Case study · Personal injury lead generation
Intake that grades itself, without the AI deciding whether a case has merit.
Personal injury lead generation for a law firm, which is not named here at its preference. Everything described below was built and is running.
The first month of paid search, measured
September 17 to October 16, 2026. Read from the ad account and the firm's lead records.
from the first ad served to the first signed retainer
per signed case, 2 cases on $4,565 of search spend
per lead, 13 leads from paid search in the first month
from a form submit to a graded lead in front of the firm
- under 15 minutes: how long the firm took to pick up each of its first two paid leads.
- $2,080: spent in one market that produced no leads in its first 13 days, so it was paused. It was relaunched on October 11. Both signed cases came from a different market.
- 495 blocked search terms and 62 landing pages built so the ads only show for, and land on, what the firm actually takes.
A month and two cases are an early result, not a monthly average, and I do not project one from them. The firm is not named at its preference.
The situation
The firm had a live website, a real caseload, and an intention to start advertising. What it did not have was a reliable path from an inquiry to a decision. Inquiries arrived by phone and by form, were triaged from memory, and left no record of why anyone had said yes or no.
Before any of the automation work, an audit of the existing site turned up several problems that would have quietly wasted the entire advertising budget.
Almost nowhere to convert
A single phone link per page, sitting near the very bottom of the document, and a homepage with no inquiry form at all. Paid traffic would have arrived and left.
Results claims with nothing behind them
Several prominent figures about past recoveries and case volume that could not be substantiated on request. In a jurisdiction with strict attorney advertising rules, that is a live risk, not a copywriting quibble. They came down before any spend started.
A repealed statute quoted as current law
A practice-area page described a serious-injury threshold using language that was no longer accurate. A prospective client could have relied on it.
The site itself was unreliable on a phone
A structured audit found a long list of layout defects across tablet and mobile, which traced back to about a dozen underlying causes once grouped properly.
None of that is glamorous, and all of it had to be fixed before automation was worth building. This is a fairly typical finding: the constraint is rarely the absence of AI.
The decision the rest depends on
The model extracts facts. Code decides what they mean.
A law firm cannot have software forming a view on whether someone has a case. That is a judgment reserved to lawyers, and a system that blurs it creates a problem far more expensive than the one it solves. So the grading engine is split in two, and the split is architectural rather than a policy written in a document.
The language model
Reads the message. Extracts facts. Stops.
What kind of incident, when it happened, where, whether there was medical treatment, who else was involved. Structured fields out of unstructured text. It is never asked whether the matter is good, strong, or worth taking.
Ordinary deterministic code
Applies the firm's own rules to those facts.
Limitation periods calculated from dates. Statutory thresholds checked. Conflicts matched against existing matters. Same facts in, same grade out, every single time, and a lawyer can read the rule and change it.
The practical consequence is that the firm can describe exactly what its software does and does not do, in a sentence, to a regulator or an insurer. The grade is a sorting mechanism over facts, not an opinion about a case. Everything else in the build sits on top of that distinction.
What was built
One intake path
Website inquiries and calls captured into a single pipeline, so nothing depends on who happened to pick up.
Response clocks with real teeth
Urgent matters carry a thirty minute clock, standard matters a day. The clock is visible, and an unmet one escalates rather than expiring quietly.
Escalation for the serious categories
Certain matters route immediately to a named human regardless of what any grade says. Some things should never wait in a queue.
Fail toward a person, always
If the model is unavailable or unsure, the inquiry goes to a human rather than being guessed at. An outage cannot silently drop work.
An append-only audit trail
Retained for ten years. What arrived, what was extracted, which rule applied, who acted, when. Written once and never edited.
Acquisition connected to outcomes
Paid search, profile listings and call tracking wired so that spend can eventually be judged against signed work rather than clicks.
The findability layer
Structured data, canonical fixes, practice-area content, and the plain-text files that AI answer engines read when someone asks them for a lawyer.
Compliance handled as scope
Attorney advertising requirements, business messaging registration, disclaimers, and sign-off from the firm's own counsel where it was warranted.
Reporting with an approval gate
Internal digest daily. Anything going to the client is drafted for review and published only on approval. Nothing goes out unreviewed.
Everything in the firm's name
Hosting, database, analytics and ad accounts all owned by the firm. If they replace me tomorrow they change the passwords and keep operating.
Recently shipped
The work did not stop at launch. These went live in September 2026, each with the date it shipped.
September 21, 2026
Grading became the primary path
New inquiries are graded on the extracted facts as they arrive, under the same rules as before: the grade describes facts, never merit, and a person still reviews the top grade.
September 21, 2026
A weekly one-page report for the firm
Reworked on September 25 to lead with results, then where the firm shows up in Google search and how it compares with competing firms. It carries no money figures, and it goes out only after review.
September 23 and 24, 2026
Checking the grader, every day
The grading was tested against 66 made-up inquiries written to the firm's own rules. Every day, a short digest lists each inquiry where the AI's grade and a person's grade disagreed, so the rules get fixed rather than argued about.
September 21 to 24, 2026
Weekly search-term reviews
Each week the searches that triggered ads are read, and the ones that should never have, such as other firms' names and off-target phrases, are blocked. 121 were added across three reviews.
September 29 and 30, 2026
Callbacks from the number the lead dialed
Every call lead now arrives as a card with one button: call back from the local number. One tap rings the staffer, then connects the caller, whose phone shows the same local number they dialed rather than a stranger's. The firm's caller name has been filed with the carriers so those calls are less likely to be flagged as spam.
Built, switched off
Missed-call text-back
A caller who hangs up gets a text back, with opt-out words in English and Spanish. It is built but stays off until the firm attaches a phone number to the messaging service.
What carries over to other engagements
Fix the leaks before you turn on the tap
Sending paid traffic to a site with one phone link and no form converts budget into nothing. The unglamorous fixes came first, and they were the highest-return work in the project.
Put the compliance boundary in the architecture
A policy saying the AI will not decide merit is a promise. A system where the model physically cannot reach the decision is a guarantee. Only one of those survives a hard question.
An audit trail is much cheaper to build than to retrofit
It went in on day one. Adding one after the fact means reconstructing history that was never recorded, which usually means it cannot be done.
Client ownership is a feature, not a concession
Building on accounts in the firm's name removed the entire question of what happens if this does not work out, which made every other decision easier to agree.
Further reading
legal intake automation that grades facts and never merit
legal intake automation should extract and structure facts for review, never score a case's merit. here's the mechanism and why the line matters.
an AI that grades facts vs one that grades merit
AI can extract facts from a file reliably. Assessing legal merit is a different task entirely, and confusing the two is how firms get burned.
how to explain an automated decision to a regulator
a regulator does not need to trust the model. they need a decision they can trace to a rule. here is how to build that trail.
Most of this started as a two week audit.
The conversion holes, the compliance problems and the build sequence all came out of mapping the workflow before writing any code.
See the service: AI for law firms and law firm intake software. Same firm, later: a second firm website, built in a day.
Questions people ask
Which firm is this?
A personal injury firm, and that is as specific as this page gets. The agreed level of disclosure does not include the firm name, the attorneys, where it practices, or the commercial terms. Those are the client's to disclose, not mine.
Is this personal injury lead generation, or is it intake?
Both halves now, and being precise about that matters. Most personal injury lead generation is sold as volume: someone sends you leads and you find out afterward how many were worth answering. This firm owns the whole path instead. Paid search ads bring the inquiry in, each one is graded on its facts against the firm's own rules, and the callback goes out from the local number the person dialed. In the first month the search ads cost $4,565, brought in 13 leads, and produced 2 signed retainers: about $351 per lead and $2,283 per signed case.
Where do the numbers come from?
From the ad account and the firm's own lead records, read on October 17, 2026, covering September 17 to October 16. Spend is Google search ad spend only. A lead is a real inquiry from the ads, by form or call, with test entries removed. Signed means a retainer the firm reported as signed. A month and two cases are an early result, not a steady-state one, and the page says so rather than projecting a monthly figure.
What does lead generation for personal injury lawyers look like when the firm owns it?
The rules that grade each inquiry are the firm's own, written in its words, and every decision is recorded where the firm can read it. An inquiry from a paid search ad is graded on its facts within seconds, a person is alerted on a response clock, and the callback comes from the local number the caller dialed. The point is that no lead the firm paid to get is lost at the last step, the callback.
Does the AI decide which cases the firm takes?
No, and the system is built specifically so it cannot. The model reads an inbound message and extracts facts, then stops. Rules the firm owns decide what those facts mean, and a person makes the actual call. Anything uncertain routes to a human rather than being resolved automatically.
How does that stay inside the attorney advertising rules?
By keeping the machine away from anything that constitutes legal judgment. Grades describe facts, never the merit or value of a claim, every A-grade lead is reviewed by a person before contact, and the whole path is inbound only. Every step writes an audit record, so the firm can show what happened and when.
Would this work for a practice that is not a law firm?
The shape carries over: extract facts with a model, decide with rules the client owns, fail toward a person, and log everything. What does not carry over is the rule set, which is specific to how a given firm or practice actually decides. That is what the audit stage is for.