From minutes saved to value realised
A demonstration that produces a report in five minutes is usually honest. It is also usually answering a different question from the one you are being asked to buy.
What does "time saved" actually mean?
At least six different things, and they are not interchangeable. A percentage that does not say which one it means cannot be checked by anybody.
| The number says | What it counts | What it does not tell you |
|---|---|---|
| Touch time | Minutes a person is actively working | Whether the report left any sooner |
| Elapsed time | Instruction to approved issue | Whether anyone worked fewer hours |
| Throughput | Acceptable work finished per hour | Whether hours or cost fell |
| Quality divided by time | Graded output against time used | How much faster anything got |
| Released capacity | Hours freed for something else | What the hours were then used for |
What does the controlled evidence really show?
Bounded professional tasks commonly get 10 to 40 per cent faster. In Science, 453 professionals wrote 40 per cent faster at 18 per cent higher graded quality. In Organization Science, 758 consultants worked about 25 per cent faster inside the model's capability, and were 19 percentage points less likely to be right on the one task outside it. Elsewhere, 5,172 support agents resolved 15 per cent more issues an hour and 4,867 developers completed 26 per cent more tasks.
Then the other half of the same literature. Sixteen experienced developers took 19 per cent longer on code they knew well, while believing they had been 20 per cent faster. In one randomised trial of two ambient clinical scribes, one cut documentation time by 9.5 per cent and the other did nothing measurable. Same technology, same year, opposite results, which is roughly how construction AI actually fails as well.
Very large numbers are not automatically marketing. A 2026 legal trial reported gains of 34 to 140 per cent because productivity there means graded quality divided by time used. The time reductions underneath were 20 to 28 per cent. That is a legitimate measure and a different one.
Why does a task saving shrink across a whole job?
Because a survey, inspection or progress report is not prose. It is capture, evidence selection, interpretation, drafting, checking, communication and accountable judgement, and a tool reaches only some of that. Net hours saved is baseline times eligible share times measured saving times adoption, less verification, exceptions and the cost of running the system.
Every one of those is a number you can measure on your own reports in a fortnight. None of them is a number a supplier can know about your firm in advance.
I found no end-to-end randomised trial of construction or surveying report production. Until there is one, 5 to 10 per cent across a whole role is a planning hypothesis to test, not a benchmark to quote.
Where should the automation stop?
Photograph handling, standing data, dictation into controlled fields and completeness checks can be compared against a source, so a machine can do them. Evidence-linked drafting sits in the middle, where model output is a draft and never a source. Cause, significance, contractual position and safety stay with a person, which is the human in the loop matrix doing its job.
The RICS standard on responsible AI use has been mandatory since 9 March 2026 wherever AI materially affects a service. It requires a recorded reliability decision by a named surveyor, an AI register and sampling of high-volume automation. That is a running cost. It is also the control system that makes the saving defensible when somebody challenges the report.
For scale, 13 per cent of UK construction businesses with ten or more employees used any AI technology in June 2026, against 35 per cent across all sectors.
What should you ask before you sign anything?
Demand the denominator. Which outcome, over what boundary, at what quality bar, on whose reports, with adoption and correction time counted in.
"Across 42 matched condition reports in a 90-day pilot, median touch time from evidence pack to approved issue fell 18 per cent after corrections, with no measured quality loss, on 64 per cent of eligible reports" is a claim you can audit.
"Eighty per cent faster" is not a claim at all.