Recent articles by a Wall Street Journal reporter and by a highly regarded veteran journalist suggest that generative AI could render management consulting firms obsolete. The headline in the WSJ begins, "AI Is Coming for the Consultants," and refers to an "existential" threat to McKinsey. The headline's dek reads: "If AI can analyze information, crunch data and deliver a slick PowerPoint deck within seconds, how does the biggest name in consulting stay relevant?"
But let's not get carried away just yet.
From my perspective both as a long-time follower of technology and of insurance and as a former partner in a 2,000-person management consulting firm, gen AI will require adapting the business model for consultants and, perhaps a bit later, insurers. But gen AI will actually accentuate the value of the sharpest consultants and the best underwriters, claims professionals, and agents and brokers — not put them out of business.
As a way of sorting through the hype and the reality surrounding generative AI, I thought it would be worth taking a look at what's really likely to happen.
Here are the two articles: the WSJ piece and one written for the Free Press by veteran business reporter Joe Nocera. The text of the WSJ article isn't as alarmist as the headline but still makes the case that gen AI can produce presentations as slick as a consultant can, so who needs human consultants? Nocera, scarred from his experience as an employee at Time Warner during the disastrous, consultant-heavy takeover by AOL, makes much the same argument, and rather gleefully. The headline of his piece reads: "The Consulting Crash Is Coming: Bloated, overpaid, and outpaced by AI—big firms confront a future they can’t outsource."
I see two major problems with the thesis, neither of which undercuts the power of gen AI but which suggest tempering expectations at least about how quickly gen AI will take over the world. Both issues also apply in insurance.
First, the articles get caught up in the glitz of AI and ignore a question that gets ignored all too often: What if the AI is wrong?
Yes, an AI can do a remarkable amount of analysis quickly and can turn it into a slick PowerPoint presentation with breathtaking speed, leaving the usual army of 20-something MBAs way behind. But what if the analysis is wrong?
I see this sort of assumption of AI-as-oracle all the time. But in a lot of cases, it isn't even possible to know whether the AI will be right. It can't know for sure how competitors will react to a new strategy by a client. It can't know for sure whether predictions about morality for life insurers will be accurate, not for years or even decades.
Sure, those armies of MBAs make mistakes, too—plenty of mistakes, including some big ones—and the AI will keep learning and improving over time. But the AI will have to earn its stripes just as consulting firms have over time, so we can't just assume the speed of AI will win the day.
Second, the articles gloss over the value of the seasoned, senior consultants and miss how those senior consultants—like their counterparts among underwriters, claims professionals, and agents and brokers—became so valuable.
Much of the value comes from the sort of behind-closed-doors interactions that an AI simply doesn't have access to, even as it scans terabytes of public information. With management consulting, the issue isn't just about producing the right strategy. It's about assembling the right team to implement it. Senior consultants get good at reading the room based on decades of experience across companies and across industries and help CEOs make often-radical changes in their leadership teams during a major pivot.
There is also an issue of comfort level. Consultants become valuable not just because they're smart but because they're trusted, based on years of successful interactions with a particular client, sometimes on a personal level.
Yes, the comfort issue can lead to unseemly fawning. I vividly remember sitting at a bar at Pebble Beach after a day of golf as a Fortune 500 CEO client ordered rounds of different grappas for a few of us and explained their nuances. I nodded along and carefully sipped each shot, even though I hate grappa. It smells like nail polish remover to me.
The need for comfort can also, as both the WSJ and Nocera noted, lead to CEOs hiring major consulting firms for huge fees just to ratify decisions they would have made anyway. (At my firm, Diamond Management & Technology Consultants, acquired by PwC in 2010, we griped that McKinsey got most of that business, while we, as a far smaller brand, had our noses pressed against the window, watching with envy.)
But management teams do need support as they make decisions that risk the company's future—and their own careers—and experienced consultants at firms with long histories can provide that support in a way that an AI cannot.
Deeply experienced human underwriters, claims professionals, and agents and brokers provide the same sorts of emotional intelligence and comfort/support to those they serve, as well, making a lot of their work safe from gen AI.
All that said, gen AI will require changes to consulting firms' business model. The firms don't just make gobs of money for their partners by charging such hefty fees for their time. The firms are built as pyramids, with thousands of young MBAs at the lower tiers being billed at rates far above their salaries—the excess going to the partners. But much of the work done at the lower levels, including those slick PowerPoints highlighted in the WSJ headline, can, in fact, be automated by AI. A big chunk of revenue will disappear.
I say good riddance to a lot of that work. I remember teams at Diamond staying up all night because, as members combined their work, they found that fonts didn't match or that someone had messed up the page numbers, and they had to get everything perfect for the presentation first thing in the morning. Who needs to pay a bunch of 20-somethings $1,000 an hour to fix formatting when an AI could do the work in seconds and let them get some much-deserved sleep?
But the thinning out of lower ranks doesn't just take away revenue. It takes away talent, too—an issue for insurers, as well. If you don't need so much research done by the lower tiers in a firm, you don't need as many people, and you don't have as much talent fighting its way up the pyramid in the up-or-out culture of a consulting firm. Those who make it will still be impressive, but what have you lost?
The same goes for insurers. What do you lose among your senior underwriters, claims professionals, and agents and brokers in a decade if gen AI means you're drawing from a smaller base?
I remain a huge fan of the potential of gen AI, and I could tell you loads of stories mocking consultants and their firms, but I think the glee about the supposed comeuppance for consulting firms is overdone. They, like key players in insurance, will have to adapt plenty, but I predict they'll be as influential as ever.
Cheers,
Paul