The Complete Guide to AI Lead Scoring for Home Service Businesses

Pull 100 roofing permits filed in the Dallas-Fort Worth metro on any given day. How many are actually worth calling on right now? The honest answer is: maybe 30–40. The rest are commercial jobs, minor repairs, permits with no contact data, or properties where the work is nearly complete.

The problem with raw permit data is the noise. You can spend an entire morning calling permits and wonder why nothing is converting — not because permit leads don't work, but because you're calling the wrong ones first. AI lead scoring solves this by ranking every permit so your team always calls the highest-probability leads first.

What the score actually measures

Permitlify's lead score runs from 0 to 100 and is assigned an letter grade from A+ to F. The score reflects the probability that a given permit will result in a booked job for a home service contractor who calls within 12 hours of filing. Here are the core factors the model weighs:

• Permit type and trade match (25%). Residential roofing, HVAC replacement, and plumbing rough-in permits consistently outperform commercial or mixed-use permits. The model assigns base scores based on 18 months of conversion data segmented by permit type.
• Declared job value (20%). Higher declared job values signal higher homeowner commitment and larger average deal sizes. Permits with declared values under $2,000 are typically minor repairs. Permits with values over $10,000 are full replacements.
• Contact quality (25%). This is often the biggest differentiator. A permit with a verified mobile phone number, confirmed email, and high match confidence (we matched the owner name to property records) scores dramatically higher than one with just an address. No contact data means the lead is nearly uncallable regardless of other factors.
• Property signals (15%). Single-family residential scores higher than multi-family or commercial. Homes built 15–30 years ago are in the sweet spot for HVAC and roofing replacements — old enough to need work, but not so distressed that the owner isn't investing.
• Data freshness (15%). A permit filed in the last 6 hours scores higher than the same permit filed 3 days ago. Freshness decays exponentially — permits older than 5 days receive a significant score penalty.

Reading the score grades

A+ / A (score 90–100): Residential replacement permit with high declared value, verified mobile + email, filed within 6 hours. Call immediately. These convert at 30–40%.

B+ / B (score 75–89): Good residential permit with solid contact data but slightly older or lower declared value. Call today. Converts at 20–28%.

C+ / C (score 55–74): Decent permit but missing a key signal — either older, lower value, or contact data is partial. Call if your A/B queue is cleared. Converts at 10–18%.

D / F (score below 55): Commercial permit, very low value, no contact data, or very old. Skip unless you have a specific reason to pursue.

How to use the score in your workflow

The most effective approach we've seen from Permitlify customers is a tiered calling system. A-grade permits get called within 2 hours of the 6 AM delivery. B-grade permits get called by noon. C-grade permits go into an automated email/text sequence. D/F grade permits are ignored unless a salesperson manually reviews and decides to pursue.

This approach ensures your best callers spend their time on the highest-probability leads, not on commercial permits that were miscategorized or contacts that don't exist.

The model keeps learning

Permitlify's scoring model is retrained monthly on conversion outcomes reported by subscribers and validated against permit lifecycle data. A permit that gets "claimed" by a contractor and results in a booked job becomes a positive training signal. Permits that generate calls but no bookings help the model identify false positives.

Over time, the model improves at predicting which specific permit types, city combinations, and property profiles convert best — and your lead feed gets progressively more accurate.