SaaS MVP development cost in the UK, and what it buys

A SaaS MVP in the UK typically costs somewhere between £8,000 and £80,000 to build, on the ranges agencies publish, depending mostly on how much it has to do. One type of user, a few core features and a way to take payment sits at the bottom. Several roles, integrations and an admin area sit in the middle. Real-time data, several platforms or regulated data push it well past that. What moves the number most isn’t anyone’s rate; it’s how many features the first version tries to include.

This piece sets out the published ranges, what a first version actually needs, what it costs each month once it’s live, using the vendors’ own pricing, and how to cut scope without cutting the things that force a rebuild. My own build work is priced per project on what it’s worth, so there’s no rate card here.

What the market publishes

There’s no survey of UK MVP prices, only agencies publishing their own bands. Two recent ones agree on the shape, if not every number.

Why the ranges overlap

TulipTech defines a simple MVP as one platform, three to five core features and no integrations beyond login and payment; its standard tier adds several user roles and light integrations with CRM, email and analytics. Quantel calls the size of the feature set the single biggest driver. Both are describing the same thing: price follows what the product has to do.

Regulated products are different

TulipTech puts products with serious compliance needs, such as healthcare, fintech, or anything touching FCA rules, NHS data, ICO obligations or accessibility requirements, at £150,000 and up. The build isn’t much bigger; the evidence, testing and controls around it are.

Timelines matter as much as cost

Eight to twelve weeks for a simple MVP is a long time to wait for proof that anyone wants the product. The shorter the build, the sooner you learn, which is the whole point of an MVP.

What an MVP actually needs

A minimum viable product has one job: prove that people will use and pay for the core idea. Everything else can wait.

The core feature, done properly

The one thing the product does that nobody else does, well enough that a paying customer would miss it. This is where most of the budget should go.

Accounts and login

Sign-up, login, password reset and basic account settings. It’s not the product, and it should come from a proven library or service rather than be written from scratch.

Taking payment

Subscriptions, invoices and failed-payment handling. Stripe Billing does most of this out of the box; building it yourself is rarely worth it at MVP stage.

A basic admin view

Enough to see customers, fix their problems and turn accounts on and off. It can be plain. It can’t be missing, because the first customers will need help.

What it doesn’t need yet

Several user roles, a mobile app, integrations with every tool a customer might use, dashboards, and settings for cases nobody has hit yet. Each is a reasonable second-version feature and an expensive first-version guess.

What it costs to run

Running a SaaS MVP is cheap compared with building it. These are the published prices for a common set-up; there are good alternatives to each.

Hosting and database

Vercel’s pricing puts its Pro plan at $20 a month and says the free Hobby plan is for personal, non-commercial use, so a paying product belongs on Pro. Supabase starts its Pro plan from $25 a month, with 8 GB of database storage and 100,000 monthly active users included, and pauses free projects after a week of inactivity.

Payments

Stripe Billing charges 0.7% of billing volume on pay-as-you-go, on top of card fees of 1.5% plus 20p for standard UK cards. For an MVP with 100 customers paying £100 a month, that’s about £70 for Billing and about £170 in card fees: roughly £240 a month on £10,000 of revenue.

Everything else

Email sending, error tracking, a domain, and whatever the product itself needs. Most have free tiers that cover an MVP’s first customers. Add them up before launch, and look at what each costs once you pass its free limit.

What pushes the cost up

Most overruns come from a short list of decisions made early.

More than one kind of user

Every role, such as admin, manager, member and customer, multiplies the screens, permissions and tests. TulipTech’s jump from its simple to its standard tier is mostly this.

Integrations

Connecting to a customer’s CRM, accounting system or data source means handling someone else’s data, errors and limits. One integration is manageable; five is a project of its own.

Several customers on one system

A product that keeps each customer’s data separate, with their own users and settings, needs that designed in from the start. It’s far cheaper to build in on day one than to retrofit.

Compliance

Personal data, payments, health or financial data bring obligations that shape the build: where data is stored, who can see it, how it’s logged and deleted. They’re cheaper to design for than to add.

A mobile app as well as the web

Two platforms is close to two builds. For most B2B SaaS, a web app that works well on a phone is the right first version.

Cut scope, not corners

There are two ways to spend less. One works; the other costs more later.

Do some of it by hand

Plenty of good products started with the founder doing part of the work behind the scenes: sending reports, onboarding customers, processing a step manually. Automate it once you know people want it.

Why corners cost more

An MVP built without the basics works until the first real customers arrive, then needs rebuilding just as it starts to earn. Rescuing a build like that, a stalled build rescue, often costs more than building it properly would have.

Technical leadership

If there’s no technical founder, someone still has to make the platform and architecture decisions. What a fractional CTO costs covers the options.

How I price a build

I don’t have a rate card or a day rate. Each build is scoped around what the product has to prove, priced on what that’s worth, and agreed in writing with a deposit before work starts. I build in Laravel, Django and Next.js, and the bespoke builds page covers how.

Getting a price

A short conversation about what the first version has to do is enough to scope it, and to say whether a no-code tool would do the job better. Book a free call.