Personal injury is one of the highest-volume practice areas for inbound leads — and one of the highest-noise areas. People reach out about incidents that are well outside the statute of limitations. About injuries that required no medical treatment. About accidents where fault is clearly theirs. About matters in states where you're not licensed.
Without a qualification system, your staff evaluates every one of these leads manually. That's 15–20 minutes per lead — most of which are unqualified before the third question. Multiplied across 50, 100, or 200 monthly inquiries, that's a significant portion of your intake coordinator's week spent on cases that will never be filed.
AI qualification changes this: every lead is evaluated against your exact criteria within 60 seconds of submission. Only the leads that meet your standards reach your staff. Everything else is handled automatically.
What Makes a Personal Injury Lead Qualified
Before you build any system, you need to define your qualification criteria explicitly. The most common PI qualification factors are:
- Statute of limitations: Is the incident within the applicable SOL for the state where it occurred? (Typically 2–3 years for auto accidents; varies by state and claim type)
- Medical treatment: Did the claimant receive medical treatment? (Untreated injuries are very difficult to value and typically indicate minor or non-existent injury)
- Jurisdiction: Did the incident occur in a state where your firm is licensed and practices?
- Liability: Is there a reasonably clear third-party fault? (Single-vehicle accidents where client was at fault, injuries from self-inflicted actions, etc. are typically disqualified)
- Insurance coverage: Is there a viable source of recovery? (Uninsured defendants with no assets limit recovery options)
- Severity: Does the injury severity support the economics of litigation? (Minor property damage without injury may not support attorney fees)
Your specific criteria may differ. Some PI firms handle low-severity claims at volume; others only take high-value cases with serious injuries. Define your criteria based on your business model before building the qualification logic.
Two Approaches to AI Qualification
There are two ways to use AI in lead qualification. They serve different purposes and you may want both.
Approach 1: Rule-Based Scoring (No AI Required)
For firms with clear, binary qualification criteria, you don't need an AI model. You need a scoring matrix:
- Incident within SOL → +3 points
- Medical treatment received → +2 points
- Licensed state → +2 points
- Third-party fault indicated → +2 points
- Insurance coverage present → +1 point
Scoring thresholds:
- 8–10: Qualified — route to immediate booking
- 5–7: Borderline — route to staff review within 2 hours
- 0–4: Unqualified — automated polite declination
This logic runs in n8n or Zapier in under 5 seconds per lead. It's deterministic, transparent, and easy to audit. If a lead is wrongly qualified or disqualified, you can trace exactly why.
Approach 2: LLM-Based Evaluation (For Complex Cases)
Some leads don't fit neatly into a binary matrix. The client described an accident but didn't know if there was medical treatment. The incident date is unclear. The fault question is ambiguous based on what they wrote.
For these cases, an LLM (like GPT-4o via OpenAI API) can read the full intake submission and return a structured evaluation:
- Does this incident appear to be within the statute of limitations? (Yes/No/Unclear)
- Was medical treatment mentioned? (Yes/No/Unclear)
- Did the incident occur in a licensed state based on the address provided? (Yes/No)
- Is third-party fault indicated? (Yes/No/Unclear)
- Overall qualification recommendation: (Qualified/Borderline/Unqualified)
- Reasoning: [brief explanation]
This structured output feeds into your routing logic exactly like the rule-based score. The difference: the LLM can reason about ambiguous descriptions rather than just matching fields against rules.
In n8n, this looks like: intake form webhook → OpenAI node with your qualification prompt → parse the JSON response → route based on the verdict. Total processing time: under 10 seconds per lead, 24 hours a day.
This is the kind of system we build in our personal injury intake automation service.
Building the Intake Form That Captures What You Need
AI qualification is only as good as the data it receives. Your intake form needs to collect the specific information your qualification criteria require. For PI, this means:
Date of incident: Exact date (not "about a year ago"). Use a date picker field, not a text box — unstructured dates are harder to process automatically.
Type of incident: Dropdown with options (auto accident, slip and fall, medical malpractice, workplace injury, dog bite, other). This determines which sub-set of qualification criteria apply.
State where incident occurred: Dropdown with all US states. This checks against your licensed states and determines applicable SOL.
Medical treatment: Yes/No/Not yet — "Did you receive medical treatment for your injuries?" Simple and critical.
Description of what happened: Open text field. This feeds the LLM for complex case evaluation.
Current representation: "Have you spoken with another attorney about this matter?" If yes, either they're shopping (fine) or already represented (potentially problematic).
Keep the form focused. Every additional field reduces completion rate. Ask only what you need to qualify — your attorney will gather additional details in the consultation.
The Three Routing Paths
Qualified Lead Path:
- Score meets threshold
- Immediate SMS: "Hi [Name], thank you for reaching out. Based on your inquiry, we'd like to schedule a free consultation. Here's a link to book a time: [Calendly]"
- Lead created in CRM with intake data and AI qualification score pre-populated
- Coordinator notified via Slack/email: "[Name] — qualified PI lead — consultation booked for [date]"
- 24-hour reminder sent automatically if consultation not booked within the hour
Borderline Lead Path:
- Score in middle range or AI returns "Borderline" verdict with reasoning
- Lead created in CRM with "Needs Review" tag and AI notes attached
- Coordinator receives task: "Review borderline PI lead — [Name] — within 2 hours"
- Automated email to client: "Thank you for reaching out. We're reviewing your inquiry and will be in touch within [X hours]"
Unqualified Lead Path:
- Score below threshold or AI returns "Unqualified" with clear reason (e.g., SOL expired)
- Automated email sent immediately: polite declination with reason if appropriate and referral resources
- Lead logged in CRM as "Disqualified" with reason code
- No further contact
What to Do With Disqualified Leads
Disqualified doesn't mean worthless. Some firms generate goodwill (and referrals) by directing unqualified leads to appropriate resources:
- SOL expired: "Unfortunately, the statute of limitations may have passed for your type of claim. We recommend speaking with [state bar referral service] about your options."
- Outside jurisdiction: "We're not currently licensed in [state], but we can refer you to [trusted local firm]."
- No injury: "Based on your description, your matter may be better handled as an insurance claim. We recommend contacting [state department of insurance]."
A disqualified lead who gets a helpful, specific response is far more likely to refer a qualified friend than one who receives a generic "we can't help you."
Measuring the System
Once your qualification system is live, measure these metrics monthly:
- Qualification rate: What % of submissions qualify? If it's under 20%, your marketing is off-target. If it's over 80%, your criteria may be too loose.
- Borderline conversion rate: Of leads that went to staff review, what % converted to retained clients? This tells you whether your threshold is calibrated correctly.
- Disqualified accuracy: Spot-check 10 disqualified leads per month. Were they actually unqualified? If you're wrongly disqualifying 10%+ of leads, tighten the criteria.
- Response time: Average time from form submission to first automated response. Should be under 2 minutes.
Adjust thresholds based on what the data shows, not on gut feel.
Common Mistakes
Qualifying on too few criteria: If you only check one or two factors, you let in low-quality leads that burn intake capacity. Define your full criteria set before building the system.
Letting the LLM make the final decision alone: AI qualification is a tool, not a judge. For borderline cases, always have a human review. The LLM provides a recommendation and reasoning — your experienced staff makes the call.
Not tracking why leads are disqualified: If 60% of your disqualified leads are SOL-expired, that's a marketing signal: your ads are reaching people who had their accidents too long ago. Adjust targeting.
Building a perfect system before testing it: Start with rule-based scoring and 3 criteria. Run 30 leads through it. Review the outcomes. Add complexity only where the simpler system is clearly wrong.
Get a Built AI Qualification System for Your PI Firm
If your intake team is spending more than 30 minutes per day on calls or emails with leads who turn out to be unqualified, you need an AI qualification layer in front of them. It should be screening leads before staff gets involved — not after.
We build complete AI intake and qualification systems for personal injury firms in 7 days, integrated with your existing CRM and calendar.
Book a free law firm automation audit and we'll map out your exact qualification criteria and the routing logic that fits your practice.