Fixed price, against a written scope
We quote a fixed price for a defined phase after discovery. That means we carry the risk of it taking longer than we expected, and we price for that — typically a margin over our expected cost.
You get budget certainty and we get an incentive to specify carefully. A fixed price without a written scope is not a fixed price; it is a number that will be renegotiated.
The five answers that move the number
- How many systems does it touch? Each integration is design, build, error handling and testing.
- Do those systems have usable APIs? This changes a price more than any design decision we make.
- How many exceptions? The happy path is a fraction of the work.
- How many user roles? Each one multiplies the permission matrix and the testing.
- Who owns it internally? Projects without a decision-maker slip, and slippage costs money.
Typical ranges
| Scope | Price | Timeline |
|---|---|---|
| One process, one intake channel, one system | £5,000–£12,000 | 4–8 weeks |
| One process, two or three systems, real exceptions | £12,000–£25,000 | 8–12 weeks |
| Multi-process or awkward legacy interfaces | £25,000–£60,000 | 3–5 months |
What is included
- Specification and the exception design
- The build, with fortnightly demonstrations
- Exception handling and monitoring, not as an extra
- Parallel running against your manual process
- Documentation good enough for another developer to maintain it
- A 90-day warranty on defects after launch
Testing and exception handling are roughly a quarter of what you pay for and the first thing clients ask us to cut. We will not, because cutting it moves the cost to your first month live at a much higher rate.
What you can cut to reduce the price
Scope, honestly. One intake channel instead of three. One role instead of two. The admin interface deferred while a developer runs queries. Historical data left where it is.
Each of those reduces the price without affecting whether the thing works, which is the distinction that matters.