Briefing · 27 pages
Enterprise AI in 2026, read across ten studies
Research briefing · 2026
Every major enterprise AI study of 2026 has been summarised somewhere. Almost none of them have been read against each other, which is the only place the interesting material is.
Ten sources from nine organisations, grouped into nine categories of measurement — adoption, individual speed, delivery, value, work design, foundations, governance, perception and geography. Each category gets the same treatment: the figures with their sources and sample sizes, what they mean together, and where twenty-five years of building enterprise software confirms the finding or runs against it.
The single most useful thing in it. Three independent measurements, from three firms asking three different questions, land on the same fifth of the market getting measurable value from AI: 19%, 20% and 21%. Nobody designed those questions to agree. Everything else is an attempt to explain why the other four fifths are not there.
Two conventions apply throughout. Every figure carries its source and sample size in the same sentence, and where a source has a commercial interest in its own conclusion, that is stated where the number appears rather than in a footnote at the back.
What's inside
- Nine categories of measurement, ten studies, roughly 45,000 respondents
- Six derived ratios that appear in no single report, each with the arithmetic that produced it
- Where the studies contradict each other — and which contradictions have an explanation
- Nine first-hand accounts from twenty-five years of enterprise delivery
- A bibliography naming each source, its method, its sample and its commercial interest