The objection surfaces in every evaluation meeting, usually as a joke that is not one: “if your tool does the review in a tenth of the time, I bill a tenth of the hours.” It deserves a serious answer, because it is really a question about what clients are buying.

Clients do not buy hours

They buy outcomes: the closed deal, the filed opinion, the answered question. Hours are a proxy — and a proxy that AI degrades. When the reading layer of a matter compresses, the firms that cling to hour-counting are competing against firms that sell the same outcome for less, delivered faster, with a citation trail a client can check. That is not a technology argument; it is a market one.

What firms are actually doing

Value pricing on defined workstreams — the due-diligence review priced as a review, not as 140 hours.
Blended arrangements where the tool's speed widens margin instead of shrinking the invoice, and the client shares the saving.
Fixed-fee products built on top of the compression — a standard NDA triage at a published price, for example.
Reinvested partner time: the hours not spent on first-pass reading move to strategy, business development and the work that cannot be compressed.

Vaultr's own commercial model is built for this world deliberately: charges agreed in writing before installation, no automatic renewal, nothing metered on your matter data — in local mode we could not meter it if we wanted to. Pricing that depends on watching your usage and pricing that promises confidentiality are in tension; we picked a side.

Every profession that automated its reading layer — medicine's labs, accounting's spreadsheets — billed more for the judgement, not less for the work.

The firms that lose in this transition will not lose to software. They will lose to other firms that used software to make the judgement layer cheaper to reach.