How to Use Roof Age and Hail History to Prioritize Inspections After a Storm

The storm came through Tuesday night. By Wednesday morning your phone has eleven voicemails, three competitors already have yard signs up two streets over, and your sales manager is standing in the office asking the only question that matters: where do we send the crews today?

Most shops answer that question badly. They knock the street closest to the office, or the one where the first call came from, or wherever the canvasser felt like parking. Then they spend the next three weeks burning gas and goodwill inspecting roofs that either weren't touched by the hail or are too new to have taken meaningful damage in the first place. The good roofs, the ones that are both old enough to fail and squarely under the worst of the core, get knocked on day nine by somebody else.

The storm doesn't decide who wins the neighborhood. The routing does. And the two variables that should drive that routing more than any others are roof age and hail history — specifically, where the damaging hail actually fell and how that lines up with roofs old enough that the impact matters.

This is a long guide because the topic deserves one. By the end you'll have a repeatable system for ranking post-storm inspections: how to read a hail swath, how to estimate roof age in a range you can trust, how to combine the two into a priority score, how to route a crew around that score, and the edge cases that quietly wreck the math. There's a worked example, several checklists, and an honest section on the tools — including ours — and where each one stops being useful.

Why "knock everything" is the most expensive mistake in storm work

The math on a brute-force sweep is brutal once you actually run it.

Say a canvasser can do 25 to 40 real door conversations in a productive day — actual conversations, not bare knocks. In a hail-restoration neighborhood, the inspection-to-sign ratio swings wildly based on whether the home was actually hit and whether the roof was a candidate to begin with. Knock a street of eight-year-old architectural shingles that caught pea-sized hail and you'll inspect a dozen roofs to find one with a marginal claim. Knock a street of 19-year-old three-tab under the storm core and the ratio flips hard in your favor.

The difference between those two streets is not effort. It's selection. And selection is exactly what gets skipped when the office is in panic mode and the instruction is "go work the storm."

There are three real costs to working the wrong streets first:

1. Time decay. Insurance carriers and most state statutes give homeowners a limited window to file a storm claim — frequently one to two years depending on the policy and state, and far less in practical terms because adjusters get harder to convince as the visible evidence weathers and as competitors muddy the neighborhood. The roofs you want most are the ones where you can document fresh, attributable damage. Every day you spend on low-probability streets is a day a high-probability roof ages out of obvious attribution or gets signed by someone else.
2. Crew morale and churn. A green canvasser who knocks 40 doors of new roofs and gets nothing quits. A canvasser who knocks 25 doors of genuinely worn, genuinely hit roofs and books four inspections stays. Roofing already fights brutal turnover; the industry's own labor data shows construction-trade churn running far above the all-industry average, and storm-chasing crews churn worse. Bad routing manufactures churn.
3. Reputation burn. Every inspection you run on a roof that wasn't damaged is a homeowner you've trained to distrust roofers. Multiply that across a neighborhood and you've salted the ground for your own follow-up.

The fix isn't to knock harder. It's to knock the right roofs first — and "right" is defined precisely by the intersection of age and hail exposure.

Run the unit economics once and it stops being a philosophical argument. Take a canvasser you're paying — base plus commission, plus the truck, plus the fuel, plus the manager's time dispatching them. Whatever your fully loaded cost per canvasser-day is, divide it by the number of quality inspections that day produces. On a well-selected street, that denominator might be five or six booked inspections. On a poorly selected street, it might be one, or zero. The cost per inspection isn't a little worse on the wrong street; it's multiples worse, because the fixed cost of putting a human on the pavement is the same either way and only the yield changes. Selection is the single biggest lever on your cost-per-inspection, and it's free — it's just a decision about where, made before anyone turns the key.

And the cost compounds in a direction most owners never tally: opportunity cost. The hour your crew spends on an eight-year-old roof that catches pea hail is an hour they didn't spend on the 20-year-old three-tab two streets over that a competitor is about to sign. In a storm market the inventory of genuinely worn, genuinely hit roofs is finite and it's shrinking by the hour as other trucks roll in. You're doing worse than wasting a low-yield hour; you're forfeiting a high-yield one. That's the real price of "go work the storm" with no list behind it.

The two variables that actually predict a productive inspection

Let's define the two inputs cleanly, because most of the bad decisions in storm work come from sloppy definitions of exactly these.

Hail history: where it fell, how big, how recent

"Did it hail here" is the wrong question. The questions that matter are:

• Did damaging-size hail fall on this specific roof? A hail report at the county airport tells you almost nothing about a roof four miles away. Hail cores are small and erratic; it's routine for one side of a subdivision to get golf-ball ice while the other side got nothing worth a ladder.
• How big was the hail, and at what angle did it come in? Damage potential climbs steeply with stone diameter. Sub-1-inch hail rarely functionally damages asphalt shingles in good condition; 1.25-inch and up starts bruising mats and fracturing granule beds; 1.75-inch and above does it reliably. Wind-driven hail at a steep angle concentrates impacts on the windward slopes, which is why two identical roofs across the street from each other can show very different damage.
• How recent and how attributable? A roof can carry old hail scarring from a storm three years ago. Fresh, attributable damage from this event is what holds up. Knowing the storm date and the roof's prior storm history is what lets you tell the difference on the ladder.

The authoritative public record for hail starts with NOAA's Storm Prediction Center storm reports and the National Centers for Environmental Information Storm Events Database. Those are real, citable, and free — but they're point reports, usually tied to a trained-spotter or public observation at a specific location and time, not a continuous map of where every stone fell. Radar-derived hail products (the kind built on Maximum Estimated Size of Hail, or MESH) fill the gaps between reports by estimating hail size across a grid from dual-polarization radar. Both have real error bars. Treat any single source as a hypothesis to confirm on the roof, not a verdict.

Roof age: a range, not a birthday

Here's the trap that sinks most age-based targeting: people go to Zillow or the county assessor, read "Year Built: 1998," and treat that as roof age. It isn't. It's the age of the house. A 1998 house may be on its second or third roof. Re-roofs are invisible to the assessor in most jurisdictions, and they're exactly the homes you need to identify.

What you actually want is an estimate of when the current roof covering went on — and the honest version of that number is a range, not a date. Aerial and street-level imagery, granule loss, the generation of shingle product visible, the condition of flashing and boots, and permit records (where they exist and are reliable) can together bracket a roof into something like "16 to 22 years old" with real confidence. Anyone selling you a single exact install date off imagery alone is overselling.

Why the range still beats a guess: asphalt shingle lifespan is well understood. Three-tab shingles typically run a 20-to-25-year rated life and often underperform it in hot, sunny, or hail-prone climates; architectural/laminate shingles run longer, commonly 25 to 30 years rated. The functional vulnerability curve matters more than the warranty number. A shingle near or past the back half of its service life has lost granule cover, the asphalt has oxidized and gone brittle, and the mat fractures far more easily on impact. The same hailstone that bounces off a four-year-old roof cracks a 20-year-old one. That's the entire thesis of age-weighted prioritization: age sets how much damage a given hail impact will produce, and hail exposure sets whether any impact happened at all. You need both.

The prioritization model: scoring a street before you knock it

Here's the system. It's deliberately simple enough to run on a clipboard if you have to, and it scales up cleanly if you have software doing the heavy lifting.

For every address in the storm footprint, you're estimating two scores and multiplying them.

Step 1: Hail exposure score (0-5)

Step 2: Roof age/vulnerability score (0-5)

Use the midpoint of your age range, adjusted for shingle type and climate.

Step 3: Priority = Hail x Age

Multiply the two. The product runs 0 to 25. That single number is your knock order.

• 20-25: Knock first. Old roof, core hail. This is the heart of the storm response.
• 12-19: Knock second wave. Strong candidates, worth a same-week inspection.
• 6-11: Knock if time allows or if a neighbor signs and you're already on the street.
• 1-5: Deprioritize. Either too new to have meaningful damage or too lightly hit.
• 0: Skip. One of the two factors is effectively zero, which makes the product zero. A brand-new roof in the core still scores low because new roofs shrug off most hail; an ancient roof outside the swath scores low because nothing hit it. The multiplication enforces the logic you actually want.

The multiply-don't-add choice is the important part. If you added the scores, a pristine new roof in the dead center of the core would still rack up a 5 and pull crews toward a low-yield inspection. Multiplying forces both conditions to be true: there has to be damaging hail and a roof old enough to show it. That's the real-world condition for a productive, attributable, claim-supporting inspection.

Optional refinements when you have the data

The base Age x Hail score is enough to run a great storm response. If you're tracking more, three modifiers sharpen it without overcomplicating the math:

• Slope-direction weighting. If you know the storm cell's track, bump addresses whose windward slopes faced the cell, because wind-driven hail concentrates impacts there. A roof "in the core" whose damageable slopes faced away may under-document.
• Prior-storm history. If an address took a documented hail event two years ago, flag it. Not to skip it — a roof can be hit twice — but so the inspector knows to separate fresh, attributable damage from old scarring on the ladder. Attribution is everything in storm work.
• Shingle-type penalty/bonus. Already baked into the age score above, but worth making explicit on the route sheet: tag three-tab roofs so the crew knows they're more brittle and likelier to show damage than a same-age laminate. It changes how thoroughly they document.

Resist the urge to add more than these. A score a crew can't compute or trust in the field is a score they'll ignore. The whole value of the model is that it's simple enough to actually drive behavior at 6 a.m.

A worked example: routing a real morning

Let's run a concrete neighborhood. Tuesday's storm dropped a hail core that clipped the north third of the Maple Ridge subdivision. You've got Wednesday and one three-person crew. Here are eight representative addresses (anonymized, but this is exactly the shape of a real list).

Read the column on the right and the day plans itself.

First: 408 Birch (25) and 412 Birch (20). Old roofs, dead in the core. These are your highest-probability, most-defensible inspections. Start the truck here.

Second: 1140 Maple (12). A laminate roof in the core, mid-life — real candidate, slightly tougher shingle, worth the ladder same day.

Third, opportunistic: 305 Cedar (10). Old roof but only edge hail; the age makes it worth a look, but don't drive across the subdivision for it — catch it if a Birch neighbor refers you or you're passing.

Deprioritize: 1155 Maple and 320 Cedar (both 4). One's too new, one's barely hit. Not today.

Skip entirely today: 77 Oak and 92 Oak (both 0). Note that 92 Oak is a 25-plus-year roof — exactly the kind of address a pure age-targeting tool would scream about. But it sits outside the swath. There's no fresh, attributable storm event to document here. It might be a great retail re-roof lead for your off-season list, but it is not a storm-claim inspection, and treating it like one is how you end up coaching a homeowner toward a claim that has no event behind it — which is both a waste of your morning and a road you don't want to be on with a carrier.

That last point deserves its own section, because it's where good intentions and bad targeting collide.

How to actually read hail data without fooling yourself

The hail score is only as good as your read of the hail data, and the data is easier to misread than people assume. Here's how the sources actually behave and how to combine them.

Point reports (SPC / NCEI). These are individual observations — a trained spotter, a public report, a news crew — logged at a specific location and time with an estimated stone size. They're authoritative in the sense that something really was observed there. They are not a map. A 2-inch report at one intersection tells you a damaging core passed near there; it does not tell you the size three blocks away, and it certainly doesn't fill in the streets where nobody happened to be standing with a ruler. Point reports are anchors, not coverage.

Radar-derived hail estimates. Dual-polarization radar products estimate hail size across a continuous grid — the most common being MESH (Maximum Estimated Size of Hail), which models the largest stone the storm was likely producing in each grid cell. This is what fills the space between point reports and gives you a swath shape. But "estimated" is doing real work in that name: radar infers hail aloft, not what reached the ground, and the estimate has genuine error bars. A grid cell flagged at 1.75 inches might have dropped 1.75 inches, or 1.25, or melted some of it on the way down. Use the grid for shape and relative intensity, and confirm magnitude on the roof.

Combining them. The reliable workflow is layered:

1. Draw the swath from the radar grid to get the shape of where damaging hail likely fell.
2. Anchor and sanity-check that shape against the point reports — if the grid says core but three nearby spotters reported nothing, distrust the grid there.
3. Confirm magnitude per roof at inspection, weighting on-roof evidence (and soft-metal denting) above any remote estimate.

No remote source is ground truth for a specific roof. The denting on a homeowner's gutters and downspouts, the bruising in your test square, the spatter marks on the AC unit's fins — those are the real evidence. The data's job is to tell you which roofs are worth climbing, in what order. The ladder's job is to tell you what's true. Keep those two jobs separate in your head and you'll neither skip a hit roof nor oversell a missed one.

A word on "the storm on each roof." Where remote data gets genuinely more useful is when it stops treating a neighborhood as one uniform blob and starts accounting for how hail and wind interacted with each individual roof — the cell's track, the wind component, which slopes faced the impacts. A flat "it hailed in this ZIP" read sends you to every door. A per-roof model of the storm narrows it to the doors where damaging impacts most plausibly landed. That's the difference between a hail map and a prioritization tool, and it's worth understanding even if you're building the score by hand: the more your hail input reflects this roof rather than this area, the more your route reflects reality.

Where this gets legally and ethically sensitive — read this part

Using age and hail data to prioritize where you inspect is squarely legitimate. You're deciding where to spend your crew's time. That's just smart operations.

Where contractors get themselves in trouble is sliding from "prioritize" into "manufacture." A few hard lines, because they protect your license and your business:

• A forecast or a hail map is not proof of damage. Radar said 1.75-inch hail fell here; that's a strong reason to inspect. It is not evidence that this roof is damaged. Only an actual roof inspection documents actual damage. Never represent storm data as proof of damage to a homeowner or an adjuster. It's the inspection findings that carry weight.
• You document conditions. The carrier decides coverage. The homeowner owns the claim. Your job on a storm inspection is to find and document the roof's condition — photos, measurements, dated storm correlation, the works — and hand the homeowner clear evidence. You are not the one who approves the claim, adjusts it, or negotiates it. In many states, a roofer acting as the homeowner's claims advocate runs straight into unlicensed-public-adjuster law. A 2024 Texas case made it explicit that even labeling yourself an insurance or claims specialist can cross that line. Document, inform, and let the homeowner and their carrier handle the claim. Check your own state's department of insurance rules; they vary.
• Say nothing about the homeowner's deductible. Offering to waive, absorb, cover, or "eat" a deductible is insurance fraud in many states, full stop. It is not a sales tactic; it's a felony in the wrong jurisdiction. Keep it out of your scripts entirely.
• No promises about approval or "free roofs." You can't promise a claim will be approved, and you shouldn't imply the roof will be free. You can promise a thorough, documented inspection. That's the honest pitch, and it's a strong one.

None of this is legal advice, and any storm-claim language your crew uses should get a once-over from counsel familiar with your state. But the operating principle is simple and keeps you clean: age and hail data tell you where to knock; the inspection tells you what's true; the carrier decides what's covered. Stay in your lane and the whole model is bulletproof.

How to estimate roof age in the field (and from the desk)

You can't score age without estimating age, so here's the practitioner version of how to bracket a roof into a range you'd defend.

Desk signals (before anyone leaves the office)

• Aerial/satellite imagery, current and historical. The single most useful desk signal is change over time. If imagery from 2011 shows a discolored, streaked roof and 2013 shows a clean, uniform one, you've bracketed the re-roof to a two-year window without ever leaving your chair. Many free and paid imagery archives carry multiple historical captures. Streaking, color uniformity, and granule sheen read surprisingly well from above.
• Shingle generation. The type of shingle visible narrows the era. Three-tab dominance suggests an older install or a budget re-roof; certain laminate profiles and ridge treatments map loosely to product generations. This is approximate, but it moves the range.
• Permit records. Where a jurisdiction reliably permits and digitizes re-roofs, a roofing permit is the closest thing to a birth certificate the covering will ever get. The catch: enforcement and digitization are wildly inconsistent. Plenty of re-roofs are never permitted, and plenty of permits never make it into a searchable database. Use it when it's there; never rely on its absence to mean anything.
• Assessor / "year built." Useful only as a ceiling check — the roof can't predate the house — and as a tiebreaker on original-roof homes. Never as the roof's age. This is the single most common mistake in age targeting; do not make it.

Ladder/ground signals (to confirm and tighten the range)

• Granule loss and mat exposure. The clearest age tell. Heavy granule loss in the gutters and bald mat patches push the estimate older.
• Shingle brittleness and curling. Oxidized, brittle, curling shingles read late-life. A shingle that cracks when you lift it is telling you its age regardless of the assessor's number.
• Flashing, boots, and sealant. Cracked pipe boots and crumbling sealant age with the roof and corroborate the covering's vintage.
• Layers. A tear-off down to one layer vs. evidence of a layover changes both the age math and the eventual scope.

The output of all this is a range — "18 to 22 years" — and that's correct, not a weakness. A range is honest, and a midpoint is plenty precise to drive a 0-5 vulnerability score. Anyone promising a single exact roof-install date from a desk is selling certainty that the data doesn't support.

A note on shingle type and climate, because they move the range

Two roofs of identical calendar age can sit in completely different places on the vulnerability curve, and a good estimator adjusts for it rather than treating age as a flat number.

Shingle type. Three-tab (the flat, single-layer, organic-or-fiberglass-mat shingle that dominated budget and older installs) generally carries a shorter rated life and oxidizes to a brittle state faster than architectural/laminate shingle. A 15-year-old three-tab is functionally older than a 15-year-old laminate. If you can read the shingle profile from imagery or the ground, let it pull your vulnerability score up or down half a step.

Climate and orientation. Heat and UV are the real killers of asphalt shingle. A south-facing slope in a hot, high-sun climate ages measurably faster than a north-facing slope in a mild one. Shingles in hail-prone, high-temperature-swing regions also tend to underperform their rated life. If you operate in such a market, your age scores should skew older than the nominal number suggests, and your local outcome log (more on that below) will confirm by how much.

Ventilation and prior repair. Poor attic ventilation cooks shingles from below and shortens life; a roof with a history of patch repairs is often a roof that's been limping for years. Both push the vulnerability estimate older. None of these are precise, but together they keep your range honest and your score calibrated to how roofs actually fail in your territory.

Building the actual route: from scores to a crew's day

A priority list isn't a route. Turning scores into an efficient crew day takes a couple more moves.

Cluster by score, then by geography

Don't send a crew zig-zagging across town hitting the single highest score, then the next, then doubling back. Instead:

1. Pull every address scoring 12+ (your knock-worthy set).
2. Cluster those geographically into tight pockets — a few streets that can be worked on foot.
3. Rank the clusters by their density of high scorers.
4. Send the crew to the densest high-score cluster first and let them work it block by block.

Density beats raw score for route efficiency. A street with six 16-scores beats a lone 25 three miles away, because the crew converts more inspections per hour of daylight. You're optimizing inspections-per-day, not the single best roof.

Seed the cluster with your warmest signal

If you have a homeowner who already called in from a high-score cluster, start the crew at that house. A booked inspection anchors the cluster, gives the canvasser a real reason to be on the street ("I'm here inspecting your neighbor's roof at 408"), and tends to cascade. Social proof on a storm street is real and it's fast.

Set a same-day re-route rule

Give the crew lead a simple rule: if a cluster is converting below some threshold — say, fewer than one inspection per ten quality conversations — after the first hour, bail to the next cluster. Storms are heterogeneous; sometimes the radar oversold a pocket, or the homes turn out newer than the imagery suggested. Don't let a crew grind a dead cluster out of stubbornness. The score is a hypothesis; the doors are the test.

Log outcomes back against the score

This is the step almost nobody does, and it's the one that compounds. For every inspection, record the score you assigned and what you found. Over a few storms you'll learn how your local hail and your local housing stock actually behave — maybe your 1.25-inch threshold should be 1.1 because your region's older three-tab is unusually brittle, or maybe your imagery vendor's captures run stale enough that you should lean more on ground signals. The model gets sharper every storm only if you feed it outcomes.

Keep the log simple enough that a tired crew lead will actually fill it out: address, assigned score, hail confirmed (Y/N and size), age confirmed (range), inspection result, booked (Y/N). Six columns. After three or four storms you'll have a few hundred rows, and the patterns jump out — which score bands convert, where your hail input runs hot or cold, which subdivisions consistently over- or under-perform their nominal age. That dataset is a quiet competitive advantage no lead vendor can sell you, because it's specific to your market and your crews.

Putting the right list in the canvasser's hand

A prioritized route does something a generic storm map never can: it makes a green canvasser sound like a veteran on the first knock. That matters more than most owners admit, because the door conversation is where the whole system either converts or dies.

Walk through what a new hire faces with no list. They knock a random door, the homeowner asks "why my house?", and the rep has nothing better than "we're working the area after the storm." That answer reads as exactly what it is — a sweep — and it triggers the homeowner's salesperson defenses. Conversion craters, the rep gets demoralized, and three weeks later they quit. Bad routing manufactures both low conversion and high churn at the same time.

Now give that same green rep a per-roof reason to be at the door:

• They know, before they knock, that this roof is in their old-roof, real-hail priority band.
• They can lead with something specific and true: "I'm inspecting your neighbor's roof at 408, and the storm Tuesday tracked right over this block." Specificity disarms.
• If you hand the homeowner a plain-language, branded summary — the estimated age range and the dated storm history for their address — the rep walks in informed instead of empty-handed, and the homeowner gets something real to look at.

The difference in the field is stark. A rep with a reason books inspections; a rep without one talks themselves out of the next door. And the rep who books inspections makes money and stays — which is how good routing quietly fixes the churn problem the industry can't seem to beat. The list isn't just a routing tool; it's a retention tool and a training tool, because it lets someone who's never climbed a ladder still knock the right door for the right reason.

Two guardrails on the door conversation, straight from the legal section above: the homeowner-facing summary documents conditions and storm history — it is not a damage verdict and not a claim. And the rep never, ever talks about the homeowner's deductible or promises an approval. The honest pitch — "the storm hit this block, your roof is at an age where that matters, let me get up there and document exactly what's going on" — is both clean and strong. It's also true, which is why it converts.

A field-ready post-storm inspection checklist

When the crew is on the ladder, the documentation is what turns a prioritized inspection into something a homeowner can actually act on. Keep it disciplined:

• Date and storm correlation. Note the storm date you're attributing to and the hail signal for the address. This ties the inspection to a specific event.
• Overview photos of every slope, oriented and labeled (front/back/left/right).
• Test square documentation. Mark a representative area (commonly a 10x10) and document impacts within it on the worst-hit slope. Photograph the marked square wide and the individual impacts close.
• Impact characterization. Photograph bruising, fractured mats, granule displacement, and the directional pattern (windward concentration). Soft metals — vents, gutters, downspouts, flashing, AC fins — often show clear, datable hail dents and are excellent corroborating evidence.
• Collateral. Window screens, fences, decking, painted surfaces, and gutters establish that hail of damaging size hit the property.
• Age corroboration. Granule loss, brittleness, layer count, boot/flashing condition — the same signals that fed your age estimate, now documented.
• Measurements and roof condition sufficient for an accurate scope.
• Hand the homeowner a clear, plain-language record of what you found. Documentation that supports a claim the homeowner chooses to file — not a claim you file, handle, or negotiate.

That last bullet keeps you on the right side of the line from the earlier section. You're the one who climbed up and documented the truth. What happens with the carrier stays between the homeowner and their insurer.

What pros get wrong (the edge cases that wreck the math)

The model above is sound, but a handful of edge cases quietly poison real-world execution. Here are the ones that bite hardest.

Treating airport hail reports as roof-level truth. A single SPC point report four miles away is not your roof's hail size. Hail cores are small and the gradient is steep. If your entire targeting is built on "the news said golf balls hit the city," you'll knock a lot of untouched roofs. Use point reports plus radar-grid estimates plus on-roof confirmation, and weight the on-roof finding heaviest.

Confusing house age with roof age. Said it before; it's worth repeating because it's the number-one error. The 1998 build with a 2015 re-roof scores low on age vulnerability, and a tool that reads the assessor will send you there anyway. Re-roofs are invisible to assessors. Estimate the covering, not the structure.

Ignoring shingle type in the age score. A 15-year-old premium laminate and a 15-year-old budget three-tab are not equally vulnerable. The three-tab is closer to the end of its functional life and far more brittle. Adjust the vulnerability score for product type, not for calendar age alone.

Forgetting wind direction. Wind-driven hail concentrates on windward slopes. A roof can be "in the core" and still show little on its leeward faces. If a crew only looks at the easy slope, they'll under-document a genuinely damaged roof. Note storm-cell movement and check the windward exposures first.

Chasing the old roof outside the swath. The 25-year roof with no storm over it is a retail lead, not a storm-claim inspection. It belongs on your off-season age-targeting list, not your post-storm crew route. Mixing the two muddies your conversion data and risks coaching a homeowner toward an event that didn't happen.

Letting old damage masquerade as new. A roof can carry scarring from a storm two years ago. Knowing the address's prior storm history is what lets your inspector distinguish fresh, attributable impacts from old ones — which matters enormously for whether the documentation holds up.

Working stale imagery. If your aerial captures are three years old, a recent re-roof won't show, and you'll mis-score age. Know how fresh your imagery is and lean harder on ground signals when it's stale.

Routing by feel instead of by score. The most expensive edge case is human. After a storm, adrenaline says "just go." A crew that knocks by instinct reverts to proximity and vibes within an hour. The score has to be in the crew's hands, on the route, or it doesn't change behavior.

Doing this at scale: where the data and the tools fit

Everything above is runnable by hand for one neighborhood. The problem is that a real storm doesn't hit one neighborhood — it smears a hail swath across dozens of subdivisions and tens of thousands of addresses, and the window to work it is short. Hand-scoring 30,000 roofs for age and hail exposure before your competitors knock is not a clipboard job.

This is the gap where data tooling earns its keep, and it's worth being precise about which tool does which job, because they're easy to confuse:

• NOAA SPC / NCEI Storm Events / radar-hail products give you the hail layer — where damaging stones fell and roughly how big. Authoritative, free, and essential, but they're hail data, not roof data. They don't know which roofs are old.
• Measurement tools (EagleView, Hover, Roofr and similar) give you precise roof geometry and material takeoffs for accurate estimating. Indispensable for scoping the job — but they measure the roof; they don't tell you which roof to knock or how old the covering is. Different category entirely.
• The assessor / Zillow gives you house age, which, as covered at length, is not roof age.

What's been missing is the layer that does the multiplication for you across an entire swath: a per-address estimate of roof age as a range fused with hail exposure modeled on that specific roof, so the priority score gets computed for every door before your crew leaves the lot.

That's the layer RoofPredict is built to be. It takes aerial imagery plus weather data per home and returns, address by address, a roof-age range (not a fake exact date), a storm/hail history for that location, and a risk score — which is exactly the Age x Hail product this whole guide is built around, computed at scale. The distinction we lean on hardest: a hail map shows you where it hailed; we model the storm on each roof, rather than only where the cell passed — hail trajectory and wind scored house by house, paired with the roof's estimated age. That pairing is the entire point of post-storm prioritization, and it's the part the public hail products and the measurement tools each leave half-done.

In practice that turns the morning-after scramble into a ranked list: the streets and addresses where an old roof and damaging hail overlap, sorted, so the crew works the highest-probability, most-defensible inspections first and skips the new roofs and the missed pockets.

Honest limits, because overselling this would be its own kind of dishonesty:

• Roof age is a range, never an exact install date. Anyone claiming otherwise from imagery is bluffing. The range is what you can trust, and it's enough to drive the score.
• A hail/storm model is odds, not proof. It tells you where damage is likely and worth inspecting. It is never proof that a given roof is damaged — only the inspection establishes that. Don't represent the model as evidence to a homeowner or adjuster.
• It ranks doors; it doesn't climb ladders. The inspection, the documentation, and the homeowner conversation are still yours. The data decides where; your crew decides what's true; the carrier decides what's covered. RoofPredict is not a lead-buying service and doesn't touch the claim — it sharpens the targeting you already do so your crew spends its daylight on the right roofs.

Used that way — as the prioritization engine in front of a disciplined inspection process — age-and-hail data stops being a vague intuition and becomes a routable, defensible system.

A 30-minute post-storm operating procedure

Put it all together and here's the actual sequence to run the morning after a storm, whether you do it by hand or with software computing the scores:

1. Define the footprint (5 min). Pull the hail swath from radar-hail products and SPC/NCEI reports. Draw the realistic core, not the whole county.
2. Score the footprint (varies). For every address, estimate hail exposure (0-5) and roof-age vulnerability (0-5), and multiply. By hand for a neighborhood; with tooling for a swath.
3. Filter to 12+ (2 min). Everything below 12 drops off today's route entirely.
4. Cluster geographically (5 min). Group the 12+ addresses into tight, walkable pockets. Rank clusters by high-score density.
5. Seed and dispatch (3 min). Start the crew at the densest cluster, anchored on any warm inbound call in it.
6. Set the re-route rule (1 min). Below one inspection per ten quality conversations in the first hour, move to the next cluster.
7. Document every inspection to the checklist above — dated, storm-correlated, claim-supporting.
8. Log score vs. outcome (end of day). Feed results back so next storm's scoring is sharper.
9. Park the off-swath old roofs on your retail age-targeting list, not the storm route.

Run that loop, storm after storm, and two things happen. Your inspections-per-day climbs because you stop knocking new roofs and missed pockets. And your documentation quality climbs because every inspection on the route is one where an old roof and real hail actually overlap — which is exactly the roof where the truth is on your side.

The storm doesn't pick the winner of the neighborhood. The roofer with the better list does. Build the list on roof age and hail history, in that disciplined Age x Hail way, and you'll be on the right ladders while everyone else is still knocking by feel.

If you'd rather not hand-score thirty thousand roofs at six in the morning, that ranked, per-roof list — age range plus storm-on-each-roof, sorted by priority — is the thing RoofPredict hands you. Book a demo and bring a street you already know; you decide whether we called it right.

FAQ

Why use roof age and hail history together instead of just one?

Because each one alone misleads you. Hail history tells you where damaging stones fell but not whether the roof was old enough to be hurt by them — a four-year-old roof shrugs off hail that cracks a twenty-year-old one. Roof age tells you which roofs are vulnerable but not whether anything actually hit them. A productive, defensible post-storm inspection requires both: damaging hail and a roof old enough to show it. Multiplying the two scores (Age x Hail) forces both conditions to be true and gives you a clean knock order.

How accurate is a roof age estimate from aerial imagery?

Accurate enough to drive prioritization, as long as it's expressed as a range rather than an exact date. Historical imagery (spotting the capture where a streaked roof became a clean one), shingle generation, condition signals, and permit records together can bracket a roof to something like '18 to 22 years' with real confidence. Anyone claiming a single exact install date from imagery alone is overselling — the honest, useful output is a range, and a range is plenty precise to score vulnerability.

Isn't the year built on Zillow or the county assessor good enough for roof age?

No, and this is the most common and most expensive mistake in age targeting. 'Year built' is the age of the house, not the roof. Re-roofs are invisible to most assessors, so a 1998 house with a 2015 re-roof reads as ancient on paper while actually carrying a nine-year-old covering. Use year built only as a ceiling check (the roof can't predate the house) and as a tiebreaker on original-roof homes — never as the roof's actual age.

Where can I find reliable hail history for a specific address?

Start with NOAA's Storm Prediction Center storm reports and the NCEI Storm Events Database for authoritative point observations, and layer in radar-derived hail-size estimates (MESH-based products) to fill the gaps between reports. Remember that point reports are tied to specific spotter locations and times, not a continuous map — hail cores are small and erratic. Treat any single source as a hypothesis to confirm on the roof, never a final verdict on a given address.

What hail size actually damages a roof?

It depends heavily on the roof's age and condition, which is exactly why you pair the variables. As a rough guide, sub-1-inch hail rarely functionally damages asphalt shingles in good shape; around 1.25 inches and up starts bruising mats and fracturing granule beds; 1.75 inches and above does it reliably. But an old, brittle, granule-depleted shingle can be damaged by smaller hail than a new one. Wind-driven hail also concentrates impacts on windward slopes, so angle matters as much as size.

How do I turn priority scores into an efficient crew route?

Don't chase the single highest score across town. Pull every address scoring 12 or above, cluster them into tight walkable pockets, rank the clusters by how densely they hold high scorers, and send the crew to the densest cluster first. Density beats raw score for daily efficiency — a street with six mid-high scores converts more inspections per hour than one lone top score three miles away. Seed the cluster with any warm inbound call, and set a re-route rule if a cluster converts poorly in the first hour.

Can I use a hail map or forecast as proof of damage for a claim?

No. A hail map, radar estimate, or forecast is a strong reason to inspect a roof — it tells you damage is likely. It is never proof that a specific roof is damaged. Only an actual inspection documents actual damage. Representing storm data as proof of damage to a homeowner or adjuster is a mistake; let the inspection findings, with photos and dated storm correlation, carry the weight.

What's the line between prioritizing inspections and improper claims involvement?

Using age and hail data to decide where to send your crew is straightforward operations. The line you don't cross is acting as the homeowner's claims advocate. You document the roof's condition and hand the homeowner clear evidence; the homeowner owns the claim and the carrier decides coverage. In many states, a roofer handling, negotiating, or adjusting a claim runs into unlicensed-public-adjuster law, and saying anything about waiving or covering a deductible can be fraud. Document, inform, stay in your lane, and have counsel review any storm-claim language for your state.

Should I knock an old roof that's outside the storm swath?

Not as part of your storm route. An old roof with no storm over it is a retail re-roof lead, not a storm-claim inspection — there's no fresh, attributable event to document. Park those addresses on your off-season age-targeting list. Mixing them into your post-storm crew route muddies your conversion data and risks pushing a homeowner toward a claim that has no event behind it.

How does software like RoofPredict fit into this versus tools like EagleView?

They solve different problems. Measurement tools like EagleView, Hover, and Roofr give you precise roof geometry for accurate estimating — they measure the roof but don't tell you which roof to knock or how old the covering is. Public hail products tell you where it hailed but nothing about roof age. RoofPredict fills the gap by computing the Age x Hail priority across an entire swath: a per-address roof-age range fused with storm history modeled on each roof, sorted so your crew works the highest-probability inspections first. It ranks doors; your crew still does the inspection and documentation, and it never touches the claim.