Most AI-disruption talk points inward: which workflows get automated, which functions get leaner, which roles change. The bigger story points the other way, at the firms that sell advice to the enterprise.
The billable hour, the seat license and the retainer are the same instrument. Each one meters human time, because until now human time was where the value came from. Agents break that instrument.
When analysis that took 60 hours takes six, the hour stops measuring worth. Every business that bills time now has one choice: reprice, or shrink.
This is no longer a forecast. The shift already shows in the numbers.
The hour is failing in public
Professional services firms are repricing in real time:
- McKinsey reportedly now earns more than 30% of global fees from outcome-tied arrangements rather than time-and-materials.
- BCG's leadership has said AI-related work would grow from roughly 20% of 2024 revenue toward 40%.
- Bain puts AI- and tech-enabled work near a third of its business and climbing.
- EY leaders openly describe a "service-as-software" future: advice delivered as a product, not a timesheet.
Then the platform vendors walked onto the same field:
- Microsoft launched Frontier Company on July 2: $2.5 billion and roughly 6,000 embedded experts, described as an "outcome-driven engineering organization," not a consultancy.
- AWS stood up a $1 billion embedded-engineer unit two days earlier, on the same premise.
When the hyperscalers sell outcomes rather than licenses, they are not entering the software market. They are entering the advisory market, and pricing it the way incumbent firms are only starting to.
The direction runs one way. Time is losing value as a unit at the same moment AI raises what one expert can produce. Where those two lines cross, the hours-based model stops being a business and becomes a liability.
Why AI breaks time-based pricing
Every earlier productivity wave made professionals faster, and the hourly model absorbed it: faster work meant more matters billed.
AI is different. It separates output from hours, instead of shrinking the hours per piece of work.
A tax memo, a market model, a due-diligence pack, a controls assessment: each once needed a person to spend the hours. The client paid for those hours as a stand-in for the judgment inside them. An agent that drafts a defensible first version in minutes does not make the judgment worth less. It makes the hour a false signal of where the value is.
A firm that keeps billing by time in that world has two options, and neither works:
- Bill for hours it no longer spends. That cannot last.
- Hand the whole productivity gain to the client for free. That is a slow liquidation.
From billable hours to billable assets
The way out is not to defend the hour. It is to change what gets sold: from time spent to owned, reusable work that compounds across clients and quarters.
Finance shows the shape most clearly, in three layers:
- Reason. Frontier models supply judgment on demand. They are bought, interchangeable and cheaper every quarter. This is exactly where not to look for an advantage: every competitor buys the same thing on the same terms.
- Delegate. Agents run multi-step workflows under human oversight. The advantage goes to whoever encodes the process well, but a well-run agent is still, on its own, a cost.
- Manufacture. Each engagement leaves behind something the firm owns and can sell again: an ontology, a diagnostic, an agent suite, a benchmark. This is where hours become assets, and where the margin stops flowing to the client.
Spend that expires is a cost. Spend that becomes an owned, auditable capability is equity. A firm that has run a diagnostic 200 times and sells the 201st in an afternoon is no longer billing time. It is earning back an asset. That is the whole game.
What it means for whoever owns the P&L
For the firms that sell advice: stop measuring utilization and start measuring reuse.
- The number that predicts survival is not hours billed per head.
- It is the share of revenue that comes from reusable assets rather than one-off effort, and how fast that share is growing.
- A partnership that cannot answer "what did we build this quarter that we can sell again next quarter?" is running a business that is melting.
For the enterprises that buy advice: stop paying for hours and contract for outcomes and reusable work.
- Buy the diagnostic that stays, the agent that keeps running, the benchmark that updates.
- That way the firm's productivity gain reaches you, instead of disappearing into a timesheet.
For everyone: Microsoft's $2.5 billion is the warning. Platform vendors will sell outcomes over the heads of anyone still selling hours.
The productivity gain from AI is real and large, and most of it currently leaks (Workday, Jan 2026):
- 85% of employees already save one to seven hours a week with AI.
- Only 14% consistently turn that into positive net outcomes.
- Roughly 40% of the time saved is lost reworking the output.
The open question is who captures the gain. Either the firm that turns expertise into an owned, governed asset it can price, or the client and the platform that route around the one still counting time.
The model was never the moat. Neither was the hour. What you build and own is.