McKinsey’s head of L&D on how AI is changing how its consultants learn and how they serve their clients

Hey ChatGPT, tell me why my company should pay an expensive consulting firm to advise us when I can just ask you for free?

That’s the “existential” question haunting the consulting world, and even highly successful giants like McKinsey are grappling with it.

More AI tools are coming out that can do the work of consultants, arguably for cheaper. And consulting firms are engaging in a delicate balancing act of selling their expertise on AI to clients, while offering advice on AI to those clients and simultaneously learning this technology themselves.

McKinsey, which currently has some 40,000 employees, is rapidly deploying AI while undergoing a significant shift in its workforce. Since the end of 2023, the firm has reduced its staff size by 5,000 employees. At the same time, it’s rolled out around 12,000 AI agents. Using these agents, teams assigned to one client project have shrunk, Kate Smaje, McKinsey’s global leader of technology and AI, previously told the Wall Street Journal.

The consulting giant has implemented several changes to adapt to AI’s disruption of the consulting industry. For example, the firm has shifted its offerings from providing strategic advice to clients to also guiding them through implementation. Around one-quarter of the firm’s work is now based on “outcomes-based arrangements,” the Journal reported in August. Additionally, McKinsey is retooling its learning and development approach, including how it teaches its staffers about AI, and using the technology to reshape learning at the firm.

“The value proposition to our clients and the value proposition to our people is that we’re going to accelerate everybody’s development,” Heather Stefanski, McKinsey’s chief learning and development officer, told HR Brew.

Learning with AI. McKinsey is well known as an academy company: its unique approach to developing its employees has allowed many of its former staffers to eventually lead top companies. Stefanski said that McKinsey being a leadership incubator is thanks to its long-time approach as a gig economy, where staffers, with the guidance of their leaders, develop a well-rounded experience by taking on varying projects that help develop lesser-used skills (such as building analytic modeling skills, or improving how they develop relationships with clients). As part of annual evaluations, employees fill out a development growth plan where they must share the skills they learned or tried to learn in the past year, and which ones they want to develop in the year ahead.

Now, instead of learning opportunities happening in human-to-human interactions, AI will be used to augment the firm’s “intense apprenticeships,” she said.

As an example, the firm uses AI agents for a myriad of tasks including instruction on how to create powerpoint presentations, take notes and transcribe, and even stick to the firm’s preferred tone of voice in their work writing, the Wall Street Journal reported. These tools can help staffers learn these skills in the moment, and allow the human-to-human learnings to focus on things like improving on empathetic leadership, Stefanski said.

Understanding AI. McKinsey, like any consulting firm, is reckoning with the demand from clients for AI expertise, while its own staff learns the novel technology at the same time. And there’s a unique pressure to stay on top of developing that expertise for the firm: AI advising reportedly makes up roughly 40% of the firm’s revenue.

McKinsey has a credentialing program where employees are certified on AI skills, ranging from levels one through five. Stefanski said the program rewards employees for learning about and using AI tools, and gives the firm an understanding of the skills employees do have.

Largely, though, the firm has focused on encouraging employees to use and experiment with AI by creating a work environment that naturally incentivizes that, she said.

For example, McKinsey has embedded “AI black belts” onto teams—employees who have a deep understanding of the AI tools available to them and know how to build tools like AI agents, and who can share that expertise to advise clients. These employees, who stay on top of developing AI skills, can act as change agents for their teams, encouraging experimentation and learning on a smaller scale.

It also focused its internal awards program on AI use cases, where the best uses are rewarded and showcased to the rest of the firm’s staff, which Stefanski said can give employees insight into how their peers are using the technology.

“It’s less about the incentives that we need, and it’s more about figuring out how to create an environment for people to experiment, learn, grow,” Stefanski said.

The firm is constantly adding information about AI to its internal knowledge management system, where employees can access articles and contact colleagues who are experts in subjects they want to learn more about. “The ability to have the reach that we have on these things, as soon as that happens, [that information] is in the organizational water,” Stefanski said.

Now, employees can use Lilli, the firm’s internal AI tool, to retrieve this information even faster. While Stefanski may have previously spent time calling and searching for articles or experts, a quick conversation with Lilli can net a wealth of information in seconds. “I’ve got the history of the McKinsey knowledge corpus at my fingertips,” Stefanski said.

Overall, though, human expertise remains far more valuable for consultants and clients than what they can only get from AI, Stefanski said.

“To me, it’s always about using AI to give you a starting point. And then, what we do at McKinsey is we take that starting point and we make it better, because you bring in the expertise, because you bring in the context, because you bring in the history of what we’ve done,” she said.