The per-seat tax: what three years of SaaS really costs, and when bespoke pays back
Nobody budgets for SaaS over three years. It arrives as a monthly line item, a few tens of pounds per person, and it's small enough that nobody questions it.
But per-seat pricing has a quiet property: it grows with your team, and it grows again when the vendor raises prices. Add the two together and the number you actually commit to is a lot bigger than the one on the pricing page.
The sum nobody does
Take a team of 25 on a tool at £35 per seat per month. That's £875 a month, or £10,500 a year. Fine.
Now let the team grow by 10% a year and let the price rise by 7% a year, both of which are ordinary assumptions. Year two costs £12,600. Year three costs £14,400. Over three years that's £37,500, and we haven't counted a single add-on, onboarding fee or premium tier.
Compare that with a bespoke tool, built once and owned outright. With a £10,000 build, £150 a month for hosting and support, and £2,000 a year for enhancements from year two, the same three years cost around £19,400. There are no per-user fees, so adding the 26th person costs nothing.
It depends on how many people use it
This is the part that matters, and it's why we don't tell everyone to build custom software. Bespoke has a fixed cost. SaaS has a cost that scales with headcount. So the question is where the two lines cross.
Using the same assumptions as above (£35 a seat, 10% team growth, 7% annual price rise, and a £10,000 build), here's how three years looks at different team sizes:
- 5 users: SaaS costs about £7,700 and bespoke about £19,400. SaaS wins, and comfortably. If this is you, stay where you are.
- 13 users: SaaS costs about £19,400 and so does bespoke. This is the break-even point, reached right at the end of year three.
- 25 users: SaaS costs about £37,500 against £19,400. Bespoke has paid for itself by month 14 and saves around £18,000 over three years.
- 100 users: SaaS costs about £150,000. Bespoke pays for itself in around three months, and the gap is more than £130,000.
So on these numbers, the crossover sits at around a dozen people. Below that, off-the-shelf is the sensible choice. Above it, the per-seat tax starts to outweigh the cost of building something that fits.
Your numbers will differ. A £140-a-seat CRM crosses over with a much smaller team than a £10-a-seat task manager, and a complex build costs more than £10,000. That's the reason to use your own figures rather than ours.
The costs that aren't on the invoice
The comparison above only counts what you pay the vendor. It ignores what the tool costs you in time, and that's often the bigger number.
- Someone re-keying data because two tools don't talk to each other
- The spreadsheet that fills the gap the product doesn't cover
- Retraining after a forced migration or a redesigned interface
Suppose a tool costs your team five hours a week in workarounds, at a fully loaded cost of £30 an hour. That's £7,800 a year. Add one disruptive change a year, costing each person three hours to adapt to, and the 25-person team's three-year SaaS cost goes from £37,500 to about £68,000.
Those figures are estimates, and you should use conservative ones. But the cost is real even if nobody tracks it, and it doesn't shrink as you add seats. It's also why the break-even point moves: with the same hidden costs, even a five-person team's SaaS bill reaches about £32,600, and bespoke breaks even in around 15 months.
Run your own numbers
We built a calculator so you can check this against your own situation, without taking our word for it. Pick your tool (or enter your own price), set your team size, growth and price rises, and see the three-year totals and the month where bespoke breaks even. There's an optional switch for the hidden costs above, and a plain explanation of every assumption underneath.
Run your own three-year comparison →
If the answer is "stay on SaaS", that's a perfectly good answer and the calculator will tell you so. If it points the other way, get in touch and we'll give you a fixed price for the build.