How Can a Small Law Firm Automate Client Intake With AI?

What parts of law firm client intake can AI actually automate?

The parts of client intake AI can automate are the answering, the collecting, the qualifying and the scheduling. The lawyer's assessment of the matter stays with the lawyer, and the line between those two is the whole design.

Answering is the first and the one that pays for the rest. A prospect who calls at 6pm on a Thursday about a car accident either reaches somebody or calls the next firm on the page. A system that answers every call and every form submission, in plain conversation, at any hour, removes the single largest leak in a small practice.

Collecting is the second. Names, dates, the other parties involved, where it happened, what documents exist, what deadlines are already running. Lawyers ask the same opening questions in a given practice area every time, and a system that asks them consistently produces a better file than a rushed phone note.

Qualifying is the third and the most valuable to a busy lawyer. Firms have criteria: practice areas they take, jurisdictions they are admitted in, matter thresholds below which the economics do not work, case types they refer out. A system applies those criteria to what it collected, so the consultation calendar fills with matters the firm actually wants and the rest are told promptly and referred rather than left waiting.

Scheduling closes it. The consultation is booked into the calendar while the prospect is still engaged, the confirmation goes out, the reminder goes out, and the intake summary reaches the lawyer before the meeting rather than during it.

Two things sit outside all of this. A system does not tell a prospect what their rights are, what their case is worth or what they should do, and it does not decide whether the firm takes the matter. Those are the lawyer's, and a build that blurs it is a liability rather than a system.

Why do so many law firms lose prospective clients at intake?

Law firms lose prospective clients at intake because responsiveness collapses at exactly the moment a prospect is deciding, and the firm never finds out it happened.

The research on this is unusually direct. In the secret shopper study inside Clio's 2024 Legal Trends Report, a third party contacted 500 US law firms as prospective clients. Only 33% responded to an email inquiry, down from 40% in 2019, and only 40% answered the phone, down from 56% in 2019. Among the firms that did reply by email, 84% replied within eight hours, and yet just 18% provided clear next steps or any cost information. The study found 73% of shoppers were unlikely to recommend the firms they contacted.

The shape of that result is worth noticing, because it is not a speed problem. The firms that answer, answer quickly. What goes wrong is that most contacts reach nobody at all, and the ones that land often arrive without the two things a person deciding whether to hire a lawyer needs: what happens next and what it costs.

The cause is not indifference. It is that intake competes with billable work and loses. A lawyer in the middle of a deposition prep is not picking up an unknown number, and in a firm with two lawyers and one assistant there is no second line of defense. Every inquiry that arrives during court, during a client meeting or after six in the evening is a coin flip.

The firm never learns the cost of this, which is why it persists. A prospect who gets voicemail does not send a complaint. They hire the firm that picked up, and the loss shows up as a quiet ceiling on new matters that gets blamed on marketing spend.

An intake system changes the default from "answered when somebody is free" to "answered", and the lawyers stay in the work they are paid for.

What are the options for automating law firm client intake, and who is each one for?

There are five ways to cover intake without building anything, and a built system is the sixth. Your own people answering when they can, a live answering service, an AI receptionist that answers and qualifies, a legal intake call center, or a legal CRM that captures the inquiry and drives the follow-up. The right answer is the cheapest one that fits the way your firm is actually losing matters, and for most small firms that is not the most capable option on the list.

Five things decide it before price does. Who answers at eleven at night, and whether it is a person. Whether the call gets qualified against your practice areas or only logged as a message. Whether the matter lands in the software you open every morning. Whether every conversation is logged and readable, since supervision under Model Rules 5.1 and 5.3 is judged on what the firm can review. And whether the questions stop at the minimum needed to decide on the matter, for the Rule 1.18(d) reason above.

Route	What it does	What it costs	Who it is for
Your own staff, nothing bought	Answers when somebody is free, takes a message when nobody is	No new cost, and the Clio secret shopper study is what it produces: 40% of 500 US firms answered the phone	A firm whose phone is covered through the day and that is not spending to bring new matters in
Live answering service Ruby	Live receptionists around the clock, lead qualification and intake, scheduling	Billed by the minute: $250 a month for 50 minutes, $395 for 100, $720 for 200, $1,725 for 500. AI features at no extra cost. No overage rate published	A firm that wants a human voice answering this month and will do its own qualifying. Minute billing punishes long intake calls
AI receptionist with a legal hand-off Smith.ai	AI answers first and can transfer to a live receptionist, books appointments, writes the call detail to your CRM	Billed by the call. AI receptionist free for 25 calls a month, $150 a month at $2.00 a call, $500 a month at $1.67 a call. Live-staffed plans from $300 a month for 30 calls, overage $11.50. First CRM integration free and Clio named. Add-ons per call include a conflicts check at $0.50	A firm with real after-hours volume that wants per-call billing and its calls landing in Clio
Legal intake call center Alert Communications	New client intake around the clock, screening to identify the right cases, web forms and online leads, CRM and case management integrations	No rate card published. Priced on volume and quoted	Plaintiff and mass tort firms with campaign-driven volume and sign-up paperwork behind every lead
Legal CRM with intake built in Lawmatics	Client intake, lead management, custom forms, e-signature and SMS on its entry tier, with workflow automation and advanced conflict checking above it	No rate card published. Every tier sends you to a demo or to sales	A firm that answers its inquiries and then loses them, where the gap is follow-up rather than coverage
A built intake system BLACKSIG	Our engineering team builds the conversation around your own intake script and criteria, writes the matter and the intake summary into the software you already run, and escalates what it should not handle	Scoped on the call. Channel count and practice area count move it far more than the size of the firm	A firm with several practice areas on different criteria, software it will not leave, and a ceiling it can feel on new matters

When is a built intake system the wrong choice for a firm?

A built intake system is the wrong choice until the intake conversation has to apply criteria that differ by practice area and the result has to arrive inside the software the firm already runs. We sit at the end of that list deliberately, and three situations send a firm somewhere else.

One phone line, one practice area and a problem you want gone this week is an answering service. Ruby or Smith.ai will be answering calls before our Discover stage finishes, and we will say so on the call. If the inquiries get answered and then go quiet, the gap is follow-up, and the cheaper fix is configuring the CRM you already pay for. A practice taking a handful of new inquiries a month does not have the volume to repay a build either, whatever the arithmetic looks like per call.

Two of the options above publish no rate card. That is ordinary in this market rather than evasive, and both will quote on volume. What the table cannot show is what happens in the first thirty seconds after the call is answered, which is where firms end up disappointed. A receptionist who does not know your practice areas takes a message, and the qualifying conversation happens later or never. Ask any option you are considering to show you what lands in your practice management software, not what the caller hears.

Where a build is the right answer, it lives on our infrastructure and we run it from there, with a report of what it handled, what it escalated and what it produced.

Every figure in that table was read off the named company's own pricing page on 23 October 2026. Vendor pricing moves faster than anything else on this page, so we recheck the table each quarter and date it here.

How does AI handle intake without giving legal advice or breaching confidentiality?

An intake system stays clear of legal advice by collecting rather than advising, and it protects confidentiality by being built so that a lawyer supervises what it does and every conversation is recorded and reviewable.

The functional line is easy to state and has to be enforced in the build rather than hoped for. A system asks questions, records answers, explains process and schedules time. It does not evaluate the merits, predict outcomes, quote a settlement range or tell somebody what to do about a deadline. Where a prospect asks any of that, the honest and correct behavior is to say a lawyer will answer it and to get the consultation booked.

Intake task	Handle it automatically	Keep it with a lawyer
Answering the call or form at any hour	Yes
Collecting facts, parties, dates and documents	Yes
Screening against the firm's practice areas and criteria	Yes, as a screen, not a decision
Running names for a conflicts check	Yes, as a search that surfaces hits	Yes, the clearance decision
Telling a prospect whether they have a case		Yes, always
Quoting a fee or a likely outcome		Yes
Accepting or declining the representation		Yes

Confidentiality deserves its own paragraph rather than a row. Prospective client information is protected before anyone signs anything, so an intake system is handling sensitive data from the first sentence of the first call. The questions that decide whether a build is acceptable are where that data lives, who can reach it, how long it is retained, whether it is used to train anybody's model, and whether the whole conversation can be produced later. Those are answerable in writing and a firm should require the answers in writing.

Two further things a firm should insist on. Every automated conversation is logged and readable, because an intake exchange that cannot be reviewed is a supervision problem rather than an efficiency gain. And the system discloses that it is an assistant rather than a person, because a prospect who later feels misled is a bar complaint waiting to be written.

We build to that line and hand the compliance judgment to the firm's own counsel, which is where it belongs. No vendor can certify a firm against its bar's rules, and any who says otherwise is selling something.

What do the bar rules say about using AI for client intake?

No rule forbids a law firm from automating intake. The ABA has published guidance that applies the existing Model Rules to AI rather than writing new ones, and the two duties that bite hardest at intake are the duty owed to a prospective client and the duty to supervise.

ABA Formal Opinion 512, issued by the Standing Committee on Ethics and Professional Responsibility on 29 July 2024, is the first ABA formal ethics opinion on generative AI. It covers competence, confidentiality, communication, candour, supervision and fees, and it treats AI as something the rules already reach. On intake it notes that the Rule 1.18(b) confidentiality duty extends to AI tools used with prospective clients.

That is the part firms underestimate. Under Model Rule 1.18(b) a lawyer who learns information from a prospective client may not use or reveal it, subject to the same exception that applies to a former client under Rule 1.9. The duty attaches during the consultation, which is before any system has decided whether the matter is a fit and before anybody has signed anything. An automated intake conversation at eleven at night is covered from the first answer.

The consequence runs against the instinct every intake form is built on. ABA Formal Opinion 510, issued in March 2024, advises asking a prospective client for the minimum information reasonably necessary to decide whether to take the matter, and warning them not to volunteer more than is asked. The reason is Rule 1.18(d): the more a lawyer learns from a prospect the firm then declines, the more likely the whole firm is disqualified from a substantially related matter. A system built to collect everything it can is working against that. The design goal is the smallest sufficient set of questions, asked in the right order, with the names of the parties taken early enough to run a conflicts check before the conversation goes near the merits.

So four things belong in the build rather than in a policy document.

Names and parties first, and a conflicts search before the questions that go to the merits.
A ceiling on what gets asked, with the invitation to tell the whole story kept back until the firm has decided it wants to hear it.
A disclosure that the caller is talking to an assistant, and a line stating that contacting the firm does not by itself create representation.
Every conversation logged and readable, because supervision under Rules 5.1 and 5.3 is judged on what the firm can review rather than on what it intended.

One caveat matters more than the rest. The ABA Model Rules bind nobody until a state adopts them, state versions differ, and a formal opinion is guidance rather than law. Both opinions are worth reading in the original alongside the firm's own counsel, who makes this call.

What does automating intake change about the firm's capacity?

Automating intake raises the number of matters a firm signs without changing the number of lawyers, which is the only lever in a practice that does not require hiring.

Law firm economics run on how much of the day turns into billable work, and the answer is less than most lawyers assume. Clio's 2025 Legal Trends Report put the average utilization rate at 38%, which works out to about three billable hours in an eight hour day. The other five go to administration, communication, scheduling and the reconstruction of time entries at the end of the day.

Two ways out of that exist. Push lawyers to bill more hours, which the profession has been trying for forty years. Or take the non billable work out of the day, which is what a system does. Intake is the best first target because it is both non billable and revenue determining. Every hour of it is unpaid, and every missed call is a matter that went to a competitor.

The gain is not abstract. A firm answering every inquiry instead of four in ten is working a fundamentally different top of funnel with the same marketing spend and the same lawyers. That is what capacity means in a practice.

What makes it stick is unglamorous. The system has to write into the practice management software the firm already runs, so the matter, the contact, the intake summary and the calendar entry land in Clio, MyCase, Filevine, PracticePanther or whatever the firm opens each morning, not in a separate tool nobody checks. And somebody has to keep it running when a vendor changes an API or the firm changes its intake criteria. Ours run on our infrastructure and we operate them from there, with a report showing what was handled, what was escalated and what it produced.

For a worked example of a system built into software a client already ran, see the Universidad Maimonides case study: about fifteen hours a week returned, running ever since on our infrastructure.

Frequently asked questions

How much does it cost to automate client intake for a small law firm?

Channel count and practice area count drive it more than firm size. One phone line, one web form and one practice area with a single intake script is a contained build. Four practice areas with different qualifying criteria, a Spanish language line, a referral partner who sends matters by email and a practice management system with a limited API is a larger one. Before asking anyone for a number, count the ways a prospect can reach you and write down the qualifying criteria for each practice area. Most firms have never written the second one down, and doing it is worth the afternoon regardless of what you build.

How does an AI intake system actually create the matter in Clio or MyCase?

Through the interfaces the practice management platform offers other software. The system authenticates as a service account with defined permissions, then creates the contact, opens the matter or lead record, attaches the intake summary and any uploaded documents, and books the consultation on the right lawyer's calendar. Every action is logged and reversible. Where a platform exposes less than the build needs, the fallbacks are scheduled file exchange and email based workflows, both of which work and both of which need more maintenance.

Is an AI intake system better than an answering service or a virtual receptionist?

An intake system and an answering service solve overlapping problems at different depths, and plenty of firms run both. An answering service gives you a human voice and a message, which beats voicemail and is the right answer for a firm that wants one thing fixed this month. Where it stops is depth: a receptionist who does not know your practice areas takes a message, and the qualifying conversation happens later or not at all. A built system asks your questions, applies your criteria, books into your calendar and writes into your practice management software. Compare them on what happens after the call is answered.

Can a solo practitioner automate intake, or does it only pay at multiple lawyers?

A solo is the strongest case rather than the weakest, because a solo has no second person to catch the call. When the only lawyer is in court, every inquiry that arrives goes unanswered, and the ceiling on the practice is set by hours in the day rather than by demand. Count the inquiries you received last month and the number that turned into a booked consultation. The gap between those two numbers is the size of the opportunity.

Can it run a conflicts check?

A system can run the names against your existing client and matter records and surface every possible hit, which is the searching part and the part that gets skipped when somebody is busy. The clearance decision stays with a lawyer, because deciding whether a hit is an actual conflict requires knowing the matters, and that is judgment the firm is responsible for. Treat automated conflicts searching as a way to make sure the search always happens rather than as a substitute for clearing it.

How do answering services and AI intake services price their work?

Three billing shapes, and which one is cheapest depends on how long your intake calls run rather than on how many you get. Per minute, as Ruby prices it, starting at $250 a month for 50 minutes. Per answered call, as Smith.ai prices it, from $300 a month for 30 calls with live receptionists, or $150 a month at $2.00 a call with its AI receptionist. Or by contact tier, the way the legal CRMs sell, which Lawmatics does without publishing a figure. A qualifying conversation that runs eight minutes is expensive on a per minute plan and costs nothing extra on a per call plan, so time your own intake calls before comparing headline prices. Figures read from each company's own pricing page on 23 October 2026.

Is it ethical for a lawyer to use AI for client intake?

Nothing in the ABA Model Rules forbids it. ABA Formal Opinion 512, issued on 29 July 2024, applies the existing rules to generative AI rather than writing new ones, and says the Rule 1.18(b) confidentiality duty owed to a prospective client extends to AI tools. The constraint firms least expect is ABA Formal Opinion 510, from March 2024: ask a prospect only for the minimum needed to decide whether to take the matter, because what the firm learns from a prospect it then declines can disqualify the whole firm under Rule 1.18(d). State rules differ, and the firm's own counsel makes the call.