Insights
How much does it cost to build an MVP in the UK?
By Rare Agentics engineering teamLast updated First published 8 minute read
There is no honest single price for an MVP, because the cost depends on what it has to do: the number of user journeys, the systems it connects to, how sensitive the data is and how much has to be decided before work starts. With Rare Agentics a first product typically takes 1 to 8 weeks at a fixed price agreed in writing, and the smallest paid step, a proof sprint, is from £3,000. Budget separately for hosting, third-party services and support after launch.
What is an MVP actually supposed to be?
A minimum viable product is the smallest version of your product that real users can use for real, and that tells you whether the idea is worth more money. It is not a demo, and it is not the whole product with the corners cut.
That definition matters for cost, because most MVP budgets go wrong before a line of code is written. The scope grows from "the smallest thing that proves it" into "everything we will eventually need", and the price grows with it. The cheapest MVP is the one with the fewest journeys that still answers the question you are asking.
A useful test: write down the one thing you need to learn from the MVP. "Will clinics pay to book patients online?" is a question an MVP can answer. "Can we build a platform for clinics?" is not, and an MVP scoped around it will cost more and teach you less.
What actually drives the cost of an MVP?
The price is set by the work, not by the label. Six things move it more than anything else.
User journeys. Each distinct thing a user does (sign up, book, pay, approve, report) is design, build and testing work. Five journeys cost far less than twenty.
Roles and permissions. One kind of user is simple. Customers, staff, managers and administrators who each see different data is a real piece of engineering, and it is where security problems hide if it is rushed.
Integrations. Every outside system (payments, accounting, email, a CRM, a booking tool) needs authentication, error handling and a plan for when it is slow or down. Systems with good APIs are quick to connect; systems without one can be the largest line in the budget.
Data sensitivity. Personal, health or financial data brings UK GDPR obligations, tighter access rules, audit trails and more testing. It is money well spent, but it is not free.
Decisions still open. If the workflow, the pricing model or the rules are still being decided, someone pays for the time it takes to decide them. Usually that is you, through changes mid-build.
Platforms. A web app that works well on a phone is one build. Native iPhone and Android apps on top are more.
| Cost driver | Pushes the cost down | Pushes the cost up |
|---|---|---|
| User journeys | A handful, each clearly defined | Many, with variations for each user type |
| Roles | One or two | Several, each seeing different data |
| Integrations | None, or one with a good API | Several, some with no API |
| Data | Non-personal, low risk | Personal, health or payment data |
| Decisions | Made and written down | Still being worked out during the build |
| Platforms | A responsive web app | Web plus native mobile apps |
If a supplier gives you a price without asking about most of those rows, the price is a guess. A guess is usually either padded to protect them or low to win the work, and you pay for both eventually.
How much does a first product cost with Rare Agentics?
A first product with us typically takes 1 to 8 weeks, at a fixed price agreed in writing before work starts. We do not print a single MVP figure, because the drivers above vary too much from one product to the next, and a number that ignored them would be wrong for most people reading this.
What we do print is the smallest paid step. A proof sprint starts from £3,000 and delivers one bounded feature, workflow or working prototype, live in your own accounts in 1 to 2 weeks. It is a deliberate first step, not a whole MVP: it proves the riskiest part of the idea and the working relationship before either of us commits to more. Every engagement and its price is set out on our pricing page.
Scoping a build you have in mind is free. Within two working days you get a first view, and within a week a written plan, a fixed price and the risks. You approve the plan and the price before the build starts, and any change to scope is re-scoped and re-priced in writing before it is built.
Larger products are delivered as a platform build, in phases of 6 to 8 weeks, each priced before it starts. You can stop after any phase and keep everything built so far.
What should an MVP quote include, and what does it often leave out?
The headline number on a quote is only useful if you know what sits behind it. A quote worth signing names all of the following.
- The journeys in scope, listed one by one, and the ones explicitly out of scope.
- Acceptance criteria for each: what "done" means in terms you can check yourself.
- Environments: a staging copy to test on and a production environment, in your own accounts.
- Testing, including tests for the rules that must never break, such as one customer never seeing another's data.
- Security basics: access rules, keys kept server-side, rate limits on sign-in and forms.
- Ownership and documentation: repositories and accounts in your name from the first commit, and documentation another team could work from.
- What happens after launch: a warranty period, and what support costs after it.
The items most often missing are the ones that only bite later: data migration from your current spreadsheets, admin screens for your own team, email deliverability, privacy notices and consent records, and the time it takes to get an app approved in the app stores. Ask about each by name.
Every build we deliver carries the launch warranty: any defect against the agreed acceptance criteria, found in the 30 days after release, is fixed at no cost.
Fixed price or day rate: which protects you?
Both models can work. They put the risk in different places.
| Fixed price | Day rate | |
|---|---|---|
| Who carries the estimating risk | The supplier | You |
| What you know before you start | The total for the agreed scope | The rate, and an estimate |
| What happens when scope changes | Re-scoped and re-priced in writing | The meter keeps running |
| Best for | A defined first product or phase | Open-ended exploration with no clear end |
| What to watch for | A vague scope that invites disputes | Estimates that drift week by week |
For an MVP, a fixed price is usually the better deal, as long as the scope is written down precisely. The discipline of agreeing what is in and out before work starts is itself worth money: it is the moment most budgets are saved.
A day rate makes sense when you genuinely cannot define the work yet. In that case, a short, fixed-price discovery or a proof sprint on the riskiest piece is often cheaper than an open-ended engagement. We cover how to tell the difference in what a discovery phase is, and whether you need one.
What will the MVP cost to run after launch?
The build is one line in the budget. Running it is another, and it starts the day you go live.
- Hosting and databases. Billed by the providers to your own cloud accounts, at their prices. For an early product with modest traffic this is usually a small monthly cost, but ask for an estimate at your expected usage.
- Third-party services. Payments, email, maps, identity checks and AI models each have their own pricing. Some are free at low volume and rise sharply as you grow.
- Support and changes. Bugs outside the warranty, small improvements, dependency updates and security patches. Our Run and improve plan covers hosting, monitoring, tuning and support with agreed response times, quoted to scope.
- Your own time. Someone has to read user feedback, decide what changes and approve releases.
Because everything runs in accounts you own, every one of these bills is visible to you. If a supplier hosts the product in their own account and charges you a single monthly fee, ask what is in it and what it would cost to move.
What are the red flags in an MVP proposal?
A few patterns are worth walking away from, or at least pushing back on.
- A price before any questions. Nobody can price an MVP they have not understood.
- No written list of what is out of scope. Everything not listed becomes an argument later.
- Code in the supplier's repositories, or hosting in their accounts. You should own both from the first commit. We cover this in how to make sure you own your code and data.
- Per-seat or licence fees for software built for you. You are paying for the build; you should not keep paying to use it.
- No mention of testing or security. An MVP with real users and real data needs both, however small it is.
- A long gap before you see anything working. You should see working software every week, not only at the end.
How should you plan your MVP budget?
Before you ask anyone for a price, work through this list. It will make every quote you receive more accurate and easier to compare.
- Write the one question the MVP must answer.
- List the journeys a user must complete to answer it, and cut the rest.
- Name every kind of user and what each may see and do.
- List the outside systems it must connect to, and check whether each has an API.
- Decide what data it holds, and whether any of it is personal, health or payment data.
- Decide whether a responsive web app is enough for the first version.
- Note the decisions still open, and who will make them.
- Set aside a budget for running costs for the first twelve months.
- Ask each supplier for a fixed price for the agreed scope, the list of what is out of scope, and who owns the code and the accounts.
If you have built a prototype already, including one made with an AI tool such as Lovable, Bolt, Cursor or Replit, say so. It can shorten the work, and it changes what the first step should be. If it already holds real user data, a security review for AI-built apps may be the right first step. Our comparison of no-code and custom builds covers when to keep it and when to rebuild.
How Rare Agentics can help
We build bespoke software for founders and growing businesses: senior engineers lead every build and AI agents do the heavy lifting. A first product typically takes 1 to 8 weeks at a fixed price agreed in writing, the code lives in your repositories and your data in your cloud accounts from the first commit, and a person signs off every release.
If you want to start small, a proof sprint from £3,000 puts one working thing live in your accounts in 1 to 2 weeks. Scoping a build you have in mind is free. Tell us what you want to build, and within two working days you get a first view, and within a week a written plan, a fixed price and the risks.