What happens to consulting economics when analysis that once took a team several days can be completed in hours?

For decades, management consulting has operated on a familiar model. Assemble a team, allocate consultants, analyse the problem, produce the recommendations and charge the client according to the people, expertise and time required.

AI changes the arithmetic. Research can be accelerated, data analysed faster, interviews synthesised almost instantly and presentations drafted in minutes. Work that once required five consultants may eventually require three. Work that took three weeks may take one.

That’s an extraordinary productivity dividend. But who gets it — McKinsey, the client or the consultants?

If BCG can deliver the same quality engagement with substantially fewer professional hours, should its margins rise? Should the client pay less? Should consultants work fewer hours? Or should the firm simply undertake more work?

This is bigger than pricing. The traditional consulting model is built partly upon leverage, with partners and senior leaders supported by progressively larger groups of consultants doing analysis and execution. AI potentially changes the shape of that pyramid.

And that raises an even more fundamental question. If clients themselves have access to increasingly capable AI, what exactly are they buying from Bain, BCG or McKinsey?

Probably what they were buying all along: judgement, pattern recognition, experience, independence, facilitation and confidence in difficult decisions.

AI doesn’t necessarily diminish those things. It may make them more visible.

The future of consulting therefore isn’t simply about consultants learning to use AI. It’s about deciding what a consulting firm is selling when it is no longer selling professional labour.