legal services under the SRA. PoCs in these sectors are the legitimate way to test before authorisation, not minimum products in the open market.
Tax and regulation shape what each stage is for. They also shape when you can compress the sequence and when you cannot.
Skipping stages is sometimes correct. Most of the time it is not. The MVP vs PoC vs prototype question often hinges on which skip is safe.
Safe to skip the PoC: the technology is mature, vendor-supplied, and used at scale by reference customers in your sector. Building a customer portal on top of Salesforce does not need a PoC. Building a real-time fraud model on a novel LLM stack does.
Safe to skip the prototype: the user journey is conventional, your designers have shipped the same pattern before, and the screens follow established conventions in the sector. A new niche-vertical CRM does not need a prototype if it copies the proven Salesforce-style flow.
Not safe to skip the PoC: novel AI, hardware integration, scale or latency challenges, or any dependency on a third-party system with limited documentation. We have seen six-figure MVPs scrapped because a payments provider returned a 0.4% edge-case error that a £7,000 PoC would have surfaced in week one.
Not safe to skip the prototype: any product that asks users to change a familiar workflow, accept a new mental model, or trust the platform with sensitive information. Trust is built through interaction, not specification.
Knowing what you can skip is half the discipline. The other half is what you check before any of the three stages starts.
Every brief we sign starts with the same checklist. It applies whether the project is an MVP, a prototype, or a PoC.
The single question: one sentence that states what the build will prove. If a stakeholder cannot state it, the project is not ready.
The pass/fail criteria: numeric thresholds for technical PoCs, completion-rate targets for prototypes, paid-conversion or activation metrics for MVPs.
The data source: the real system, dataset or user pool the work will run against. PoCs against synthetic data prove nothing real.
IP ownership at handover: the contract must confirm the client owns the code, design files and documentation produced. Ambiguity here costs more in a dispute than the build did originally.
NDA before disclosure: the agency or partner signs an NDA before any confidential commercial detail is shared. This is non-negotiable.
Project management line: 10% to 15% of total cost named separately, not absorbed into developer days. A PoC without a PM still needs scheduling and reporting.
Run a brief past this checklist before any agency starts work, regardless of the stage.
The MVP vs prototype vs PoC question is not about which one is best. It is about which question you are most exposed to right now. Technology risk lives in a PoC, experience risk in a prototype, and market risk in an MVP.
Pick the stage that retires the largest unknown. Sequence the rest only when the first one returns a clear answer. Skip a stage only when the risk it addresses is already retired by previous work.
If you want a second opinion on which stage your project actually needs, we run no-cost scoping conversations through our software development team. We have built PoCs, prototypes and MVPs across UK fintech, healthtech, retail and logistics for 14 years. Tell us the question, and we will tell you which stage answers it.
A PoC proves a specific technology can do a specific thing and is shown to engineers. A prototype proves the design is usable and is shown to a small group of target users. An MVP is a live product in the market that proves customers will pay.
Only when the MVP depends on technology that has never been built, never been integrated at scale, or never been used in your sector. If you are building on a mature stack (Stripe, Salesforce, AWS managed services) the PoC adds time without retiring real risk. If the MVP hinges on a novel model or a hardware integration, the PoC is the cheapest insurance you can buy.
A prototype is a Figma file or a clickable flow with no real backend, no real data and no real payments. An MVP is a live product on a real domain. The journey from prototype to MVP is a full build, not a conversion.
A standard UK MVP costs £30,000 to £80,000 over three to six months at the median UK developer day rate of £500 (ITJobsWatch). A complex multi-role MVP in a regulated sector reaches £70,000 to £150,000. Regional UK agencies sit at £350 to £550 per day; London agencies at £600 to £900 per day.
The PoC comes first when there is technical risk to retire. The prototype comes first when the technology is mature but the experience is unfamiliar. The two are rarely sequenced together, since most projects need one or the other plus the MVP that follows.
Contact Us
Get in touch with our team anytime today.
Our team is always here to listen, support, and guide you.
Stay informed with instant updates delivered straight to your inbox.
Providing top-notch technology solutions that you can trust to transform your growing business
Services
Get In Touch
Ahmedabad
714, Shivalik Shilp, Iscon Cross Road, SG Highway, Satellite, Ahmedabad – 380015, Gujarat, India.
+91 63532 74556
