Digital transformation ROI: evaluation and measurement
- McKinsey classified 6% of 1,993 survey respondents as AI high performers under a defined EBIT and value threshold.
- MIT's 5% production finding is preliminary and directional, based on a mixed-method research programme rather than a representative market survey.
- Revenue, operating-model and project-abandonment findings are shown with their own populations and methods. They are not combined into one ROI benchmark.
This evidence review focuses on enterprise AI transformation ROI. It separates observed financial impact, implementation signals and project attrition because the studies measure different populations and outcomes. None of the figures represents an expected return for an individual programme.
Enterprise AI transformation financial impact
McKinsey's 2025 global survey provides the clearest definition of the high-performing group. MIT's preliminary report studies a different question and should be treated as directional evidence.
| Evidence | Finding | Population and method |
| McKinsey State of AI, 2025 | 6% qualified as AI high performers. These respondents attributed more than 5% of organisational EBIT and significant value to AI. | Global survey of 1,993 participants at all organisational levels, conducted 25 June to 29 July 2025. The high-performer group contained 109 respondents. |
| MIT NANDA preliminary report, 2025 | About 5% of task-specific GenAI tools examined reached production with measurable profit-and-loss impact. | Preliminary mixed-method research drawing on more than 300 public initiatives, 52 structured interviews and surveys of 153 senior leaders. It is not a representative market survey. |
The two percentages are not comparable. McKinsey classifies organisations by reported EBIT contribution and value. MIT describes task-specific deployments found through a smaller mixed-method study.
Operating patterns associated with AI value
| Evidence | Finding | Population and limitation |
| McKinsey State of AI, 2025 | AI high performers were nearly three times as likely as other regular AI users to report fundamentally redesigned workflows. | Association within the same 1,993-participant global survey. It does not establish that workflow redesign alone caused the financial result. |
| BCG AI Radar, 2026 | Trailblazer CEOs were about twice as likely as followers to deploy AI agents end to end across a workstream or process. | Global survey of 2,360 executives, including 640 CEOs, across 16 markets. The groupings reflect reported confidence, investment and behaviour. |
Both studies associate broader value with changes to workflows and operating practice. They do not provide a transferable return calculation.
Reported enterprise AI revenue impact
McKinsey's Superagency report asked 118 US C-suite respondents how GenAI had affected revenue. These are executive perceptions reported in October and November 2024.
| Reported revenue change | Share of respondents | Evidence boundary |
| Increase above 5% | 19% | 7% reported more than 10%; 12% reported 6% to 10%. |
| Increase of 1% to 5% | 39% | Self-reported revenue impact, not an audited result. |
| No change | 36% | Survey finding for this US C-suite sample only. |
AI project abandonment before scale
S&P Global's Voice of the Enterprise survey examined delivery attrition rather than ROI. Its 2025 sample covered 1,006 midlevel and senior IT and line-of-business professionals in North America and Europe.
| Measure | Finding | Scope |
| Companies abandoning most AI initiatives before production | 42% in 2025, up from 17% in the previous survey. | Respondent-reported organisational experience. |
| Projects scrapped between proof of concept and broad adoption | 46% on average. | Average reported by respondents, not a measured market-wide failure rate. |
These findings make production readiness, accountable ownership and a defined value baseline part of the investment decision. They do not predict the outcome of a specific AI initiative.
Sources
- McKinsey & Company (2025). The State of AI in 2025: Agents, Innovation, and Transformation.
- MIT NANDA (2025). The GenAI Divide: State of AI in Business 2025, preliminary report.
- BCG (2026). As AI Investments Surge, CEOs Take the Lead.
- McKinsey & Company (2025). Superagency in the Workplace.
- S&P Global Market Intelligence (2025). AI Experiences Rapid Adoption, but with Mixed Outcomes.
