Automating client intake workflows

The intake you already have

The inquiry arrived on Thursday. It was a good one, referred, squarely inside the practice, and it sat in the shared inbox until Monday because everyone who could answer it was billable elsewhere. When the client finally came in, someone re-typed the intake form into three systems by hand. The conflicts check was a partner searching memory and emailing the others: “does anyone know these people?” And the matter opened the way matters open in most small firms, as an email thread with attachments.

None of this loses a case. It loses clients, quietly, before they become clients.

Referred work is forgiving; it waits. Everything else compares the firm against whoever answered first, and in legal and advisory work the first substantive reply wins more often than partners like to admit. Meanwhile the people doing the triage, reading inquiries, chasing forms, drafting the same engagement letter for the fortieth time, are the most expensive people in the building.

What client intake automation covers

Six pieces. AI where reading and drafting help, plain automation where it’s better.

• First response and qualification. Every inquiry acknowledged within minutes, around the clock, and read for matter type, urgency, and fit, so a partner sees a briefed lead instead of a raw email.
• Intake forms that file themselves. The client answers once; the answers populate the practice-management system, the billing record, and the file. Nobody re-types anything into a third screen.
• Conflicts screening. Parties and related names pulled from the inquiry and checked against the client list before anyone books a call. The result is a flag for a human, never a decision by a machine.
• Engagement letters drafted from intake data. The same approach that works for proposal writing: the letter assembles itself on your template, and a partner reviews the terms before it leaves the building.
• Matter opening on signature. The signed letter triggers folders, task lists, and billing codes. The matter exists in every system it needs to, at once.
• Follow-up that doesn’t depend on memory. Unresponsive leads and unsigned letters are chased on a schedule, in the firm’s voice, until a person closes the loop.

No prospective client should learn how stretched the firm is from how long it takes to hear back.”

The pipeline, end to end

One inquiry, moving through the firm without waiting on anyone’s inbox:

On law firm intake software

Off-the-shelf law firm intake software exists, and much of it is genuinely useful; where a tool fits, I say so in the first two weeks rather than build around it. The difference in this work is not the category, it is the fit: systems shaped to the way your firm actually takes on work, built on the practice-management stack you already run, and owned outright when the engagement ends. No per-seat license that grows with headcount, no vendor roadmap deciding what your intake does next year. And in every configuration, on every matter, a human confirms conflicts and terms before anything binds the firm.

What stays human

The system reads, drafts, files, flags, and reminds. It does not decide. Whether the client is one the firm should act for is a partner’s judgment. Fee conversations happen between people who can read a room. And the conflicts decision itself, not the search but the decision, stays with whoever carries that responsibility today. The point of automating intake is to take the clerical weight off those judgments, not to touch them.

The shape of the engagement

This page describes one industry’s version of the work a Fractional AI Officer does. The shape is the same everywhere: ninety days embedded in the firm, roughly three production systems shipped, each with monitoring, an evaluation suite so quality can’t silently regress, and a kill switch you control. Everything is documented in a one-page runbook, the IP transfers to the firm in full, and your team is trained to run what got built. How those systems keep earning their keep after I step back is its own discipline; I wrote it up in AI systems that stay in production.

The terms are stated plainly. A $5,000 two-week paid diagnostic maps the intake workflow and returns an ROI-ranked plan; it credits toward the engagement if we proceed. The engagement itself is $30,000 for the first three firms, the Founders’ Edition, with continuation at $7,000 a month, and $50,000 Standard after that. By day ninety the pattern across this work is roughly twenty hours a week reclaimed, about one FTE of capacity unlocked, and a 15–25% lift in margin. The first documented engagement, Carol Coelho Náutica, is written up end to end with the numbers the firm reported.

Not for every firm

If the firm takes a handful of matters a year, all referred, intake automation will not pay for itself; change nothing. If you already have a technology lead building this in-house, you don’t need me in the chair. And if the intake problem is really a positioning problem, too many wrong-fit inquiries because the market doesn’t know what you do, automation will only answer the wrong people faster. The fit is a founder-led firm between $1M and $10M in revenue, five to thirty people, where the partners can name the hours intake costs them and want them back.