The Widening AI Value Gap: Build for the Future 2025
S · Shapes the field BCG's annual Build for the Future study of 1,250 executives in 68 countries; its 5%-versus-60% value split became one of the most quoted findings of 2025, though maturity and value are self-reported.
Summary based on the full text·Original: about 35 minutes to read
In short
BCG's September 2025 study scores 1,250 companies on 41 AI capabilities and sorts them into four maturity bands. Only 5% are "future-built" and get substantial value from AI; 60% get little or none. The leaders spend more on IT, put more of it into AI, reshape core workflows end to end rather than run isolated pilots, and upskill more than half their staff. BCG argues the gap is compounding because leaders reinvest their gains.
Key numbers
- 5%companies BCG classes as 'future-built', achieving AI value at scale
- 60%companies reporting minimal revenue and cost gains from AI despite substantial investment
- 1.7xrevenue growth of future-built firms versus stagnating and emerging firms
- 3.6xthree-year total shareholder return of future-built firms versus laggards (Capital IQ)
- 17% to 29%agents' share of total AI value in 2025, expected by 2028
As reported by the source.
English·PDF·opens media-publications.bcg.com
What it reports
- 2025 split: stagnating 14%, emerging 46%, scaling 35% (13 points more than 2024), future-built 5%.
- Future-built vs laggards: 1.7x revenue growth, 1.6x EBIT margin, 3.6x three-year TSR, 2.7x return on invested capital.
- Future-built firms plan 26% more IT spend and up to 64% more of their IT budget on AI in 2025.
- 70% of AI value potential sits in core functions such as sales and marketing, manufacturing, supply chain and pricing; R&D alone 15%; IT's share rose 6 points to 13%.
- Agents account for 17% of AI value in 2025, expected 29% by 2028.
- Sector gaps: in less mature sectors under 50% of staff can use GenAI tools; in mature sectors over 70%. Software firms plan to upskill 55% of staff, chemicals and machinery under 15%.
- Agent experimentation or deployment: North America 51%, Asia-Pacific 45%, Europe 41%.
How they got it
Survey of 1,250 CxOs and senior AI decision makers across 68 countries and nine industries, who rated their company on 41 capabilities in four stages. Weighted scores (random forest, logistic regression and PLS structural equation modelling) placed firms in bands of 0-25, 25-50, 50-75 and 75-100. TSR and EBIT came from Capital IQ and Orbis.
Limits
- Maturity and value are self-reported, often for the business area the respondent knows best, and many figures are expected rather than realised; BCG flags perception bias.
- Financial outperformance is correlation; BCG notes company performance has many drivers beyond AI, and successful firms may simply invest more in everything.
- The playbook maps onto BCG's consulting offer, including its "10-20-70" rule.
Why it matters for companies
The study gives leadership teams a benchmark for where value concentrates: a few core workflows redesigned end to end, not dozens of pilots spread thin. The training and access gaps by sector are a quick test of whether a company is in the serious group. Treat the TSR multiple as a description of who is winning, not proof that AI caused the win.
Summary based on the full text. Read the original for the full argument and data.Open PDF on media-publications.bcg.com
- Tags
- AI agents