The forty-thousand-pound MVP is a myth in 2026

A founder asked me last week, in a perfectly reasonable email, what a "fair" budget was for an MVP. He had been quoted £40,000 by a perfectly reasonable agency. He wanted a second opinion before he wrote the cheque.

The short answer is that £40,000 is no longer the right number, and the agency quoting it is either operating on a 2022 cost model or deliberately leaving margin on the table because the buyer does not know any better. The honest 2026 answer is much lower, with caveats. This essay is the actual breakdown.

What an "MVP" actually is

Before we get to the number, we need to be honest about what we are pricing.

The word MVP has drifted. In 2010 it meant the minimum interesting thing you could put in front of a user, often a clickable Figma or a Wizard-of-Oz hack. In 2015 it had drifted to a "thin slice of real product, deployed." By 2020 it had drifted again to "a functioning version one of the product, with auth, billing, and at least one happy-path user journey." And by 2024 some agencies were quoting "MVP" prices that included two or three flagship features, integrations, and a basic admin panel — essentially version two of the product.

This drift is why MVP pricing is now incoherent in the market. The same word means very different things to different vendors.

For this essay I am going to use a specific, narrow definition: an MVP is the deployed first useful version of the product, with authentication, persistence, one billable workflow, basic email notifications, an admin/operator view, and observability sufficient to debug a paying customer. This is the version-one definition. It is what most early-stage founders mean when they say MVP, even if their vendor heard "version-three" when they read the brief.

What it actually costs to build in 2026

For an AI-native version-one product built by a single senior operator with the modern toolchain, working full-time, the honest 2026 budget is:

• Build time: 4 to 8 weeks. Some products are 2 weeks. Some are 12. The mode for the products we have worked on, defined as above, lands in the 4-8 week band.
• Senior operator cost: £8,000 to £20,000. Calculated at a £200-£250 day rate, which is what a competent senior operator running AI tooling is genuinely worth on a fixed-scope project. Lower than the agency rate, higher than the freelance market floor, calibrated to actual output.
• Tooling: £400 to £900 total. Cursor or Claude subscriptions, AI model API spend during build, hosting for the first three months, domain, transactional email, error tracker. This number tends to surprise people; it is small.
• Design: £0 to £4,000. Most products at this stage do not need bespoke design. A Tailwind UI license, a competent eye, and an evening with Figma if you absolutely need a single hero illustration. If you need a real designer, £4,000 buys five days of senior work, which is more than enough for a version-one product if scoped tightly.
• Legal: £600 to £1,500. Terms, privacy, basic contract templates. Not strictly part of the MVP build but most founders forget it. If you skip this you pay more later.

The honest top-line is therefore £10,000 to £25,000 for a properly-scoped version-one AI-native MVP in 2026, with the modal number sitting around £15,000.

That is a long way from £40,000.

Where £40,000 actually goes

Now, in fairness to the agency quoting £40,000: it is not necessarily mispriced for what they are delivering. The number is reasonable if you assume:

• A team rather than an operator. A typical agency staffs a project with a project manager, a designer, a developer, a QA. Four heads on the project carries four times the cost.
• Days lost to internal process. Stand-ups, retro, account check-ins, ticket triage, status reports to the client. In a small operator-led build none of this exists; in an agency build all of it gets billed.
• A traditional waterfall scope. Detailed Figma in advance, sign-off cycles, change requests handled formally. None of this is wrong, but all of it is expensive.
• Margin. Agencies are commercial businesses and a 30-40% margin on the line items is normal. We are not criticising it; it just means the buyer is paying for it.

So the agency quote is internally consistent with what an agency is set up to deliver. It is just a much more expensive way of getting to the same outcome than the alternatives now available, and an increasing share of buyers are noticing.

What you give up at the lower number

The honest version of any pricing essay is what you give up by paying less. So:

• You give up the agency wrapper. No account manager. No ticket system. No weekly status report. If you need that to feel comfortable, the agency rate is what you are paying for.
• You give up the team. A single senior operator can ship the version-one product. They cannot in parallel staff a marketing site, a customer success function, and a sales process. If you need all four at once, you need more than one human, and the cost goes up correspondingly.
• You give up the "neck to wring". Bigger agencies offer a kind of insurance: if it goes wrong, there is a company you can sue. A solo senior operator carries different risk. In practice the failure rate of competent solo operators on version-one builds is lower than the failure rate of mid-tier agencies, but the perception of risk runs the other way.
• You may give up the design polish. A version-one product built in eight weeks by a senior operator is not going to win a design award. It is going to be functional, clean, and convertible. If your version-one product needs to be visually stunning — luxury verticals, fashion, certain consumer cases — bump the design line by an order of magnitude and the total budget reflects that.

For most B2B SaaS, internal tools, AI-flavoured workflows, and operator-facing products, none of these trade-offs is a problem. The product will be functional, ship in eight weeks, and let you start learning what the real product needs to be.

A word on tech-for-equity

The reason we ended up running the Moonlabs Incubator the way we do is precisely the maths above. The actual cost of building a version-one product for a founder we believe in is in the range of £15,000 of our time. A 1% equity slice in a company with even a modest exit story is worth dramatically more than that, in expectation. The arithmetic works for us because the production cost has fallen, and it works for the founder because they keep 99% of the company instead of paying £40,000 of post-tax cash for the same outcome.

This is not a clever pricing trick. It is the only price the new economics allows. Agencies still quoting £40,000 in cash for an MVP are leaving the upside on the table for someone else, and that someone else is us.

What this means if you are about to write the cheque

If you are about to write a £40,000 cheque for an MVP in 2026, three honest questions before you sign.

One: what exactly is in scope, and would the same scope be possible with a single senior operator in eight weeks at £200 a day? Most of the time the answer is yes and you have just identified £20,000 of fat in the quote.

Two: do you need a team, or do you need one good operator? If your project genuinely needs four heads in parallel — multi-region launch, four flagship features at once, accelerated timeline — pay for the team. Most projects do not.

Three: can you get the same outcome with tech-for-equity? If your company has a credible exit story and the vendor genuinely believes in it, equity-based deals are increasingly available, and they align incentives much better than a cash invoice. We are not the only people offering this; there is a small but growing market of operator-led incubators on similar terms.

The £40,000 MVP is not a scam. It is just a 2022 number being quoted in 2026, and the founders who run the maths will spend half that on a better outcome.

Louis O'Connell-Bristow is co-founder of Moonlabs, the operator-led AI incubator. The Moonlabs Incubator takes 1% tech-for-equity plus 10% on funds raised. The Forward Deployment service offers fixed-scope, transparent-rate engagements for buyers who prefer to pay in cash.