The hourly rate on a freelancer's profile is the most visible number in the whole decision — and usually the smallest. What actually determines the cost of a build is everything stacked on top of that rate: the marketplace's cut, the hours you spend vetting and managing, the rework when something misses the spec, and the tail risk that the project is abandoned and you pay a second developer to finish it. This guide builds an honest total-cost-of-ownership model, then works through a thinking example against a fixed MVP build.
The headline hourly rate on Upwork is the smallest part of what a build costs you. Add the marketplace fee, the unpaid hours you spend screening and managing, the probability and cost of rework, and the risk that the project is abandoned mid-build and a second developer has to finish it — and a rate that looked like a fraction of any agency quote frequently lands at or above a fixed £12,000–£30,000 MVP build. Upwork genuinely wins for small, well-specified, low-risk tasks; on a whole product, model the total, not the rate.
Hiring is not a purchase, it is an ongoing cost. Total cost of ownership (TCO) is the discipline of counting everything a decision costs over its life, not just the sticker price. On a marketplace the sticker price is the hourly rate — and it is deceptively low, because five other costs are hidden behind it. None of them appears on the profile.
Add those to the rate and you have the real number. The point of this list is not that Upwork is bad value — it is that the rate alone tells you almost nothing about what the build will cost.
Suppose you find a capable freelancer at, say, £25/hour — a deliberately illustrative figure, not a market average. Against a fixed £12,000–£30,000 MVP band it looks like a fraction of the price. Then run the logic of the TCO model, layer by layer, and watch the gap close.
| Cost layer | The £25/hour freelance path | A fixed team build |
|---|---|---|
| Headline price | £25/hour — looks like a fraction of any agency quote | One fixed band, e.g. £12,000–£30,000 for an MVP |
| Marketplace / platform fee | A meaningful cut added on top (check the current schedule) | None — you contract directly |
| Your vetting & interviewing | Hours of your own time shortlisting and interviewing — unpaid, but real | One scoping call with a senior engineer |
| Ongoing management | You are the project manager: specs, reviews, coordination, chasing | Managed for you inside the fixed price |
| Rework | A reasonable probability of redoing work that misses the spec | Tested code built against an agreed spec |
| Abandonment tail risk | If the build stalls, a second developer to finish — often the largest single cost | Continuity: one team owns delivery to the end |
| Effective total | Frequently climbs to, or past, the fixed band — with less certainty of finishing | Known before you start |
Prices published from our Open Price Book (v1.0 · July 2026 · next review October 2026). All prices exclude VAT.
Run the logic, not a spreadsheet. Take the £25/hour as your starting point — it is genuinely lower than any senior team's effective rate, for the first hour. Then add the marketplace's cut. Add the unpaid hours you spend screening candidates and, later, reviewing and re-specifying work. Add the rework when something comes back wrong. And weight the whole thing by the chance — never zero on a larger build — that you end up hiring a second developer to finish, paying twice for one outcome. Each layer is modest on its own; stacked, they routinely turn a rate that looked like a third of the agency price into a total that meets or beats it. The difference is that the fixed £12,000–£30,000 band is a number you know on day one, and the freelance total is a number you only learn at the end.
If a freelance build has already stalled, finishing it is usually cheaper than starting over — see how an app rescue takeover works.
None of this makes Upwork a bad platform — it makes it the right tool for a specific job. When the work is small, well-specified and low-risk, the hidden costs barely apply: a defined bug fix, a landing page, a one-off script, a data clean-up, a design asset. The scope is clear, the review takes minutes, and there is nothing to abandon.
Upwork's escrow is real buyer protection, too — as a plain factual description, it holds your funds with a neutral third party and releases them when you accept the work, which caps your downside on a marketplace contract. The trouble is only that escrow protects the payment, not the outcome: it cannot make an under-specified product finish, and it does not refund the weeks you spent managing a build that went nowhere.
The moment the work becomes a whole product — vague at the edges, months long, central to the business — the TCO model takes over and the rate stops being the cost.
Genuinely. A template build or a single-gig freelancer is the right tool at this budget — an agency would be a bad deal for you.
Workable with one vetted senior freelancer — keep the repo in your name from day one and check references properly.
At this scope you need accountability, continuity and tested code — a senior team with published prices and milestone billing.
The wedge between the middle and right-hand column is where the hidden costs bite hardest. If you are weighing the two, our Upwork vs a software agency comparison lays out the trade-offs side by side, and the Open Price Book shows exactly what our bands cover — no "from £X" games, just published, versioned prices with milestone billing so you pay for delivered, accepted work.
Book a free 30-minute scoping call with a senior engineer — in your business hours. We'll size the work against our published bands and set milestone billing so you only ever pay for delivered, accepted work.
Book a free scoping call →