AI projects are producing measurable gains at many companies, but only 13% of organizations with implementations have scaled them fully in line with the original business case, a BearingPoint survey finds.

The consultancy questioned 1,050 C-suite executives and senior leaders in 13 countries across Europe, the United States and China. Among respondents with AI in use, 74% reported a measurable effect on revenue or costs, yet almost three-quarters had reduced the intended scope or achieved less scale than planned.

For executives deciding whether to fund another pilot, the result points to a more useful question than whether AI can work: can the organization connect a working tool to reliable data, accountable budgets, existing systems and workforce decisions? The survey associates those management foundations with the small group that expands AI successfully.

Four stages of adoption

BearingPoint divides respondents into four maturity groups. Explorers, 15% of the sample, are still assessing opportunities. Experimenters, 20%, run projects that have not entered operations. Implementers, the largest group at 54%, have deployed tools and generated some value. Leaders, 11%, report broad integration, measurement and a defined transformation plan.

The gap between the last two groups is substantial. The study says 47% of Leaders scale initiatives completely as planned, compared with 6% of Implementers. That correlation does not prove that BearingPoint’s maturity practices caused the difference, but it helps identify which operational capabilities accompany successful expansion.

Reported financial effects were also uneven. Nearly half of organizations put AI’s impact below 4% of costs and below 2% of revenue. About four in ten reported both revenue growth and cost reduction, while more than a quarter saw productivity improve without a measurable profit-and-loss effect.

Agents amplify the readiness gap

Interest in autonomous systems is ahead of deployment. Only 13% of organizations reported a defined agentic-AI strategy with active initiatives, and 10% said they were scaling agents across the enterprise. More than three-quarters were still learning, setting priorities or exploring pilots.

That lag matters because agents connect more systems and can take actions rather than only produce text. Legacy integration and complex regulation were the two most frequently cited barriers to AI scaling. An agent added to weak data access or unclear approval rules increases operational exposure instead of fixing the underlying process.

Productivity creates a workforce problem

The survey says 62% of organizations already report AI-related workforce overcapacity of at least 10%, with 95% expecting that level by 2030. These are executives’ assessments, not audited job counts or a forecast validated against hiring data, so they should not be read as a direct prediction of layoffs.

Even so, the result identifies a planning gap. A productivity gain affects the profit and loss statement only when a company changes capacity, prices, output or service quality. Keeping the same work design after automating part of it can leave savings theoretical and employees uncertain about how roles will change.

How to read the findings

The research was produced by a consultancy that advises organizations on technology and management. Its sample spans regions and senior roles, but the public summary does not provide response-level data, detailed weighting or independently audited performance measures. The reported business impact is therefore best treated as a structured view of executive experience, not a universal benchmark.

The durable message is narrower: adoption and value are different milestones. Companies that want to move beyond pilots need named financial owners, integration plans, governance and a workforce decision before declaring a project ready to scale. The next useful evidence would track the same organizations over time and compare reported gains with audited outcomes.

Verification

  • VERIFIED: BearingPoint says the study surveyed 1,050 C-suite executives and senior leaders in 13 countries and that 74% of implementers report measurable revenue or cost impact. BearingPoint press release
  • VERIFIED: The study summary reports a 13% full-scale rate, the four maturity groups and a 47%-versus-6% scaling gap between Leaders and Implementers. BearingPoint study
  • VENDOR-REPORTED: Workforce overcapacity, financial impact and agentic-adoption figures are self-reported survey results published by the consultancy; response-level data were not available for independent review.
  • VERIFIED: Reuters independently summarized the survey’s main scaling and return findings. Reuters
  • ANALYSIS: Recommendations about ownership, integration, governance and workforce design interpret the survey rather than report controlled causal findings.