Two ways to make work go away
You can pay someone else to do it every month, or you can build something that does it. Both are legitimate; they suit different work.
The comparison is rarely made properly because a retainer is an annual number that recurs quietly and a build is a one-off number that lands with a thud.
Three years, side by side
| Retainer | Build | |
|---|---|---|
| Year 1 | £1,000/month = £12,000 | £12,000 build + ~£1,500 running |
| Year 2 | £12,000, plus any rise | £1,500–£3,000 maintenance |
| Year 3 | £12,000 | £1,500–£3,000 maintenance |
| Three-year total | ~£36,000+ | ~£16,500–£21,000 |
| If volume doubles | Usually costs more | Usually no change |
| If requirements change | They adapt immediately | Change work, priced |
Where the retainer genuinely wins
- The work needs judgement that varies case by case
- Requirements change monthly, so any build would be obsolete on delivery
- Volume is low or seasonal and you want to pay only when it happens
- Nobody internally will own a system
The last one decides more of these than anything financial. A build with no owner degrades quietly; a retainer has a supplier who notices.
Where building wins
High frequency, low judgement, stable for a year, and errors that cost real money. At volume, a build does the work at 3am on a bank holiday and applies exactly the same rule every time.
It also scales without a conversation, which is the property that matters most when you are growing.
The hybrid that usually wins
Automate the mechanical majority and keep a smaller retainer for the judgement work and the exceptions. Most of our clients end up here rather than at either extreme.
It also gives the retainer something more valuable to do than keying, which tends to improve the relationship as well as the cost.