Most firms that buy intake tools are not trying to build a system. They are trying to stop losing calls on Friday afternoons, stop retyping the same name into four places, and stop discovering three weeks later that a promising case went to the firm down the street. The gap between that problem and what gets built is where legal intake software usually dies.
What happens is predictable. A firm maps its ideal intake process, the one it wishes it ran, then configures legal intake software to enforce it. Twelve conditional fields, four approval gates, automated conflict flags, a scoring rubric for case value. Six weeks later the intake coordinator is keeping a parallel spreadsheet because the form will not let her save a half-finished record when a caller hangs up. The system is now overhead, not leverage.
What the firm is actually trying to solve
Strip the vendor language away and most intake problems reduce to four things.
Speed of first contact. A prospective client who calls three firms generally retains one of the first to call back meaningfully. If your inbound web form lands in a shared inbox that someone checks twice a day, no amount of pipeline reporting fixes that.
Not dropping people. Leads sit in voicemail, in a paralegal's notebook, in a personal email thread. Nobody knows the current count of open inquiries, so nobody can tell whether follow-up is happening.
Stopping duplicate data entry. The same facts get typed into an intake form, a conflict check, an engagement letter, the practice management system, and the calendar. Each retype is a chance to transpose a date.
Knowing what is working. Which referral sources, ad spend, or practice areas actually convert. Most firms have opinions here and very little data.
Notice that none of those require judgement about the merits of a matter. They are clerical and logistical. That distinction matters, because the systems that fail are the ones that try to automate the parts requiring an attorney's assessment.
Where overbuilding goes wrong
The most common failure is encoding case evaluation into the tool. A firm builds a scoring model that routes personal injury inquiries by reported injury severity, treatment status, and time since the incident. On paper it prioritises the pipeline. In practice a caller describes a soft tissue injury with no treatment yet, scores low, gets routed to a slow-track nurture sequence, and the firm never learns that there was a commercial vehicle and a policy limit worth pursuing. The scoring model was not wrong about the inputs. It was wrong to think the inputs were the case.
Keep the triage. Drop the pretence that the triage is an assessment. A good intake record flags what is time sensitive and routes it to a human quickly. It does not decide.
The second failure is rigid required fields. Real intake calls are messy. Callers do not know the name of the opposing party, the exact date of the incident, or whether they signed anything. If the system will not let staff save a partial record, staff will work outside the system, and your data will be worse than it was with a legal pad.
The third failure is automating conflict checks as a gate rather than a search. Running a name against your matter database and surfacing possible hits is useful clerical work. Treating a clean automated result as a conflicts clearance is not. Under the California Rules of Professional Conduct, the conflict analysis is the lawyer's, including the parts a database cannot see: the former client whose matter was handled informally, the relationship a partner has not logged, the imputation question across offices. Build the search, keep the sign-off manual, and record who signed off.
The fourth failure is treating a deadline calculator as authoritative. A tool can flag that a matter involves a government entity and that a claim presentation requirement may apply, or that a statute of limitations window is approaching. It should flag, not calculate. California deadline practice has too many tolling rules, local court variations, and service nuances to be safely automated, and a calendar entry generated by software is still an attorney's responsibility to verify.
What actually works
The systems that survive their first year share a few traits.
They start with capture and routing. Every inbound channel, phone, web form, chat, referral email, creates one record in one place within minutes. That single change usually produces more recovered revenue than anything else on the list.
They automate the clerical middle. Pulling the caller's details into a conflict search. Generating a draft engagement letter with names and dates filled in. Creating the matter in Clio or MyCase once an attorney approves. Sending a secure link for document upload rather than asking the client to email a photo of their driver's licence.
They make follow-up a default rather than a task. If no human has touched a new inquiry in two hours, someone gets told. If a prospective client has not returned a signature in four days, a reminder goes out. These are small automations and they recover a surprising amount of work that would otherwise quietly expire.
They keep the human step where judgement lives. Case evaluation, fee discussion, scope, declination letters. A tool can prepare the material and surface the record. The decision stays with a lawyer.
A worked example
A six-attorney employment firm in Los Angeles runs roughly forty inbound inquiries a month across phone, its website, and two referral sources.
The build is deliberately small. Web form and phone both create a record in Lawmatics with a fixed set of fields: name, contact, employer, one free text summary, source, and a single flag for anything time sensitive. Nothing else is required. The record is created even if the caller hangs up mid-sentence.
On creation, two things happen automatically. The system runs the caller's name and the named employer against existing matters and returns possible hits into the record as a list, clearly labelled as a search result and not a clearance. And it posts a message into the intake Slack channel with the summary and the time sensitive flag.
An attorney reviews within one business day. If the inquiry is worth a consultation, one click sends a scheduling link and a short questionnaire. If it is declined, the system produces a draft non-engagement letter from a template for the attorney to review and send, which matters more than people think, because undocumented declinations create exactly the ambiguity about whether a relationship formed that you do not want later.
When the attorney marks the matter accepted, the engagement letter draft populates with names, scope language selected from a short list, and the fee terms the attorney chose. It is never sent automatically. On signature, the matter is created in Clio with the contact, documents, and intake summary carried across, and the intake record closes.
What is not in the build: scoring, automated deadline entry, automated conflict clearance, multi-stage approval, and a reporting dashboard beyond source and conversion counts. The firm can add those later. Most of them it will not want.
Judging the options honestly
Vendors in this space fall into rough categories. Practice management platforms like Clio and MyCase include intake modules that are adequate if your needs are simple and you value having one system. Dedicated intake and marketing platforms like Lawmatics go deeper on lead nurture and campaign attribution, at the cost of another system to administer. Litigation-heavy platforms like Filevine extend further into matter management and are a heavier commitment. Beyond that sit general automation layers and custom builds, which fit when your intake crosses tools no single vendor covers.
A few questions cut through demos quickly.
- Can a user save an incomplete record, and what does the system do with it
- Where does the data physically live, who at the vendor can access it, and what are their subprocessors, because client information triggers CCPA obligations for many California firms and your duty of confidentiality under the Rules of Professional Conduct does not transfer to a vendor
- Can we export everything, in a usable format, without vendor assistance
- What happens when the process changes, and can our staff make that change without a consultant
- Does the tool ever make a decision we would be uncomfortable explaining to the State Bar of California
That last question is the useful filter. If the answer is yes anywhere, that is a place to remove automation and insert a person.
What to do first
Spend one week counting before you buy anything. Log every inbound inquiry by channel, the time of first contact, the time of first substantive response, and the outcome. Do it on paper if necessary. Most firms find the bottleneck is not the absence of software but a specific unowned handoff, usually between the first call and the attorney review.
Then build the smallest thing that fixes that one handoff, and run it for a month before adding anything. Alphovia builds this kind of intake automation for firms, and the pattern that holds is consistent: the systems that last are the ones that were allowed to stay boring.