■ HIGH RISK ■ Management & Business
The pyramid is collapsing from the bottom. Partners who sell trust and take blame are fine; the army of analysts building slides beneath them is the part AI genuinely replaces.
“AI generates the PowerPoint. But charging $500/hour to present it? That's a uniquely human skill.”
Our AI replacement risk score — how we score jobs
Consulting economics have always rested on leverage: a partner sells the engagement, and a stack of juniors does the market sizing, the competitor benchmarking, the expert-call synthesis, the Excel model, and the two hundred slides that get cut to forty. That labor was billed at rates that assumed it took humans weeks. Much of it now takes an afternoon — desk research, first-pass market models, synthesizing interview notes into themes, drafting the storyline, even producing clean charts are all things current tools do competently enough to compress a team of four into a team of one plus review time.
Clients have noticed, and that is the real threat. Corporate strategy teams now run their own analysis in-house with the same tools, which erodes the information asymmetry that justified the fee. Benchmarking decks and 'here's what good looks like' reports — historically bread-and-butter work — are close to commoditized. Firms are responding by shifting toward implementation, where value is measured in delivered change rather than delivered documents, and toward AI transformation work, which is currently lucrative and probably cyclical.
What holds up is less flattering to the profession and more durable than the analysis. Executives buy consultants for political cover: an external name to legitimize a decision already made, to deliver news the CEO cannot say internally, and to absorb blame if a restructure goes badly. They buy access to a partner's judgment built from watching thirty similar situations fail. And they buy execution — sitting inside a client organization for nine months, navigating its factions, getting a merger integration actually done when the two IT departments hate each other. None of that is a document-generation problem. Our score of 58 splits the difference: the deliverable is cheap now, the relationship and the accountability are not.
Automatability: our editorial assessment of current and near-term AI capability
The analyst layer is under pressure now, with firms already slowing graduate intake and expecting more output per junior. Through roughly 2030 expect flatter pyramids, smaller teams on the same scope, and fee pressure on any research- or benchmarking-heavy work as clients replicate it internally. Implementation, restructuring, and transformation delivery hold up longest, because they require bodies inside the client's building making things happen.
It remains a strong training environment, but the deal has changed. You will be expected to produce senior-level output almost immediately because the grunt-work apprenticeship is compressed, and intake classes are smaller. If you join, treat it as two to three years of accelerated exposure rather than a partnership track, and push hard for client-facing and implementation experience early.
Never is a long word, but the sticky parts are political and physical: being the outside voice that makes an unpopular decision survivable, reading a boardroom where two executives are quietly fighting, and living inside a client organization for months to force change through. Those are accountability and presence problems, not analysis problems.
For research-heavy deliverables, increasingly no — corporate strategy teams can now produce a passable version themselves. Rates hold where the firm carries risk, brings scarce sector expertise, or supplies people who actually execute. The industry's shift toward outcome-linked and implementation work is a direct response to that repricing.
Get known for something specific rather than for being smart in general. Learn to deliver a workstream with half the historical team by leaning on AI for research and drafting, then spend the reclaimed time in front of the client. And build the execution muscle — running a program, managing resistance, hitting a milestone — because that is where the fees are migrating.