The £20k Licence That Costs £270k a Year
Licence fees are the sticker price, not the cost. The true annual cost of an operational SaaS product is licence plus change requests plus workaround labour plus downtime plus the innovation you cannot ship. For mid-market operators the total routinely runs 5 to 10 times the licence line, and sometimes more. Totalled honestly, a £150k to £350k custom build with a 2 to 4 year payback stops looking expensive. Run your own numbers with our payback calculator.
Ask a finance team what an operational SaaS product costs and they will quote the licence line. It is the only number that arrives as an invoice with the vendor’s name on it.
The other costs are just as real. They are simply paid in different currencies, from different budgets, and they never share a page. This post puts them on one page.
The worked example: a composite operator
The table below is a fictional composite, built from patterns we see repeatedly across multi-site and franchise businesses. The licence is deliberately modest, because that is how these stories usually start.
| Cost line | Annual cost | Where it hides |
|---|---|---|
| Licence and subscription | £20,000 | The IT budget (the visible line) |
| Change requests and customisation | £130,000 | IT or operations budget, invoice by invoice |
| Workaround labour (the spreadsheet layer) | £48,000 | Payroll, across several departments |
| Downtime at peak periods | £45,000 | Lost revenue, never a cost line |
| Error correction and reconciliation | £27,000 | Refunds, credits, and staff time |
| True annual cost | £270,000 | 13.5x the licence fee |
Every line deserves a word.
Change requests dominate this example, and that is not unusual. One multi-site leisure business we spoke to paid around £20k a year in licence fees and over £200k a year in change requests, for work that was slow, expensive, and inconsistent in quality. We have written separately about where that change request money actually goes.
Workaround labour here assumes 30 hours a week across the business at a loaded cost of £35 an hour, over 46 working weeks. That is a mid-sized spreadsheet layer; plenty are bigger. How to estimate yours is a post of its own.
Downtime gets its own line because peak-period outages hit revenue directly. The same leisure business suffered regular unavailability clustered at exactly the moments of heaviest customer demand, with a vendor unable to diagnose or fix the underlying scaling problems. When your busiest hour is your least reliable hour, the cost is not hypothetical.
Errors and reconciliation are the leakage from every manual bridge between systems: mispriced transactions, missed renewals, and the hours spent finding and fixing them.
One honest omission: the table excludes the opportunity cost of innovation you cannot ship because the roadmap belongs to someone else. It is real, but it is hard to price defensibly, so the £270k is a floor, not a ceiling.
What does the payback maths look like?
Set the true cost against a replacement build in the current market:
- Build cost: £150k to £350k covers small to large mid-market operational systems. Take £250k as the middle scope.
- Run cost of the custom system: assume 30% of current true cost, so around £81k a year for hosting, support, and continued development. That is a deliberately cautious figure, and unlike change requests, the development spend builds an asset you own.
- Annual saving: £270k minus £81k, roughly £189k a year.
Payback on the £250k build is around 16 months. At the £350k scope it is under two years. Stress-test it: assume only half the workaround, downtime, and error savings materialise, and payback stretches to roughly two to three years. Over five years, the base case compares £1.35m of continued SaaS true cost against about £655k for the build and its running costs.
The real composites behind this post landed across that range. The membership business spending around £160k a year reached payback in about two years. The franchise network paying roughly £10k a month in licence fees paid back in around three. The leisure business, with the heaviest change spend, took around four years even at that scale, and still considered it the right call because it ended the outages and bought back the roadmap.
Why has the build side of the equation moved?
Because AI-augmented delivery has cut the cost of the build itself. Our recent proposals for large greenfield Azure builds are landing at 28 to 34% of what the same work would have cost before AI, at the same quality or better (Q2 AI Velocity Report). A system that would have cost £700k to build in 2022 now sits inside the £150k to £350k range this post assumes, with full code ownership as standard.
That is the whole reason this arithmetic is worth redoing now. The SaaS side of the ledger has not changed. The build side has.
Run your own table
One worked example proves nothing about your business. Two resources will get you to your own numbers:
- Our SaaS replacement payback calculator builds this table for your figures and shows the payback period and five-year comparison, with every assumption visible and editable.
- The ROI guide for AI-augmented development explains the full cost model, including the inputs vendors tend to leave out.
When the numbers point to replacement, the AI-era SaaS replacement playbook covers when, why, and how to make the move.
Prefer to have the table built for you? Book a consultation and we will put your real numbers into this model, line by line.