home  /  marketing & business development  /  intake & lead handling
growth · ai for legal practice

Intake and lead handling.

Most firms lose matters to response time rather than to competitors. This is the fixable part — inside two firm limits.

begin here

Where is your firm?

Start a conversation with the AI Adoption Concierge, already scoped to intake & lead handling. Pick a starting point, or describe your situation directly.

AI Adoption Conciergeintake & lead handling · orientation, not legal or ethics advice
Tell me how enquiries reach the firm now and how quickly they get answered. I'll help you think about what a system can safely handle — and where it has to stop.

Intake is the most straightforwardly valuable AI application in law firm marketing, because the problem it solves is real and unglamorous: people contacting a lawyer are usually in some distress, they contact several, and the first substantive response frequently wins. A system that responds immediately at any hour, collects the relevant facts and gets a qualified enquiry in front of the right lawyer addresses that directly. Two limits define the whole design, and both are non-negotiable. The system cannot give legal advice or evaluate a claim's merits, and it must not create a reasonable impression that a lawyer-client relationship has formed. It also cannot clear conflicts, which means what it says before conflicts are cleared has to be safe on the assumption that the firm may have to decline.

mechanisms

What intake automation may and may not do.

The first four are safe. The last two are the ones that create the problems.

Respond immediately

Acknowledge, set expectations, capture contact details. Pure win over an unanswered form.

Collect the facts

What happened, when, where, who. Structured and available before the first call.

Answer procedural questions

Fees, process, what to bring, how long things take. Information, not advice.

Route and schedule

To the right practice group and onto a calendar. Removes the delay that loses matters.

Assess the claim

Whether they have a case is legal advice. The system must not go here.

Imply representation

Language a distressed person could read as "we are your lawyers". The subtler and more common failure.

methodology

What the evidence shows — and what we examine.

How firms build intake safely.

Clear, early disclaimersNot a lawyer, not advice, no relationship formed. Stated where it will actually be read.
Hard refusal on merits questionsTested deliberately with the questions people actually ask, which are all merits questions.
Conflicts before substanceNothing that reads as engagement until conflicts are cleared by a human.
Retain the transcriptsWhat was said to a prospective client is a record you may need. Keep it.
what's at stake

What intake practice decides.

Revenue on one side, and an inadvertent duty on the other.

matters won or lost on response time inadvertent lawyer-client relationships duties owed to prospective clients conflicts discovered too late lawyer time spent on unqualified enquiries the firm's first impression

Duties can attach before you take the matter.

In most jurisdictions a prospective client is owed confidentiality on what they disclose, and the disqualification consequences can be real. An intake system that collects detailed facts before conflicts are cleared can create a problem the firm did not choose.

common questions

Intake — practical questions.

Can a chatbot create a lawyer-client relationship?

It should not, and the risk is that the person on the other end reasonably believes one formed. Relationship formation generally turns on the prospective client's reasonable belief rather than the firm's intention, and someone in distress reading sympathetic, confident language about their situation may reasonably form that belief. This is why the design constraint is on tone as much as content: the system must be helpful without being reassuring about the merits, and the disclaimer has to appear where it will be read rather than in a footer.

What about confidentiality of what people tell it?

Treat it as protected from the first message. Most jurisdictions impose confidentiality duties toward prospective clients on information disclosed during consultation about possible representation, whether or not the firm takes the matter — and those duties can carry disqualification consequences if the firm later acts for the other side. Practically this means intake transcripts get the same handling as client material, and the vendor arrangements behind the system have to be adequate for that, which is a question most firms have not asked of their chat provider.

How do we stop it giving advice?

Constrain it and then test it adversarially, because the default behaviour of these systems is to be helpful and helpfulness here means advice. Instruction alone is insufficient; the design should route anything resembling a merits question to a human, and someone should deliberately try to extract advice from it using the phrasings real people use — "do I have a case", "is it worth suing", "how much is this worth". Those are the first questions almost everyone asks. Re-test after any change to the system or its underlying model.

Does this work for every practice area?

It fits high-volume consumer practices best and specialised commercial work least. Personal injury, family, immigration, employment and consumer bankruptcy have enquiry volumes where response speed genuinely determines outcomes and the initial questions are reasonably standard. Complex commercial work usually arrives through relationships and referral, where an automated first response may read as a lack of seriousness. The honest test is how your best matters actually arrived — if it was a phone call from someone who knows a partner, this is not your bottleneck.

related

Related specialization areas & resources.

Build intake without building a duty.

Describe how enquiries reach you now. The Institute will help you draw the lines.

AI adoption conciergeorientation · not legal or ethics advice
Tell me how enquiries reach the firm now and how quickly they get answered. I'll help you think about what a system can safely handle — and where it has to stop.