Kerry Brown, Transformation Evangelist at Celonis, is a strategist and thought leader who helps companies achieve organizational excellence.
A concerning pattern is emerging in corporate workforce planning. Organizations often reduce headcount based on assumptions about what AI can handle, only to discover that the work—or the expertise needed to perform it—cannot be replaced as easily as expected. Gartner and Forrester have projected that roughly half of the job cuts made in the name of AI will be reversed.
AI absolutely provides the opportunity to reduce repetitive work, improve performance and enable employees to focus on higher-value activities. As companies increasingly explore organizational transformation or process reimagination, due to market pressures or M&A transactions, realizing transformation’s full potential requires a clear understanding of how work actually gets done.
The challenge is rarely technology alone. Transformation initiatives are often missing business context: the exceptions, subtleties, nuances, informal handoffs and judgment calls that shape how work happens in practice. These rarely appear in process maps, job descriptions or organizational charts. Without true visibility, leaders risk making workforce decisions based on an oversimplified view.
The simplest conclusion is that AI has been oversold, but recent reversals in hiring point to a different problem: decisions being made without a full view of the work, expertise and institutional knowledge. In my 20 years working in digital transformation at major enterprises, I have repeatedly seen that work described on paper differs from what people actually do.
AI is bringing that gap into sharper focus. According to a recent report from my company, 45% of business leaders cite difficulty getting AI to understand their business context as a barrier to adoption. Even advanced AI agents only successfully complete about one in four real-world tasks, according to research from UC Berkeley. Essentially, work is complicated and AI doesn’t fully get it.
From Broad Assumptions To Evidence-Led Capacity Planning
Workforce optimization should begin with a clear view of how work gets done, not assumptions about where AI can replace people. This requires a data-driven approach—one that uses operational evidence to determine what can be automated, where human expertise remains essential and how existing talent can be reskilled or redeployed.
Digital records that business systems already produce—including time stamps and transactions—can be assembled into a living digital twin of an organization via process intelligence. Organization-related elements (org charts, roles, etc.) can be incorporated to reveal how work happens day to day, including the variations and exceptions that traditional process maps miss. This specificity provides the detail needed to identify and quantify opportunities to optimize the organization.
With this context, leaders can replace broad headcount assumptions with a practical three-step framework with workforce performance as a central goal:
1. Map work before you move headcount. Before making workforce decisions, identify how work is performed, where exceptions occur and which employees provide judgment or coordination that may not appear in job descriptions.
2. Recognize work that requires judgment. Be specific about which tasks AI can reliably handle and where people must remain involved (in the loop). Measure success through better outcomes across people, processes and technology—not simply volume of AI deployment.
3. Reskill and redeploy where you can. Reskilling should be an objective of AI transformation from the outset, not an option considered only after automation decisions have been made. Redeploying can be a win for the company and the employee.
A Steadier Way To Move Fast
Being evidence-led does not mean being slow. It means making the right fact-based decisions. AI will improve, but in the meantime, understanding how work gets done means you won’t pay twice: once to eliminate capacity and once to rebuild it.
Before making AI driven workforce changes, leaders must understand where technology creates value, where people remain essential and where employees can move into higher-value work. This context allows organizations to protect and enhance the expertise their business, and their future, depend on.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?