McKinsey's Workforce Transformation: The Dual Impact of AI on Employment and Productivity
As AI continues to revolutionize the consulting industry, McKinsey's CEO reveals significant shifts in workforce dynamics, highlighting a 25% increase in client-facing roles and a corresponding decrease in non-client-facing positions. This article explores the implications of these changes for employees and the consulting landscape.
TL;DR
McKinsey boss Bob Sternfels says AI has changed the firm’s staffing math: client-facing roles are up 25% while back-office roles are down about 25%, yet output still rises. He says AI saved 1.5M hours of search/synthesis work, pushing consultants “up the stack.” McKinsey now has ~40k people and ~25k AI agents—and could hit parity by year-end.
McKinsey’s global managing partner Bob Sternfels says AI is changing the firm’s staffing math: the consultancy can expand client-facing teams while shrinking support functions—and still grow overall.
He shared the numbers during a CES appearance tied to a live “All-In” podcast taping with Jason Calacanis and General Catalyst CEO Hemant Taneja.
CES spotlight on McKinsey
Speaking from CES in Las Vegas, Sternfels described AI as a force that has “fundamentally changed” how McKinsey allocates people across the business.
Instead of assuming growth must always mean adding headcount everywhere, he suggested AI now lets the firm rebalance where it hires, where it reduces roles, and how much output each side can produce.
That shift matters beyond consulting, because it mirrors what many large enterprises are wrestling with: how to modernize an established operating model without breaking what already works.
The “25 squared” staffing model
Sternfels labeled McKinsey’s new approach “25 squared,” meaning the firm is increasing client-facing roles by 25% while reducing non-client-facing roles by about 25%.
He said the non-client side represents roughly half the workforce, and despite the reduction, output from that side has still risen by about 10%—a result he connects to AI-driven productivity.
In his words, McKinsey can now “grow” the client-facing half and “shrink” the support half while still achieving aggregate growth, calling it a new paradigm for how professional services firms scale.
Productivity gains: “Up the stack”
Sternfels said McKinsey saw large efficiency gains after adopting AI, including 1.5 million hours saved in search and synthesis work over the previous year.
He also argued that when AI handles the heavy “research-and-summarize” lifting, consultants spend less time on junior-level grunt work and more time tackling complex, higher-value problems—what he called moving “up the stack.”
The implication is that early-career roles don’t vanish overnight, but the work inside those roles changes quickly, pushing teams to learn faster and operate with leaner structures.
40,000 people—and 25,000 agents
Sternfels said McKinsey had about 40,000 human employees and 25,000 personalized AI agents as of “last week,” describing agents as digital coworkers that can handle entire job functions.
He expects the firm could reach rough parity—similar numbers of agents and people—by the end of the year, signaling how fast “agentic” work models are spreading inside large organisations.
He tied this to a broader warning for incumbents: leaders have to transform existing enterprises into something different, because the choice is increasingly “transform or die,” and CEOs are fixated on organisational speed over pure strategy. Business Insider has also reported that AI is pushing consulting firms to rethink how they monetize work, including experimenting with outcome-based pricing rather than billing primarily for time and teams.
