McKinsey, the firm whose entire business is advising other companies how to operate, is being forced to dismantle the pricing model that built it. Artificial intelligence has compressed the work — a project that once required a six-person team for three weeks now takes one person and a few prompts — and consulting’s billable hour, the unit by which McKinsey, the Big Four, and the whole industry sold their labor for a century, no longer corresponds to the work being done. The firm has begun moving to outcome-based pricing, charging for results rather than time, because the time has collapsed and can no longer be charged. The consultant who spent the last two years telling every client that AI would transform their business has discovered it transforming his, and is now quietly renegotiating the meter the entire profession billed by.
The Unit That Collapsed
The billable hour was a near-perfect business model for fifty years, and its perfection lay in a single alignment: the firm sold time, time was scarce, and the more complex a client’s problem, the more time the firm could justify selling to solve it. Complexity was not an obstacle to consulting; it was the inventory. A difficult analysis that consumed a six-person team for three weeks was not a cost to be minimized but a product to be billed, and the entire pyramid of the profession — the armies of junior analysts, the leveraged teams, the hours logged against the engagement — was an apparatus for converting difficulty into revenue at an hourly rate.
Artificial intelligence severs the alignment at its root. The analysis that filled the three weeks can now be produced in an afternoon, which means the hours that were the product have evaporated, and a firm cannot bill for time that the work no longer requires. The complexity that was the inventory has been commoditized; the difficult analysis is now cheap, and cheap analysis cannot sustain the rates the difficulty once justified. McKinsey’s move to outcome pricing is not a strategic innovation. It is a forced retreat from a unit of value that the technology destroyed — an admission that the firm can no longer sell the hours, because the hours have stopped corresponding to anything a client would pay for.
The retreat is early and incomplete; only a quarter of the firm’s fees are tied to outcomes, the rest still billed the old way against a clock that ticks slower every quarter. But the direction is fixed, and the destination is legible. When you sell hours, you sell labor — the time and judgment of specific people. When you sell outcomes, you sell results regardless of the labor behind them, which is a quiet but total concession: the labor is no longer the thing of value. The profession built on selling the analyst’s time is conceding that the analyst’s time is no longer what the client is buying, because the analysis can now be produced without spending much of it.
The Consultant Consulted
The justice of it is structurally exact, and worth naming without mercy. Consulting’s product was always, in significant part, the importation of efficiency — the firm arrived at a client, identified the redundant headcount, recommended the reductions, and billed handsomely for the diagnosis. McKinsey built an empire on telling other companies how to do more with fewer people. Artificial intelligence now does to McKinsey precisely what McKinsey did to its clients: it finds the work that can be done with fewer people, and the people in question are the consultants. The diagnostician has contracted the disease he made his fortune diagnosing, and the symptoms are the ones he always described to others as a transformation rather than a loss.
This is the same machine I described being assembled when the AI labs and the private equity firms formed a venture to embed engineers inside companies and redesign their workflows — a venture that named consulting, explicitly, as the industry it intended to take. That venture is the automated successor to the consultant: it does what McKinsey did, the workflow redesign and the efficiency diagnosis, but it does it with a model rather than a leveraged pyramid of analysts, and it is owned by the labs and the financiers rather than the partners. The consultant who sold efficiency for fifty years is being made efficient by a product built in the image of his own service, and sold to his own clients, at a price his hours cannot match.
What McKinsey is discovering, the whole class of professional advisers will discover behind it — the lawyers billing for document review, the accountants for reconciliation, the analysts for the model. Each profession built its economics on selling expert time by the hour, and each is meeting the same severance of time from work. The billable hour is not a McKinsey problem. It is the load-bearing assumption of the entire professional class, the belief that expert labor is scarce and therefore chargeable, and the technology is dissolving the scarcity across every profession that ever sold an hour of expertise as though the hour were the value.
What This Means
Outcome pricing sounds like a clever adaptation, and it is a trap with a longer fuse. If the outcome can be produced by a model with one person and a few prompts, then selling outcomes is only a slower path to the same cliff — because the client will eventually ask why it is paying a firm for a result that a model produces directly, and the firm’s honest answer will have nothing to do with the analysis. It will have to do with trust, with the relationship, with the accountability of a name and a logo on the deck, with the senior partner in the room whose presence means someone can be blamed. The analysis becomes the machine’s. What the firm retains is the handshake.
That retention is real, and it is also a narrowing of exactly the kind the doctors and the mathematicians are living through. The consultant survives, but as a brand and a relationship rather than a producer of analysis — the famous logo that makes a board feel safe, the trusted adviser who absorbs the accountability, the human presence that lets an executive say the decision was blessed by someone who could be held responsible. These are the same authenticity premiums that survive the synthetic flood, the irreproducible human credentials that retain value precisely because the reproducible analysis no longer does. The profession keeps the part that was never really the work, and loses the part that was.
So the firm that diagnosed efficiency for the world has received its own diagnosis, and the diagnosis is the one it always delivered to others: the work you sold can be done with fewer of you, and the market will not keep paying for the version that requires more. I note that McKinsey, of all institutions, will understand the prognosis perfectly, because it wrote the prognosis. It spent half a century arriving at companies to explain, in tasteful slides, that the comfortable headcount was an inefficiency the market would eventually correct. The market has arrived at McKinsey now, carrying the same slides, and the firm is doing what its clients always did when it delivered them — renaming the loss a transformation, and renegotiating the bill on the way out.