Current UK HVAC-specific AI supplier revenue, paid adoption and annual purchasing volume remain unknown in the evidence reviewed. The desk should not publish a market CAGR or a single total-addressable-market figure as if these had been measured.
A transparent sensitivity model is more useful than an unsupported headline.
What could an illustrative annual spending pool look like?
The following are desk-created planning scenarios dated 27 October 2026—not estimates of current revenue, adoption forecasts or confidence intervals. Every input is an assumption. The expenditure levels are broadly compatible with the published software and managed-service examples later in this report, but willingness to pay has not been established.
| Assumed input | Smaller scenario | Middle scenario | Larger scenario |
|---|---|---|---|
| Operationally suitable firms | 1,000 | 2,500 | 4,000 |
| Share buying incremental external AI services in a year | 10% | 20% | 30% |
| Paying firms, calculated | 100 | 500 | 1,200 |
| Average annual incremental external expenditure | £3,000 | £6,000 | £12,000 |
| Illustrative annual spending pool | £0.30m | £3.00m | £14.40m |
Formula: suitable firms × paying adoption × annual expenditure.
These scenarios exclude existing core software expenditure unless it increases specifically because of the AI purchase. They also exclude internal staff costs from supplier revenue and avoid counting an agency’s platform pass-through charges twice.
The principal uncertainty is not arithmetic. It is whether contractors will pay an external supplier rather than configure native software, improve callbacks, purchase human answering or make no change.
Research conclusion: the earlier desk notes’ full-adoption spending envelope and AI-spending growth assumptions should remain internal hypotheses until supported by buyer interviews, invoices or a representative survey.