The AI-Ready Firm: Connect, Reprice, Automate — In That Order

An AI-ready law firm can see what AI is doing to its money. That’s the whole definition, and it’s why the first move isn’t buying an AI tool. It’s connecting the systems that hold your financial data, so there’s a real answer when you ask what changed. Then repricing what the data exposes. Then handing AI the friction. In that order, and the order carries most of the weight.

This is the practical companion to Episode 5 of Off the Books, where Gary walks through the playbook he’s used with firms going through it now.

Why connect first, before anything else?

The instinct is to start with the tool, because the tool is what’s being sold to you. Every AI pitch a firm hears right now is about producing work faster: drafting, research, intake, summarization. All real. None of it tells you what the speed did to your revenue.

If time capture, billing, collections, and accounting don’t share data, your firm is already deciding in the dark, and AI moves faster in the dark without improving the view. A tool that drafts in half the time hands its output into the same broken relay between work performed and cash collected. You reach the same gap sooner.

Connecting first means one place where the money is visible: what’s in progress, what’s been billed, what’s been collected, and what any of that means for a given matter. Once that exists, “did AI change our economics?” has an answer instead of a debate. We’ve written separately about what real-time financial visibility actually means for a law firm; the short version is that a number you reconstruct at month-end is a historical artifact.

The move is one connected experience across the tools you already run, not one tool that claims to do everything. Consolidation hides fragmentation inside a single interface rather than removing it.

What does repricing actually mean here?

Once you can see the numbers, the numbers start making pricing decisions for you.

Under hourly billing, AI compressing a four-hour task into one hour is revenue evaporating. Under a fixed fee, the same efficiency is margin expanding. Same technology, opposite outcome, and your operating model decides which one you get. We trace that mechanic in more detail in how AI changes your law firm’s realization rate. (Realization rate, if you don’t track it: the share of the work you performed that made it onto an invoice.)

Two disciplines make repricing survivable.

Reprice from data, not from nerves. The reflex when AI shortens a matter is to drop the price, because charging the same for less time feels indefensible. But the work is worth what it’s worth to the client, and the hours were never the value. Track effective hourly rate, the fee divided by the hours it actually took, and a fixed fee becomes an evidence-based decision. Firms that reprice downward by reflex give away margin and can’t tell, because they weren’t measuring the rate they gave up.

Start with one practice area. The predictable, repeatable work, where you can already estimate the effort within a reasonable range. Prove the fixed fee there, watch the effective hourly rate for two quarters, then expand. Repricing the whole book at once is how firms end up quietly unprofitable in three practice areas simultaneously.

Be honest about the constraint: this is slow. It takes a couple of billing cycles before the data says anything trustworthy. That’s also why connecting has to come first, since the clock on repricing doesn’t start until the data exists.

Where does AI actually belong in the workflow?

Third, and in a narrower slot than the marketing suggests.

Hand AI your friction. Drafting time-entry narratives from work already logged. Flagging time that looks missed. First-pass research. Plain-language questions about your own numbers, like which matters are dragging realization this quarter or what’s aging in AR past 60 days. Repetitive, high-volume, verifiable work, with a human reviewing the output before it goes anywhere.

Keep your judgment. Pricing calls, client strategy, and verifying AI’s own output stay with people. The ethics rules require it, and it’s what clients are paying for. A firm that hands over judgment gets harder to defend without getting more efficient.

The second half of this move is the one firms skip: redeploy the time. Capacity that AI frees up either goes to higher-value work or evaporates into the day, and usually nobody decides which. We’ve written about what to do with the time AI frees up on your paralegal team, and the same logic runs at the attorney level.

Why the order matters more than any single move

Doing all three at once is how change efforts die.

Reprice before you can see the data and you’re guessing with a straight face. Automate before you reprice and you’ve made an unprofitable model run faster. Buying the AI tool first is the most common sequence, because it’s the one being sold, and it leaves you with a capability you can’t prove did anything, which is how it gets cut in the next budget review.

There’s a human reason too. Firms don’t adopt operational change because the plan is good. They adopt it because one number visibly moved and somebody noticed. Connecting first produces that early, cheap, undeniable win: a realization figure that’s current instead of six weeks old, an AR number a partner can look at on a Tuesday. That’s what earns the second move.

So the first step is smaller than the whole playbook. Pick one number, realization or days to collect, and find out whether you can see it right now, in real time, without asking anyone to build a report. If you can’t, that’s move one. You can’t reprice or automate your way out of a problem you can’t see.

Frequently asked questions

What does it mean for a law firm to be AI-ready? The firm’s financial data is connected enough that the effect of AI on revenue is measurable. A firm can adopt AI tools without being AI-ready; it just won’t be able to tell whether they helped.

Should a law firm reprice matters because of AI? Where AI meaningfully compresses predictable, repeatable work, yes, but based on tracked effective hourly rate rather than instinct. Repricing downward by reflex gives away margin the client wasn’t asking for.

What should AI not do in a law firm? Pricing decisions, client strategy, and verifying its own output. Those require judgment the ethics rules keep with the lawyer.

Why connect systems before adopting AI? AI acts on what it can see. On a disconnected stack it can’t see across the gaps between billing, collections, and accounting, so it answers narrow questions confidently and misses the ones that matter.

What’s the smallest first step? Pick one metric, realization rate or days to collect, and check whether you can see it in real time today. The answer tells you whether you’re on move one or move two.