How to Avoid Driving to Homes That Just Got a New Roof

Every storm-season morning, a rep loads a magnet sign onto a truck door, punches an address into a phone, and burns the next nine hours doing one of the most expensive things in roofing: driving to houses. Fuel, the rep's hourly cost or draw, the windshield time that could have been spent in a conversation instead of at a stoplight — all of it gets spent the second the truck rolls. And a brutal share of it gets spent on roofs that were replaced eighteen months ago.

Nobody plans to knock a fresh roof. It happens because the list a crew works is built on the wrong signal. You bought a storm-path polygon, or you pulled a neighborhood off a hail map, or you inherited a 4,000-row spreadsheet from the last campaign, and none of those rows know whether the shingles on the actual roof are nine months old or twenty-six years old. So the rep finds out the slow way — by standing on a porch, looking up, and realizing the architectural laminates are still shedding granules from the manufacturer's wrapper. That is a wasted stop. Stack up forty of them a week across a five-person team and you are paying real money to confirm that other contractors already did the work.

This is a routing-and-targeting problem dressed up as a sales problem, and it is fixable. The roofs that just got replaced leave evidence — in aerial imagery, in permit records, in the way a street looks from the curb, and in the simple physics of how old a roof has to be before a storm can plausibly total it. Below is the full playbook for filtering those roofs out before a rep ever turns the key, the field tells for catching the ones that slip through, and the data layers that let you rank a list so the truck only rolls toward roofs that are genuinely due.

Why fresh-roof stops cost more than you think

Most owners underprice this leak because they only count the obvious line: gas. The real bill is bigger and quieter.

Start with the unit. A canvasser working a suburban storm grid covers somewhere between 60 and 110 doors in a productive day, and a meaningful slice of those doors are no-contacts, not-interested, and not-qualified before anyone even looks at the roof. When you then layer fresh roofs on top of that, you are losing more than the stop — you are losing the good stop you could have made with that same block of time. The cost of a wasted drive is the drive plus the opportunity.

Here is a way to put a number on it that you can run for your own shop.

A worked cost example

Assume a five-rep canvass team, each rep fully loaded (draw or hourly, payroll taxes, phone, fuel, vehicle wear) at $34 an hour. Assume each rep makes 80 stops a day and that 12 percent of the addresses on a typical bought-storm list turn out to be roofs replaced within the last three years — a conservative figure in a market that took a verified hail event two or three seasons back, because every contractor in the county chased the same streets you did.

The arithmetic: 5 reps times 80 stops is 400 stops a day. Twelve percent is 48 fresh-roof stops daily. At 6 minutes each that is 288 minutes — 4.8 rep-hours — vaporized every day on roofs that were never going to buy. At $34 that is about $163 a day, roughly $3,590 a month, and just over $43,000 a year. That number buys nothing. No appraisal, no signed contract, no relationship. It is pure friction.

And that is only the direct labor. It ignores the granular fact that a rep who spends the first two hours of the morning getting skunked by fresh roofs is a demoralized rep by lunch, and demoralized reps quit. Canvasser turnover is one of the most expensive numbers in a restoration shop, and few things sour a new hire faster than a list that makes them feel like the work is pointless.

The second-order cost: your reputation on the street

There is a softer cost that does not show up on a spreadsheet at all. The homeowner who just paid $18,000 for a roof six months ago and now has a third storm-chaser knocking to "inspect for damage" is not amused. They talk. They post in the neighborhood group. They tell the two neighbors who actually do have aging roofs that the guys in the white truck are clueless. Hitting fresh roofs does more than waste your time — it actively erodes the trust you are trying to build in a market, and it makes the legitimate, age-and-storm-justified knock next door land worse.

So the goal is not merely efficiency. It is precision that protects the brand.

The hierarchy of signals: from cheapest to richest

There is a clean way to think about every method for screening out fresh roofs. Each one is a signal about roof age, and signals differ in three things: how much they cost to obtain, how much of your list they can cover, and how reliable they are per address. The smart play is to stack cheap-and-broad filters first, then spend richer signals only on the addresses that survive.

Here is the full hierarchy, worst-to-best on a per-roof basis, which is also roughly cheapest-to-most-expensive to apply at scale.

Notice the trap most crews fall into: they rely almost entirely on the bottom row — the rep's eyes at the door — which is the most reliable signal but also by far the most expensive, because by the time the rep can use it, you have already paid for the entire drive. Every layer above it exists to make sure the rep's eyes only get spent on roofs worth looking at.

Let us work down the list.

Layer 1: Storm timing logic — the free filter nobody runs

The single cheapest filter costs nothing but a clear head, and most lists never get it applied. It is this: a roof cannot have been wrecked by a storm if the roof is newer than the storm.

If you are working a hail event dated, say, April 14 of two years ago, then any roof installed after April 14 of two years ago is — for that event — irrelevant. It did not exist as the current roof when the stones fell. A roof installed three months ago has no storm-damage angle from a two-year-old event, full stop. This sounds obvious written down, yet bought storm lists routinely include thousands of addresses with no age data at all, and crews knock them as if every house froze in time on the storm date.

How to actually use storm timing

The filter only works if you anchor every campaign to a specific, dated, verified event rather than a vague "we had some weather." Pull the event from an authoritative record — the NOAA Storm Events Database and the NWS Storm Prediction Center maintain hail and wind reports you can tie to a date and a rough footprint. Once you have a hard date:

1. Treat the storm date as a cutoff line.
2. Any address you can confirm was re-roofed after that date drops out of the storm campaign entirely. It may still belong on an age campaign later (a roof installed after the storm is, by definition, new and will not age into the due zone for many years), but it is dead for this storm.
3. Any address re-roofed shortly before the storm is also suspect — a roof three weeks old when the hail hit is unlikely to have been totaled, and even if it was marked, that situation usually got chased immediately. The fertile ground is roofs that were already several years into their service life when the event hit.

Storm timing is a coarse filter. By itself it will not tell you which specific roofs are new — it only tells you that if you later learn a roof is new, you can discard it without a second thought. Its value is that it makes the richer signals below decisive instead of merely suggestive.

A note on "storm physics" versus "storm path"

The other thing storm timing teaches you is that a storm path is not a storm outcome. A polygon drawn around a hail swath tells you stones fell somewhere inside the shape. It does not tell you the stones that fell on a specific roof were big enough, dense enough, or wind-driven at an angle steep enough to functionally damage that roof's particular shingle. Two roofs on the same block can take the same storm and come out completely differently depending on slope, orientation to the wind, shingle type and age, and how much life the mat had left before the event. A brittle 19-year-old three-tab will fail under hail that a two-year-old impact-rated laminate shrugs off. Path tells you where to look; it does not rank the roofs inside it. Keep that distinction in your head — it is the whole reason a smarter targeting layer beats a bigger polygon.

Layer 2: Permit records — the public paper trail of a new roof

When a roof gets replaced legally, somebody usually pulled a permit. That permit is, in many jurisdictions, a public record, and it is the closest thing to a birth certificate a roof has. If a building department shows a re-roof permit issued and finaled for an address fourteen months ago, you do not need to drive there to know the roof is new. You hard-exclude it.

Where permit data lives and how to get it

Permit availability is wildly uneven, and that is the honest catch. Some counties and cities publish a fully searchable online permit portal updated nightly; some hand you a clunky records-request form and a two-week wait; some rural jurisdictions barely enforce re-roof permitting at all, so the absence of a permit means nothing. Before you lean on permits in a given market, learn that market's specific situation.

Practical sources, roughly in order of usefulness:

• Municipal and county online permit portals. Most mid-size and larger jurisdictions now run an Accela, Tyler, or comparable portal where you can filter by permit type ("reroof," "roofing," "re-roof," "roof replacement") and date range. Learn the exact category names your jurisdiction uses; they are inconsistent.
• Open data portals. A growing number of cities publish building permits as bulk downloadable datasets on their open-data sites. When this exists it is the best case — you can pull every re-roof permit in the city for the last five years as a single file and match it against your list.
• County assessor / appraisal district notes. Some appraisal districts note major improvements. Less reliable for roofs specifically, but occasionally useful.
• Records requests. Where no portal exists, a public-records request can get you the data, but the turnaround makes it unsuitable for fast storm-season work.

Matching permits to your list without drowning in false positives

The mechanical work is an address join: take your canvass list, take the permit pull, normalize both address formats (this is where most people lose — "123 N Main St Apt 2" versus "123 North Main Street #2"), and flag every list address that has a re-roof permit dated after your relevance cutoff. Standardize to a consistent format first; the U.S. Postal Service publishes addressing standards worth conforming to so your join actually lands.

Three cautions so you do not over-trust this layer:

1. A pulled permit is not always a finaled job. Some permits get pulled and the work never happens, or happens partially. A finaled or inspected status is stronger evidence than an issued-only permit.
2. Repairs are not replacements. A permit for a small repair after a leak does not mean the whole roof is new. Read the scope where the data exposes it.
3. No permit does not mean old roof. Plenty of re-roofs happen without permits, especially in light-enforcement areas and especially right after a big storm when everybody is slammed. Permit absence is weak evidence; permit presence is strong evidence. Use it asymmetrically.

Permits are the gold standard for confirming a new roof, but their patchy coverage means they can only clean part of your list. For the rest, you need something that covers every address.

Layer 3: Aerial and satellite imagery — ranking the whole list

Permits clear the addresses where a new roof is on file. Storm timing clears the addresses that postdate the event. But you still have a huge middle — addresses with no permit record and no obvious timing exclusion — and you cannot afford to drive all of them blind. This is where overhead imagery earns its keep, because unlike permits it covers essentially every roof on the planet that has been photographed from above, which is to say nearly all of them.

The core idea: a roof's appearance from the air changes in legible ways as it ages, and by comparing imagery across time you can estimate where a given roof sits in its life. You are not getting an exact install date from a picture — you are getting a range, and a range is exactly what you need to sort the obviously-fresh from the plausibly-due.

What overhead imagery can actually tell you about roof age

From good aerial or high-resolution satellite imagery, a trained eye (or a model trained on enough examples) can read several age-correlated cues:

• Tonal uniformity and color. A brand-new asphalt roof reads as deep, even, and saturated — uniform dark gray or black, crisp lines. As shingles weather, granules shed unevenly, the surface lightens and mottles, and the even tone breaks up.
• Streaking and biological growth. Dark streaks (often algae) and lighter patches develop over years, especially on north-facing slopes in humid regions. A roof showing established streaking is not new.
• Surface texture and granule loss. High-resolution imagery can show the smoothed, thinned look of a roof well into granule loss versus the textured, full look of a fresh laminate.
• Change between captures. The most powerful cue is temporal. If imagery from three years ago shows a weathered, streaked roof and imagery from last year shows a crisp, uniform dark roof on the same footprint, the roof was replaced in between. That before-and-after flip is the clearest possible aerial fingerprint of a recent replacement.

The right mental model: a range, not a date

Be disciplined about what this signal is. Imagery gives you a probabilistic age range — "this roof reads as roughly 2 to 6 years old" or "this roof reads as 18-plus years old and weathered." It does not give you "installed on March 9." Anyone selling you a precise install date off a single satellite frame is overselling. The honest, useful output is a band wide enough to be defensible and narrow enough to be actionable. A roof that reads 0 to 3 years old comes off the storm list. A roof that reads 17 to 24 years old goes to the top of it. The middle gets ranked by everything else you know.

This is also why imagery pairs so well with the cheaper layers. Imagery says "this roof looks fresh"; a finaled permit from two years ago confirms it; storm timing explains it. Each layer makes the others more decisive.

Limits you should respect

Imagery is not magic, and a practitioner who pretends otherwise gets burned:

• Capture recency varies. If the latest available image of a neighborhood is three years old, a roof replaced last spring still looks like its old self from above. Always know the date of the imagery you are reading.
• Tree cover and angle. Heavy canopy, steep oblique angles, and shadow can obscure enough of a roof to make the read unreliable. Flag low-confidence roofs rather than guessing.
• Material confounds. Some premium or designer shingles, and some tile and metal, age differently and can read "newer" or "older" than they are. Imagery age estimation is tuned mostly to asphalt, which is fine because asphalt is most of the residential market, but keep the exception in mind.
• Resolution floors. Below a certain resolution you lose the texture cues and are left with only gross color and change detection.

Given those limits, the correct use of imagery is to rank and triage the whole list, not to make irreversible exclusions on thin evidence. Hard-exclude on permits; rank on imagery; verify the ambiguous middle with the next layer.

Layer 4: Street-level imagery — the cheap spot-check

For the addresses where the overhead read is ambiguous — the low-confidence flags from Layer 3 — a quick street-level look often settles it without a drive. Street-level imagery captured from public roads shows the roof from the side and front, which catches things the overhead misses: a freshly installed drip edge gleaming, brand-new flashing, a dumpster or material pallet in the driveway in a recent capture, crisp new ridge caps.

The workflow is simple. For each ambiguous address, pull the most recent available street-level capture and look for fresh-install tells. Note the capture date — same caveat as overhead, an old street capture cannot show a recent roof. Used as a spot-check on the few hundred genuinely ambiguous rows rather than the whole list, it is fast and keeps a lot of trucks parked.

Do not try to street-view your entire list manually; that is slower than just driving for many crews. Reserve it for the ambiguous middle that the cheaper, scalable layers could not resolve.

Layer 5: The curbside field tells — your last line of defense

Every filter above is probabilistic. Some fresh roofs will slip through — bad imagery dates, missing permits, a replacement that happened last month. So your reps still need to be trained to recognize a new roof from the curb in the first ten seconds, before they invest the walk, the knock, and the pitch. Treat this as the safety net, not the primary strategy, because it is the most expensive signal you have. But a well-trained eye saves the pitch even when it cannot save the drive.

Here is the field-tell checklist worth drilling into every canvasser.

The ten-second new-roof scan

• Granule color and sheen. A roof under roughly two years old often still has a slightly oily sheen and an unnaturally even, deep color. No fading, no streaking, no mottling.
• The fresh-cut lines. New architectural shingles have crisp, sharp shadow lines and dimensional definition that softens with age. Crisp definition plus even color is a strong new-roof signal.
• Bright, unstained flashing and drip edge. New metal at the eaves, valleys, and penetrations is bright and unweathered. Old roofs have dulled, sometimes rust-streaked metal even when the field shingles look okay.
• New pipe boots and vents. Replacement jobs almost always include new boots and vents. Black, supple, uncracked boots on an otherwise mature-looking street are a tell.
• Clean, full ridge caps. Fresh ridge and hip caps sit full and uniform; aged ones thin and curl.
• The yard evidence. A nail-magnet pass mark in the grass, a few stray shingle bits in the flowerbed, a faint rectangle in the lawn where the dumpster sat, a yard sign from a competitor — all recent-job residue.
• The mismatched neighbor pattern. On a street where a storm came through, you often see clusters of obviously-new roofs (the ones already chased) next to obviously-aged ones (the holdouts and the missed). Reading the block, rather than only the house, speeds the rep up.

Train it as a go/no-go, not a judgment call

The mistake crews make is treating the new-roof read as a fuzzy gut feeling. Make it a hard go/no-go: if two or more of the fresh-roof tells are present, the rep marks it new, logs it, and moves on without knocking. Logging matters — every fresh roof a rep catches at the curb should flow back into your data so the address is permanently flagged and never re-served to another rep next campaign. A field catch that does not update the list will just get re-driven by the next person.

Putting the layers together: a screening pipeline

Individually each layer is partial. Stacked in the right order they compound into a list where the truck almost never rolls toward a fresh roof. Here is the end-to-end pipeline a disciplined shop runs before a single rep is dispatched.

The seven-step pre-dispatch workflow

1. Anchor to a dated event (or an age thesis). For storm work, fix the verified storm date and rough footprint from authoritative records. For non-storm "aging-out" work, define the age band you are targeting (for example, asphalt roofs reading 18 years and older). Everything downstream filters against this anchor.
2. Apply storm-timing exclusion. Drop any address you can already tie to a post-event replacement. Cheap, instant, and it shrinks the list before you spend money on it.
3. Join permit records and hard-exclude. Pull re-roof permits for the jurisdiction, normalize addresses, and remove every address with a finaled re-roof permit after your relevance cutoff. These are confirmed dead.
4. Score the survivors on imagery age. Run the remaining addresses against overhead imagery to assign each a roof-age range. Drop the clearly-fresh band (for example, anything reading 0 to 3 years). Rank the rest oldest-first.
5. Spot-check the ambiguous middle with street-level imagery. For low-confidence imagery reads only, do a quick street-level pass and resolve them up or down.
6. Sequence the route by score, not by geography alone. Build the day's route to hit the highest-age, highest-storm-likelihood roofs first, while still respecting drive efficiency. A perfectly tight loop full of mediocre roofs loses to a slightly looser loop full of due roofs.
7. Close the loop with field logging. Whatever the rep learns at the curb — confirmed new, confirmed old, already-signed-with-a-competitor — flows back into the master list so the next campaign starts smarter. The list should get better every time a truck rolls, not start from zero.

A before-and-after of the same list

Picture a 4,000-address bought storm list run through this pipeline.

The team now spends its best hours on roughly 900 high-probability roofs instead of spraying 4,000. The 1,840 deprioritized rows are not thrown away — they are worked later or on slower days — but the fresh roofs are genuinely gone. That is the difference between a list and a ranked list.

Where RoofPredict fits

The pipeline above is real and you can assemble it by hand: pull permits yourself, eyeball imagery, build the joins in a spreadsheet. Plenty of sharp shops do exactly that. The reason a dedicated layer exists is that the imagery-age step and the per-roof storm step are the two pieces that are genuinely hard to do well at scale and by hand, and they are the two that most directly determine whether you drive to a fresh roof.

RoofPredict is built to do those two jobs across a whole list at once. It reads aerial imagery to estimate a roof-age range for each address — a band, not a fake exact date — so the clearly-fresh roofs sort straight to the bottom and the aging-out roofs sort to the top. And it models storm exposure per roof rather than per polygon: instead of telling you a hail swath touched the neighborhood, it weighs the storm against each individual roof's age and exposure so you get odds that a given roof was plausibly affected, not a blanket assumption that every house under the swath is fair game. The output is a ranked, enriched version of the list you already work — your own CRM or mailing list, with roof-age and storm signal attached to each row — so a rep's day is sequenced toward roofs that are actually due.

What it does not do, and should not be sold as doing: it does not promise that a high-ranked roof is damaged, and it does not replace the inspection. Roof age is a range, not a certificate. A storm score is odds, not proof. The roof still has to be looked at, documented, and measured by a human before anyone writes a number. What the data buys you is that the human looks at the right roofs — the ones old enough and exposed enough to be worth the drive — and stops looking at the ones that just got wrapped in new laminate last spring. It tightens the front of the funnel; it does not pretend to be the whole funnel.

If you already run permits and imagery by hand and it works for you, keep doing it. If the imagery-age and per-roof-storm steps are where your hand-built process falls down — and for most shops they are — that is the specific gap this fills.

The documentation and estimate side: doing it right when the roof IS due

Filtering out fresh roofs is half the job. The other half is what happens when a rep reaches a genuinely aged, genuinely storm-exposed roof and the homeowner is interested. This is where a lot of restoration shops wander into trouble, so it is worth being precise about the lane you are allowed to operate in — because doing this part cleanly is also what makes your targeting pay off instead of blowing up.

A roofing contractor's legitimate role on a storm-damaged roof is to inspect it, document the condition thoroughly, and prepare an accurate repair estimate for the work the contractor would perform. You photograph the damage, you measure the roof, you write a scope aligned to standard estimating practice, and you hand that documentation and estimate to the homeowner. The homeowner is the one who files with their insurer. The insurer is the one who decides coverage. You are the expert on the roof and the repair; you are not the homeowner's representative against their insurance company.

The do-not-say list — keep your team on the legal side

The line between legitimate contracting and unlicensed public adjusting is sharp, and crossing it is a real liability. Train every rep on what they may not say or do, because an enthusiastic canvasser will cross it without realizing. A contractor may not, for a fee:

• Negotiate, adjust, or "handle" the claim on the homeowner's behalf. You document your scope; you do not work the claim.
• Interpret the homeowner's policy or coverage. "Your policy covers this" is not your statement to make. You speak to the roof, not the contract.
• Promise a specific payout, approval, or outcome. "We'll get this approved" and "the carrier will pay X" are off-limits.
• Promise the deductible is waived, absorbed, eaten, or gone. The deductible is the homeowner's responsibility; absorbing it is fraud in most states and an unfair-practice violation besides.
• Advertise or imply a "free roof." It is not free; somebody pays the deductible, and the carrier pays only what it covers.
• Represent the homeowner against the insurer. That is public adjusting, which is licensed in most states, and doing it without a license is illegal.

What you may do is the meat of the job: inspect thoroughly, take complete photo documentation, write an accurate estimate aligned to standard estimating line items, and state facts about your scope to the carrier when appropriate. You are the roof expert. Stay on the roof-expert side of the line and you are both compliant and more credible — homeowners trust the contractor who explains the documentation honestly far more than the one promising a free roof.

A clean documentation workflow

When a targeted roof checks out as aged and storm-exposed and the homeowner wants to move, run a documentation process tight enough to stand on its own:

1. Full photo set, organized by slope and component. Overview shots of each elevation, then close-ups of damage with a reference for scale, then every penetration, valley, edge, and accessory. Mark hail strikes and wind damage clearly. Date and geotag where your tools allow.
2. Accurate measurement. Pull a measured diagram or measure on-site so squares, ridges, hips, valleys, and penetrations are correct. The estimate is only as good as the measurement under it.
3. An estimate aligned to standard line items. Write the scope the way the industry writes scope — recognized line items, correct quantities, regional pricing. An estimate that mirrors standard estimating practice is one everyone can read and check.
4. A plain-language summary for the homeowner. Explain what you found and what your estimate covers, in facts, without interpreting their policy or promising an outcome. Hand them the documentation. They file; the insurer decides.
5. Keep your own records. Retain the photos, measurements, and estimate. Good documentation protects you on warranty, on disputes, and on your own workmanship later.

This is the honest, durable way to run storm restoration, and it is exactly the work your targeting is supposed to feed. Precise targeting means your documentation team spends its time on roofs that justify the documentation — not driving across town to a roof that turns out to be a year old.

Common mistakes that put trucks back in front of fresh roofs

Even shops that know better backslide. The failure modes are predictable.

Re-buying the same stale list every season

The most common one: a shop buys a storm list, works it, learns a thousand things about those addresses in the field — and then next campaign buys a brand-new list from the same vendor and throws away everything it learned. The fresh roofs it confirmed by hand last spring are right back in the new list, unflagged, ready to be re-driven. If your field intelligence does not persist in a master record you control, you will pay to rediscover the same fresh roofs forever. Own your list; enrich it; never start from zero.

Trusting a storm polygon as if it were a damage report

A hail swath is a starting point, not an answer. Treating every roof under the polygon as a qualified lead is how you end up knocking the two-year-old impact-rated roofs alongside the twenty-year-old three-tabs. Rank inside the polygon by roof age and per-roof exposure; do not work it flat.

Using imagery dates without checking them

A crew pulls up overhead imagery, sees an old-looking roof, and dispatches — not noticing the imagery is four years stale and the roof was replaced two years ago. Always read the capture date alongside the roof. A stale image is worse than no image because it gives false confidence.

Letting reps freelance the new-roof call without logging

Reps who eyeball a fresh roof and move on are doing the right thing in the moment, but if that judgment never flows back to the list, the next rep drives there again. Make field logging mandatory and make it trivially easy, or the field intelligence evaporates.

Over-trusting permits in light-enforcement areas

The flip side of relying on permits: in a market where re-roofs often happen without permits, a clean permit record means almost nothing, and a shop that hard-excludes only on permits will leave plenty of fresh roofs in the list. Match the layer to the market — lean on permits where coverage is good, lean on imagery where it is not.

Optimizing the route for tightness over quality

A dispatcher who builds the prettiest, tightest geographic loop while ignoring roof quality will produce an efficient march past mediocre roofs. Drive efficiency matters, but it is the tiebreaker among good roofs, not the primary sort. Rank by roof quality first, then tighten the route within the high-quality set.

A field-ready checklist

Pin this where the people building lists and routes can see it.

Before you buy or build a list

• Is there a specific, dated, authoritative storm event anchoring this campaign? (Or a clear age thesis if it is non-storm work.)
• Do you own the master list, or are you renting someone else's that you cannot enrich?

Before you dispatch

• Storm-timing exclusion applied — post-event replacements dropped?
• Permit records joined and finaled re-roofs hard-excluded (where coverage is good)?
• Imagery age range scored for every survivor, fresh band dropped?
• Imagery capture dates checked for staleness?
• Ambiguous middle spot-checked with street-level imagery?
• Route sequenced by roof quality first, drive efficiency second?

At the door

• Reps trained on the ten-second new-roof scan?
• Two-or-more-tells go/no-go rule in place?
• Every field finding logged back to the master list?

When a roof qualifies

• Documentation workflow run: full photos, accurate measurement, standard-line-item estimate?
• Reps trained on the do-not-say list (no claim handling, no coverage interpretation, no payout/deductible/free-roof promises)?
• Homeowner files; insurer decides; you stay the roof expert?

The bottom line

Driving to homes that just got a new roof is not a sales problem you fix with a better pitch. It is a targeting problem you fix before anyone leaves the lot. The roofs that were replaced last year leave a paper trail in permits, a visible flip in aerial imagery, a tell at the curb, and a logical impossibility against the storm date — and every one of those signals is cheaper to read than the drive is to make. Stack the cheap, broad filters first, spend the rich signals on the survivors, and let your reps' eyes be the last line of defense instead of the first.

Do that and the truck rolls toward roofs that are genuinely due — old enough, exposed enough, worth the documentation and the estimate — while the fresh roofs stay off the route where they belong. That is where the wasted forty-three thousand dollars a year goes back into the business: not into more windshield time, but into more conversations on roofs that can actually become work.

RoofPredict exists to do the two hardest parts of that filtering at scale — the per-roof age range and the per-roof storm exposure — and to hand them back to you attached to your own list. It will not tell you a roof is damaged, and it will not write your estimate. It will tell you which roofs are old enough and exposed enough to be worth the drive, so the only roofs your crew confirms in person are the ones it should have driven to in the first place. See how it ranks your list at roofpredict.com.

FAQ

How can I tell if a roof is new before driving to the address?

Stack signals you can read from a desk. Storm-timing logic drops any roof replaced after the storm you are working. Public re-roof permits confirm replacements with near-certainty where the jurisdiction publishes them. Aerial imagery gives every address a roof-age range based on color, streaking, and change between captures, so clearly-fresh roofs sort to the bottom. Street-level imagery spot-checks the ambiguous ones. By the time a rep reaches the curb, only roofs that survived all of those filters should be on the route.

Can aerial or satellite imagery tell me the exact age of a roof?

No, and be wary of anyone who claims it can. Imagery gives a probabilistic age range, not an install date. A roof might read as roughly 2 to 6 years old, or 18-plus years and weathered. That range is exactly what you need to separate clearly-fresh roofs from aging-out ones and to rank a list, but it is a band, not a certificate. Always also check the imagery capture date, because a stale image cannot show a recent replacement.

Are roofing permit records reliable for finding recently replaced roofs?

They are the strongest single confirmation where they exist, and weak evidence where they do not. A finaled re-roof permit dated after your storm cutoff is near-proof the roof is new, so you can hard-exclude it. But coverage is uneven: some jurisdictions publish searchable permit portals, others require slow records requests, and in light-enforcement areas many re-roofs happen with no permit at all. Use permit presence as strong evidence and permit absence as almost no evidence.

What does a fresh roof look like from the curb?

Look for an even, deep, slightly oily color with no fading, streaking, or mottling; crisp dimensional shadow lines on architectural shingles; bright, unstained drip edge and flashing; new black supple pipe boots and vents; full uniform ridge caps; and yard residue like a nail-magnet pass mark, stray shingle bits, or a dumpster rectangle in the lawn. Train reps on a two-or-more-tells go/no-go rule so the call is a hard decision, not a gut feeling, and log every catch back to the list.

Why do bought storm lists send crews to so many new roofs?

Because a purchased list is usually built on a storm path or a neighborhood polygon, and a polygon does not know the age of any specific roof. After a verified hail event, every contractor in the county chases the same streets, so a year or two later a large share of those roofs are already replaced. Unless the list is filtered by roof age and confirmed replacements, crews re-drive roofs other contractors already finished. The fix is ranking the list by per-roof age and storm exposure before dispatch.

How much does driving to newly replaced roofs actually cost?

More than fuel. For a five-rep team at $34 loaded per hour making 80 stops a day, if 12 percent of addresses are freshly replaced and each wasted stop burns six minutes, that is roughly 4.8 rep-hours a day, about $163 daily, near $3,590 a month, and just over $43,000 a year on roofs that were never going to buy. That excludes the morale hit on reps who get skunked and the reputation damage from knocking homeowners who just paid for a roof.

Does RoofPredict guarantee a roof is damaged or needs replacing?

No. It estimates a roof-age range per address and models storm exposure per individual roof rather than per polygon, then attaches those signals to your own list so the roofs most likely to be due sort to the top. Age is a range, not a certificate, and a storm score is odds, not proof. The roof still has to be inspected, documented, and measured by a person. What the data buys you is that your crew looks at the right roofs and stops driving to ones that were just replaced.

Can a roofing contractor handle the homeowner's insurance claim for them?

No. A contractor may inspect the roof, document damage thoroughly, prepare an accurate repair estimate aligned to standard line items, and state facts about that scope to the carrier. A contractor may not negotiate or handle the claim, interpret the homeowner's policy or coverage, promise a specific payout or approval, promise the deductible is waived or absorbed, advertise a free roof, or represent the homeowner against the insurer. Those last items are unlicensed public adjusting in most states. The homeowner files; the insurer decides coverage.

What is the right way to document a storm-damaged roof for the homeowner?

Run a tight, self-standing process: a full photo set organized by slope and component with damage marked and a scale reference; an accurate measurement of squares, ridges, hips, valleys, and penetrations; a repair estimate written with recognized line items and correct quantities; and a plain-language summary of what you found and what your estimate covers, stated as facts without interpreting policy or promising an outcome. Hand the homeowner the documentation. They file with their insurer, who decides coverage. Keep your own copies.

Should I delete deprioritized addresses from my list?

No. Hard-exclude only confirmed fresh roofs and confirmed dead addresses. Roofs that simply rank lower today should be deprioritized, not deleted, then worked on slower days or in later campaigns as they age. The bigger principle is to own and persistently enrich one master list rather than re-buying a stale list each season, so every field finding and data update makes the list smarter instead of starting from zero every campaign.