Skip to content
Dom Conte
← All writing
Legal AI in practice 4 min read

The billable hour isn't dying. It's being unbundled.

Everyone predicts the death of the billable hour and keeps being wrong. The hour isn't dying. The work it used to price is being pulled apart, and the hour only survives on one of the pieces.

The death of the billable hour has been predicted so confidently, for so long, by so many people, that the prediction has become a kind of industry folklore. It is always about to happen. It never quite does. Alternative fee arrangements arrive, get adopted at the margins, and the hourly rate sails on.

I think the people predicting its death have the right intuition and the wrong noun. The billable hour isn’t dying. The bundle it used to price is being unbundled, and the hour is quietly retreating to the only part of the work where it still makes sense.

What the hour was actually pricing

Spend any time looking closely at a legal matter and you notice that “an hour of a lawyer’s time” was never one thing. It was a bundle. Inside that single, uniformly-priced unit you’ll find at least three different kinds of work, sold together because historically they were inseparable:

  • Assembly. Gathering documents, populating templates, cross-checking references, formatting, comparing versions. Necessary, valuable, and almost entirely mechanical.
  • Application. Running the firm’s known method over a known situation. Issue-spotting against a familiar checklist, applying the standard treatment, drafting the standard clause. Skilled, but repeatable.
  • Judgement. The genuinely contingent bit. What this specific client should do, given this specific risk, in this specific context, where the answer isn’t in any precedent and being wrong is expensive.

For a hundred years these travelled together at one blended price, because the only thing that could do the assembly and the application was the same person who held the judgement. The bundle wasn’t a pricing philosophy. It was a technological constraint.

The reason this is the moment, and the previous predictions were early, is that AI does something the alternative fee arrangement never could. AFAs tried to change the pricing of the bundle while leaving the bundle intact. AI breaks the bundle itself.

The assembly layer is the first to go. A capable system gathers the documents, populates the template, cross-checks the references, lines up the versions. The application layer is going next: run the firm’s method over the situation and produce a strong first pass. What’s left, sitting on top, is the judgement - and judgement is the one thing in the stack that genuinely does scale with the time and attention of an expensive human being.

So the hour doesn’t die. It un-bundles. It gets stripped off the assembly and the application, where it never made sense anyway, and concentrates on the judgement, where it always did. The category that survives the longest billing by the hour is the category where the hour was actually measuring the right thing.

The pricing problem this creates

Here’s the part firms find genuinely uncomfortable. Once you unbundle, you cannot keep charging the old blended rate for the assembly and application, because the client can now see roughly what those cost - which is almost nothing. The value didn’t disappear; the visibility changed. The market is going to insist that the productised layers get priced like products: fixed, per-output, per-seat, per-outcome.

That’s a margin shock for any firm whose profitability secretly depended on charging judgement prices for assembly work. A lot of practices are more exposed to this than they admit, because the comfortable matters - the high-volume, repeatable ones - were quietly subsidised by the bundle. Unbundle them and the subsidy evaporates.

The firms that come out ahead are the ones that get there first and on purpose. They unbundle their own work before a client or a competitor does it for them. They price the assembly and application layers honestly and cheaply, win volume on those, and reserve the hour for the judgement, where they can charge confidently because the value is legible and real.

What to actually do with this

If you run a practice, the useful exercise isn’t to debate whether the hour lives or dies. It’s to take your three most common matter types and cut each one along these lines. How much of this is assembly? How much is application? How much is genuine judgement?

For most repeatable work, partners are shocked by how thin the judgement slice actually is - and relieved, because the thin slice is the defensible, high-value, genuinely-bill-by-the-hour part. The rest was always going to be productised. Better to be the one who productises it, prices it deliberately, and keeps the margin, than to wait for the unbundling to be done to you at a price someone else sets.

The billable hour is not on its deathbed. It’s just being told, after a century, that it was only ever good at pricing one of the three things it sold - and that from now on, that’s the only thing it gets to price.

Written by Dom Conte

Legal-tech founder, builder and speaker. More about me →