"How much will it cost to build?" is the first question every founder asks, and the honest answer depends on things nobody knows on a first call. So here are the real ranges, what moves them, and how to keep the number down without building something you'll throw away.
A focused MVP, built properly by a UK team, usually lands between £15,000 and £40,000. A single-purpose tool can start around £5,000. Anything with complex roles, integrations or data will be more. These are typical ranges, not quotes: the only honest number is one based on your scope.
What "properly" means
An MVP is the smallest product that real customers will use and, ideally, pay for. It isn't a prototype or a clickable demo. "Properly" means it has:
- Real accounts with sign-up, login and password reset that work
- A database that keeps each customer's data private from the others
- Payments, if you're charging, that the customer can't tamper with
- A design people trust, because people don't hand their card details to something that looks unfinished
- Somewhere to run it, with backups, monitoring and a way to deploy changes safely
You can get a convincing demo for far less, and Lovable or Bolt will produce one in a weekend. The gap between that and the list above is where most of the cost of a real MVP goes. We wrote about that gap here.
What moves the price
The number of user types. One kind of user is simple. Customers, admins, team members with different permissions and a back office for your staff multiplies the screens and the rules.
Integrations. Every third-party system (accounting, CRM, a marketplace API, an AI model) adds build time and, more importantly, testing time, because you don't control the other end.
Payments complexity. One subscription price through Stripe is straightforward. Usage-based billing, trials, discounts, invoicing and VAT across countries are not.
Data that has to be right. Anything financial, medical or legal needs more care, more testing and more thought about who can see what.
Design ambition. A clean, consistent interface built from a good system is affordable. A highly custom, animated experience is a different budget.
What you can safely leave out of version one
The cheapest way to build an MVP is to build less of it. Things you can usually leave for later:
- An admin panel. Use the database dashboard, or a ready-made admin tool, until the same task is costing you hours a week.
- Every edge of billing. Start with one plan and a Stripe-hosted checkout. Add annual pricing, coupons and seats when customers ask.
- Native mobile apps. A responsive web app works on every phone and costs a fraction to build and maintain.
- Automation of rare tasks. If something happens once a month, do it by hand until it doesn't.
- Settings and preferences. Make sensible decisions for the customer and add options when people actually ask for them.
What you can't leave out: security, data privacy, payments that can't be manipulated, and backups. Those are cheap to do from the start and expensive to retrofit after something goes wrong.
Three ways to pay for it
A fixed-price project
An agency scopes the MVP and quotes a fixed price. You know the number up front, which is genuinely useful when you're raising or budgeting.
The catch is that an MVP is the product most likely to change while it's being built. You learn things from the first users, and every change becomes a change request. Fixed price works best when the scope really is fixed, and an MVP rarely is.
A freelancer
Experienced UK freelancers commonly charge £350 to £600 a day. A three-month build at three days a week comes to roughly £14,000 to £23,000, before design.
It can be the cheapest route. The risks are that one person is designing, building and deciding all at once, and that the knowledge leaves with them when the project ends.
A product team on subscription
This is what we sell, so weigh it accordingly. You pay a flat monthly fee, add work to a queue, and a designer and developer ship it one request at a time. Ours is £4,500 a month, with founding places at £3,500 a month for the first six months.
As an illustration, four months on a founding place is £14,000, design included, with no change requests because changing your mind just means reordering the queue. When the MVP is live you can keep going with the next features, pause, or cancel. The full model and the maths are here.
The trade-off is that you don't get a fixed total up front. You get a fixed monthly cost and control over what gets built next, which suits founders who expect to learn as they go.
How to keep the cost down, whichever you choose
Write down the one thing it has to do. Not the vision, the job. If your MVP does one job well, it can be small.
Prototype first. Use Lovable, Bolt or a Figma prototype to test the idea with real people before paying anyone to engineer it. If you already have a prototype, a developer can often build on it rather than starting again.
Decide quickly. The biggest hidden cost in any build is waiting: for feedback, for content, for a decision. A founder who replies within a day can halve the timeline.
Keep it boring underneath. Mainstream, well-supported tools (we use Laravel) mean any competent developer can pick it up later. Exotic stacks are expensive to hire for.
Plan for version two. The MVP is the start of the cost, not the whole of it. Budget for the months after launch, when real users start asking for things.
Where to start
If you have a prototype already, our two-week kickoff sprint is £1,950: one defined piece of work, designed and shipped, often enough to take a working prototype to something you can put customers on. Subscribe within 30 days and it comes off your first month.
If you're comparing options, see how the product team works, read what it costs to hire a developer instead, or book thirty minutes. We'll tell you honestly what your MVP needs, even if the answer is a smaller build than you expected.
