Methodology & Trust
How Permit-Backed Roof Age Data Works
By Jesse DeLuca — Founder, ReroofGen
Permit-backed roof age works because replacing a roof almost always requires a building permit, and that permit's issue or finalization date is a dated public record filed by address. By aggregating county and municipal permit records — and falling back to property records when no permit exists — you get a real install date instead of a model estimate. The honest limit: coverage varies by jurisdiction, records can be missing, and when no re-roof permit is found the home's build year is only an upper bound on the roof's age.
Key takeaways
- A re-roof almost always requires a permit, and the permit's issue or finalization date is a dated public record tied to the property address.
- Aggregating county plus municipal permit records produces a real install date — a recorded event — rather than a statistical model's guess.
- When no re-roof permit exists, the home's build year becomes an upper bound: the roof is at most that old, possibly newer from unpermitted work.
- Coverage is uneven — permit records vary by jurisdiction, and some are incomplete, mislabeled, or simply missing.
- ReroofGen treats the permit as the primary signal and property records as the fallback, so you verify the roof story before you spend time on outreach.
Roof age is the number every roofing decision turns on — and most tools give you a guess dressed up as a fact. A statistical model looks at a home's build year, maybe some satellite imagery, and infers a likely age. That's useful as a starting point, but it isn't evidence. Permit-backed roof age is different: it's built on a dated public record of the actual re-roof.
Here's the chain of logic, end to end. Replacing a roof almost always requires a building permit. That permit is filed by address and carries a date — when it was issued, and often when the work was finalized and inspected. That date is the single most reliable signal of when the roof was last replaced, because it records something that happened rather than something a model predicts. This page explains how that record becomes a verified install date, and — just as importantly — where it falls short.
The permit is the primary signal
Start from the event itself. When a roofing contractor tears off and replaces a roof, the jurisdiction almost always requires a permit before the work begins. The building department records that permit against the property address, assigns it a type — "reroof," "roof replacement," "roofing" — and stamps it with dates: the issue date, and frequently a finalized or inspection-passed date once the job clears.
That finalization or issue date is the anchor. It's a government-recorded, address-tied event, which is why it beats any inference drawn from build year or imagery. The most recent re-roof permit on a property tells you, with real confidence, when that roof went on. Everything else in the methodology exists to find that record, or to be honest when it isn't there.
Why aggregating county and municipal records matters
Permits aren't issued from one national database — they live in county and city building departments, each with its own portal, formats, labels, and history. A single source almost always has holes. A property inside city limits might have its re-roof on file with the municipality but not the county, or vice versa. Some departments publish decades of clean searchable history; others went digital recently and have thin back-records.
Aggregating across both layers — county and the municipalities inside it — recovers permits a single search would miss, and reconciles them against one address so you don't double-count or pick the wrong one. The goal is one clear answer per property: the most recent verified re-roof date, with the source it came from so the claim is auditable rather than a black box.
The fallback: build year as an upper bound
No methodology finds a permit on every home, and pretending otherwise is where roof-age data loses trust. When there's no re-roof permit on file, the most defensible signal is the property record's build year.
Treat that build year as an upper bound, not an install date. The roof is at most that old — it could be the original roof, or it could have been replaced without a permit, or replaced under a record that never made it online. Stating it as a ceiling keeps the data honest: you know the roof can't be newer than the house, and you know you don't yet have proof it was redone.
- Re-roof permit found — verified install date, the strongest signal
- No permit, known build year — roof is at most that old (upper bound)
- Permit present but mislabeled or undated — flag for review, don't overstate
- Conflicting county vs. municipal records — reconcile to one address before trusting either
Being honest about the limits
Permit-backed roof age is the best widely-available signal, but it isn't perfect, and the limitations are worth naming plainly. Coverage varies by jurisdiction — some building departments simply have better records than others. Records can be missing, incomplete, or mislabeled, so the absence of a permit never proves a roof is original; it only means the record isn't there to find. And small repairs or unpermitted work won't show up at all.
The right posture is to lead with the permit when it exists, fall back to the build-year ceiling when it doesn't, and never dress a fallback up as a verified date. That's what lets a roofer trust the number enough to act on it.
What this means for working roofs
For a roofing company, the practical payoff is simple: you can verify the roof story before you spend a dollar on outreach. A verified re-roof from two years ago tells you to skip that address. A home at the end of an asphalt roof's typical 15-to-20-year window with no recent permit tells you it's worth a knock. The data turns a list of addresses into a ranked picture of which roofs are actually worth working.
That's the entire point of doing this with recorded permits instead of estimates — to target the right properties and close more work with less waste, on evidence you can stand behind rather than a model you can't explain.
Frequently asked questions
Why is a permit date more reliable than a roof age estimate?
A permit date is a recorded event — a government building department logged a re-roof at a specific address on a specific date. An estimate is a model inferring age from build year, satellite imagery, or material wear. The permit tells you what actually happened; the estimate tells you what is statistically likely.
What happens when there's no permit on file?
With no re-roof permit, the most defensible signal is the home's build year, which acts as an upper bound: the current roof is at most that old, and could be newer if work was done without a permit or the record never made it online. It's an honest ceiling, not a precise install date.
Does every roof replacement get a permit?
Most jurisdictions require a permit for a full re-roof, so the large majority are recorded. But not all work is permitted — small repairs, some rural areas, and unpermitted jobs can slip through. That's why missing permits don't prove a roof is original; they just mean the record isn't there.
Why does coverage vary by jurisdiction?
Building permits are issued at the county or city level, and every department runs its own system, formats, and online portal. Some publish clean searchable records going back decades; others have gaps, inconsistent labels, or limited digital history. Aggregating across both county and municipal sources fills more of those gaps than any single source.
See how ReroofGen surfaces roof age across an entire territory