For as long as professional services have existed, the deal has been simple: you buy hours, and you trust that more hours mean more value. AI has quietly broken that arithmetic. When a piece of work that took six weeks now takes two, billing by the hour means sending a smaller invoice for the same result, and asking the client to reward you for being slow. The incentive is upside down, and everyone in the room knows it.
The top of the market has already started to move. BCG now runs variable, outcome-tied fees on around three quarters of its largest AI engagements, with payment linked to what the client actually gains rather than the hours spent. Across the sector, firms are experimenting with tokenised and outcome-based pricing. Clients, meanwhile, have learned to wait, on the reasonable bet that the next model makes the project cheaper. Both sides are being pushed off the hour at the same time.
The hour is not dying because it is unfair. It is dying because AI broke its logic. The signature insight · Outcomes
Why the hour is breaking
Time-based billing only works when time is a fair proxy for value. AI has severed that link. The same output now takes a fraction of the hours, so the more capable the tool, the smaller the bill for identical work. A firm that prices on time is now penalised for its own efficiency, and a buyer who prices on time is paying for activity rather than result. When a proxy stops tracking the thing it stood for, the proxy gets replaced. What replaces it is the outcome itself.
But pricing on outcomes is far harder than pricing on hours, which is why most firms talk about it and few do it. An hour is easy to count. A result has to be defined, measured, and honestly attributed, and that is a discipline, not a billing change.
What the pattern actually looks like
- The invoice shrinks as the tool improves. Same result, fewer hours, less revenue, for being faster.
- Buyers defer. Expecting each model release to lower the price, clients hold projects back, and the sales cycle stretches.
- Value decouples from time. What the client gains and what the clock says diverge, and the clock loses.
- Outcome and tokenised pricing surface at the top of the market. The largest firms are already tying fees to results, which sets the expectation for everyone beneath them.
What outcome pricing actually demands
Whether you are buying transformation in this market or selling it, the shift only works on three conditions. They are also a fair test of whether a partner can really do it.
- Agree the outcome and how it is measured, up front. A small, real set of metrics in the client's own data, decided before the work starts, not reverse-engineered afterwards to flatter the result.
- Baseline before, re-baseline after. The delta only exists if there is a clean starting point to measure against. No baseline, no proof, no honest outcome fee.
- Be honest about attribution. Real change happens in a noisy system, and claiming laboratory-grade causation is how you lose a serious buyer. Calibrated evidence, openly caveated, is more credible than a confident number. That honesty is the discipline that makes outcome pricing trustworthy.
The takeaway
The billable hour is ending not because it was wrong, but because AI dissolved the link between time and value it depended on. The replacement is pricing on proven change, and that is only ever as good as your ability to measure the change honestly. The firms that can set a baseline, show the delta, and be straight about what they can and cannot claim will set the terms of the next decade. The rest will keep selling hours, for less and less.