Boston Consulting Group’s September 2025 report, The Widening AI Value Gap, surveyed more than 1,250 companies worldwide. The headline is blunt: only about 5% achieve AI value at scale, while most investment still fails to produce material returns.
The split BCG reports
In BCG’s Build for the Future 2025 global study (n = 1,250), only 5% of companies are “future-built”—achieving substantial AI value at scale through revenue, cash-flow, or workflow gains large enough to matter. Fully 60% report little or no material value. Another 35% are scaling efforts and seeing some returns, but many say they are not moving far or fast enough.
Within the broader maturity mix BCG shows for 2025, a large emerging/experimenting share (about 46%) sits below the scaler and future-built tiers—active enough to spend, not mature enough to compound value.
Why the gap widens
BCG argues the top 5% reinvest early wins into capabilities—strategy, tech, people, and innovation—while laggards stay in pilots and shallow automation. Future-built firms also lean harder into agentic use cases as a share of AI value.
The practical reading: scattered tools without operating-model change do not move the P&L.
What “value at scale” implies
Scale here is not a demo count. It is measurable impact across workflows and financial outcomes. If your dashboard only shows seat counts and prompt volume, you are not measuring what BCG’s top cohort is measuring.
Start with one owned workflow, a baseline, and a review path—then expand only when the metric moves.
How to use the research
Share the 5% / 60% / 35% split with leadership to reset expectations. Treat the 46% stuck-experimenting band as a warning against endless pilots without production owners.
Read the primary PDF before citing secondary summaries; figures above follow BCG’s September 2025 publication.