How to Find the Worst-Hit Streets After a Storm (Field-Tested Workflow)

Two crews can chase the same storm and post wildly different numbers. One signs eleven jobs in a week off four streets. The other spends six days knocking a whole zip code and limps home with two repairs and a stack of "we're fine, thanks." The storm didn't play favorites. The map did.

Finding the worst-hit streets is the single highest-leverage skill in storm restoration, and almost nobody teaches it well. Most contractors treat a hail event like a starting gun: the alert pops, they load the trucks, they drive toward the dot on the radar map, and they start knocking wherever they happen to park. That works often enough to be dangerous, because it feels like the storm did the work when really you just got lucky with where you stopped.

The difference between a good storm and a forgettable one is rarely the size of the hail. It's how fast and how precisely you can collapse a 40-square-mile warning polygon down to the dozen blocks that actually got cored, and then how disciplined you are about working those blocks before the other six companies in town figure out the same thing. What follows is the workflow professional storm chasers and mature restoration shops actually run, including the parts that go wrong, the signals that lie to you, and the unglamorous fieldcraft that separates a high-close territory from a wasted tank of gas.

Why "drive toward the radar dot" fails

Radar shows you where precipitation and hail cores were aloft. It does not show you where damaging hail reached the ground, how big the stones were when they landed, or which direction they were driven. Those gaps are exactly where money is won and lost.

Hail aloft is not hail on the ground. A radar core can light up bright red at 18,000 feet and melt to pea-sized slush before it lands, especially in warm-season storms with a high freezing level. Meanwhile a smaller-looking signature can drop dense, wind-driven stones that wreck north- and west-facing slopes. The radar is a clue, not a verdict.

Then there's the polygon problem. A severe thunderstorm warning or an SPC outlook covers enormous ground to keep the public safe. If you canvass the whole warned area uniformly, you are spreading a finite crew across territory that is mostly undamaged. Storm damage is not smooth. It comes in a swath, often a long narrow ribbon, and inside that ribbon there are hot cells where the worst stones fell. Your job is to find the ribbon, then find the cells.

The third trap is timing. The first 72 hours after a storm are the richest window you'll get. Damage is fresh and visible, homeowners are rattled and receptive, and the field hasn't been picked over yet. Every hour you spend knocking a street the storm skipped is an hour a competitor spends on the street it cored. Precision early is worth more than effort late.

The mental model: swaths, cells, and aspect

Before any tools, get the picture right in your head. Three concepts do most of the work.

The swath. Hail falls along the storm's path as it moves, producing an elongated footprint, often many miles long and sometimes only a few hundred yards to a couple of miles wide. Picture a paint roller dragged across the map. The long axis follows the storm motion. Your highest-damage streets cluster along the spine of that roller mark, not the edges.

The cells inside the swath. Damage is not uniform along the swath. As the updraft pulses, the storm dumps its biggest stones in bursts. You get hot spots, gaps, then another hot spot. This is why one subdivision gets shredded and the identical subdivision a mile up the road barely needs a repair. Inside a swath you're hunting for these intensification points.

Aspect and wind drive. Hail almost never falls straight down. It rides the storm's wind, so it strikes some roof faces hard and shelters others. If stones were driven from the west-northwest, the west and north slopes take the beating and the south and east slopes can look nearly clean. This matters enormously for both finding damage and documenting it, because an inspector who only walks the leeward slopes will wrongly call a hammered house undamaged. Aspect also tells you which side of a street to favor: on an east-west road, the houses presenting their broad face to the wind-driven side get hit hardest.

Hold those three ideas together and the goal becomes concrete. You are looking for the spine of the swath, the intensification cells along it, and the slope aspect that took the drive. Everything below is about pinning those three down quickly and confirming them with your own boots.

How a swath actually forms, and why width matters

It helps to understand the physics a little, because it changes how you read a map. A supercell or a strong multicell carries a tilted updraft. Hail embryos cycle up and down inside that updraft, accreting ice on each pass, until they get heavy enough to fall out, usually on the storm's forward-right flank. Because the parent storm is moving the whole time, the fallout traces a line on the ground rather than a point. That line is the swath.

The width of the swath tells you something useful. A narrow swath, a few hundred yards to half a mile, usually means a tight, intense core, the streets under it can be brutally damaged but the footprint is small and easy to miss if your map work is sloppy. A wide swath, two or three miles across, usually means a larger or slower storm, more total damaged roofs but often a gentler gradient from spine to edge. Either way the spine is where the energy concentrated, and the gradient off the spine tells you how fast damage falls off as you move to side streets.

This is also why two storms with the same peak stone size produce very different territories. A fast-moving storm smears a long thin swath across many neighborhoods; a slow-moving or training storm can park over a few subdivisions and pound them repeatedly. Pull the storm motion, and you'll know which kind of territory you're about to work before you load the trucks.

Reading the gradient instead of a single point

The rookie habit is to find the one biggest hail report and drive straight to it. The professional habit is to read the gradient, the way intensity climbs toward the spine and falls toward the edges. A single two-inch report is a data point; a run of reports climbing from one inch to 1.75 to two and back down maps the actual shape of the energy. When you canvass, you want to be inside the part of the gradient where damage is consistent across most roofs, not out on the fringe where it's one bruised roof per block. The gradient, not the single peak, is what you're mapping.

Step 1: Pull the storm data before you touch a doorbell

Do your desk work first. Fifteen minutes of map work saves you two days of aimless knocking. Gather these inputs and lay them over each other.

Hail reports and estimated stone size

Start with ground-truth reports. The NOAA Storm Prediction Center publishes daily storm reports with hail, wind, and tornado observations, including reported stone sizes and times. Pair that with the NWS local storm reports for the affected office. These are spotter and public reports, so coverage is uneven, but a cluster of 1.5-inch and 2-inch reports along a line tells you where the swath spine probably ran.

Radar-derived hail size products add resolution between the sparse ground reports. Maximum Estimated Size of Hail (MESH) is a radar algorithm that estimates the largest hail a storm could produce per grid cell. Several commercial storm-mapping vendors package MESH and similar fields into swath maps you can buy by event. The estimate is not gospel, but a continuous MESH swath fills the gaps between scattered spotter reports and shows you the ribbon shape.

Wind data

For wind events, pull measured gusts from nearby ASOS/AWOS stations and the SPC wind reports. Straight-line wind damage also runs in swaths and corridors, and it concentrates where terrain funnels or where a downburst hit. Damage indicators on the ground, fence lines down in one direction, debris fields all pointing the same way, confirm the corridor.

Storm motion and time

Note the direction the storm tracked and roughly when it crossed each area. Storm motion gives you the long axis of the swath. Timing helps you reconstruct which neighborhoods got the mature, high-energy part of the storm versus the weakening tail.

Building stock and roof age

Hail doesn't read property records, but your close rate does. Two streets can take identical hail and produce very different job counts because one is full of 22-year-old roofs at the end of their service life and the other was re-roofed after the last big storm five years ago. An older asphalt shingle near the end of its life bruises and fractures far more readily than a roof with plenty of mat strength left. A street of aging roofs in the swath spine is the richest target on the map. A street of recently replaced roofs in the same spine will frustrate you. More on layering this in below, because it's the signal most contractors ignore and the one that most reliably separates a knockable street from a dead one.

Lay it over a parcel map

Finally, drop all of the above onto a street-level map with parcels. You want to see swath plus streets plus, ideally, roof age together. County GIS and assessor sites give you parcels and sometimes year-built. Now you can draw your first-pass priority zones before the wheels even turn.

A field checklist for the desk phase

Keep this short and repeatable so you actually run it every storm instead of winging it:

• SPC storm reports pulled for the event date and time window
• NWS local storm reports for the affected office, noting reported stone sizes and times
• MESH or commercial swath map overlaid to fill gaps between reports
• Nearby ASOS/AWOS gust measurements pulled if wind was a factor
• Storm motion direction noted (gives the swath long axis)
• Approximate time the storm crossed each neighborhood reconstructed
• Wind-drive direction estimated from reports and damage indicators
• Parcel and year-built data pulled for the brightest third of the swath
• First-pass priority zones drawn and ranked before anyone leaves

If you can't answer the wind-drive direction, that's fine, you'll confirm it on the windshield survey by reading bent A/C fins and the side of the street that got hit. But everything else on the list should be settled at the desk. The crew should never leave the shop without a ranked street list in hand.

What free data can and can't do

Be realistic about the public sources. SPC and NWS reports are free and authoritative, but they're sparse, a rural swath might have two reports across nine miles, and report locations are approximate. They tell you a storm was real and roughly how big; they don't draw you a clean ribbon. MESH and commercial swath products give you the continuous shape but cost money and carry their own error, MESH is an estimate of the largest possible hail per grid cell, not a measurement of what actually fell on a given roof. The sound approach is to triangulate: use the free reports to anchor reality, the MESH to draw the shape, and your own windshield survey to confirm the ground truth. No single source is trustworthy alone.

Step 2: Build your priority map

With the data assembled, score and rank. You're converting a fuzzy swath into a ranked list of streets so the crew always knows the next-best block.

Here's a simple scoring frame that works without fancy software. Rate each candidate street on four factors and total them.

A street sitting on the swath spine with two-inch reports, full of early-2000s roofs, presenting west faces to a west-northwest drive scores near the top and goes to the front of the line. A street on the swath edge, recently re-roofed, with stones driven away from its visible slopes drops to the bottom even if it's technically inside the warned area.

This is also where you set boundaries that keep your crew efficient. Define the territory edges so canvassers aren't tempted to wander off the swath chasing a single damaged house surrounded by clean ones. One bruised roof on the swath edge is usually noise; ten in a row is the spine. Knock the spine.

A worked example

Say a storm tracks east-southeast across the north side of a metro at 7:40 p.m. The SPC reports show 1.75-inch hail near a golf course, 2-inch near a high school three miles down-track, and 1-inch at the far edge. A vendor MESH swath draws a 9-mile ribbon roughly 1.5 miles wide connecting those points, brightest in the middle third.

You pull parcels for the brightest third. Two subdivisions sit dead center. One was built in 1999 to 2003 and shows no recent permits. The other was rebuilt after a 2021 storm and shows a wall of re-roof permits. Same hail, very different prospects. You rank the 1999-era subdivision first, work it slope-by-slope favoring the west and north faces, and only after you've worked the spine do you consider the edges. That ranked list, drawn before anyone knocks, is the whole game.

Put numbers on it and the ranking gets obvious. Take three streets from that storm:

Oak Ridge wins decisively, hit hard, on the spine, old roofs, right aspect. Maple Crossing took the same hail but the fresh roofs gut its score, you'll burn an afternoon there hearing "my roof's only four years old." Birch Ln has old roofs but sits on the edge where damage is spotty. The ranking writes your route: Oak Ridge today, Birch Ln as a fill-in, Maple Crossing only if you run out of better doors. That's the entire value of scoring before you knock, it stops the crew from treating every street in the polygon as equal.

Re-scoring as the storm reputation develops

Your first-pass scores are a hypothesis built on data. After a day of knocking and a handful of inspections, you'll have real signal, which streets are producing functional-damage inspections and signed jobs, and which are coming back cosmetic. Feed that back into the scores. A street that scored 12 but keeps inspecting cosmetic gets marked down; a street that scored 8 but is quietly converting gets marked up. By day two or three your ranked list should be driven as much by your own results as by the original map. The map starts the hunt; your knock data finishes it.

Step 3: Drive the swath to ground-truth it

Maps get you close. Your eyes confirm it. Before you commit the crew to a street, do a windshield survey of the candidate zones. You're looking for the tells that a roof got hit, visible from the road at 15 miles an hour.

What to scan for from the truck:

• Soft metals. Dented gutters, downspouts, fascia, and especially metal roof vents, turtle vents, ridge vents, and chimney caps. Soft metal dents are the loudest roadside signal of hail size and density. A street where half the vents are pocked is a street that got cored.
• Window screens and wraps. Torn or dimpled aluminum window screens and dinged garage doors and A/C condenser fins. The A/C fin combing is a quiet giveaway, hail bends the fins in the direction it was driven, which also tells you aspect.
• Shingle debris and granules. Granule wash at downspout outlets and in the street gutters, plus shingle fragments or whole tabs in yards after a wind event. Fresh asphalt granules look like coarse dark sand piling at the bottom of downspouts.
• Collateral. Shredded leaves and stripped branches under trees, split or pocked wood fences, holes in trampolines and patio screens, splatter marks on decks and concrete that haven't weathered away yet. Splatter marks are time-sensitive evidence and one of the cleaner ways to confirm a fresh, sizable event.
• Other roofers. Competitor yard signs and lawn-stake flags going up, and other crews already canvassing. If three companies are already on a street, the storm is real there, but you may be late. If a clearly hammered street has no signs yet, you found it first.

Drive the spine first, then a cross-street or two to feel the swath edges. The goal is to confirm the worst blocks and trim any streets your map over-rated. Ten minutes of windshield work can save a canvasser a wasted afternoon.

What windshield surveys get wrong

Two failure modes to watch. First, clean-looking south and east slopes on a wind-driven storm. From the street you often see the leeward face, which can look untouched while the windward face behind it is destroyed. Don't write off a street because the road-facing slopes look fine, check the soft metals and the collateral, which don't lie about aspect.

Second, old damage masquerading as new. Faded splatter marks, weathered dents, and oxidized exposed mat are from a prior event. Fresh hail damage has bright, clean fracture points and unweathered splatter. Calling old damage fresh is how contractors get themselves and homeowners into trouble with carriers and inspectors. If you can't tell, that's a roof to inspect carefully, not a street to skip.

Reading stone size from the ground

You can estimate the hail size that hit a street without ever climbing a ladder, and that estimate sharpens your ranking. Soft-metal denting scales with stone energy. Pea to dime-sized hail rarely dents gutters meaningfully and seldom produces functional roof damage on a healthy shingle. Quarter-sized hail starts to dent soft metals and bruise older shingles. Golf-ball and larger leaves obvious, deep dents in gutters and downspouts, cracks window screens, splits fence pickets, and reliably produces functional damage across roof slopes regardless of age. When you see deep, dense denting and split wood from the street, you're on a street that almost certainly has functional roof damage, push it up the list. When the soft metals look nearly clean, be skeptical even if the area was warned.

Aspect read from the curb

The collateral also hands you the wind-drive direction for free. Bent A/C condenser fins point the way the hail was driven, comb them with your eye and you'll see the lean. Splatter marks on the windward sides of poles, mailboxes, and fence posts but not the leeward sides tell the same story. Once you know the drive direction on one street, it usually holds across the neighborhood, so you can predict which roof faces to inspect first on every house and which side of each street took the worst of it. That one read can double a canvasser's efficiency, because they lead every conversation pointing at the slope that's actually damaged.

Mapping the swath edges with cross-streets

Don't just drive the spine. Run a couple of cross-streets perpendicular to the swath so you can feel where the damage falls off. The point where soft-metal denting fades from obvious to occasional is your practical swath edge, the streets beyond it aren't worth a full canvass even if the warning polygon technically covered them. Marking those edges keeps canvassers from drifting off the productive blocks chasing the occasional damaged outlier.

Step 4: Where roof age data changes the game

Here's the layer most storm chasers leave on the table. The swath tells you where the stones fell. It does not tell you which of those roofs were already vulnerable. And vulnerability is half of whether a street converts.

A roof's response to hail depends heavily on the condition and age of the shingles. As asphalt shingles age, the mat gets brittle, the asphalt loses volatiles, and granule adhesion weakens. An older shingle fractures and bruises under impacts that a newer, more pliable shingle of the same product would shrug off. So inside one swath you can have a 6-year-old roof that took 1.5-inch hail and shows only cosmetic marks, and a 21-year-old roof one street over that took the same stones and now has functional damage across multiple slopes. The second street is your street.

The problem is that roof age is invisible from a swath map and laborious to pull house by house from permit records. This is exactly the gap that per-address roof-age intelligence fills, and it's worth being precise about what it does and doesn't tell you.

Tools like RoofPredict estimate a roof's age as a range from aerial and satellite imagery, house by house, and model storm exposure per roof rather than per zip code. Practically, that lets you do two things you otherwise can't:

1. Rank streets by how many end-of-life roofs sit inside the swath. Instead of guessing which subdivision is the 1999 build and which is the 2021 rebuild, you can see the age profile of the block and push the older, more storm-vulnerable streets to the top of the canvass list.
2. Stack storm exposure on top of age, per roof. A roof that is both aging out and sat in the high-energy part of a recent storm is the highest-probability door on the map. Working those first concentrates your crew where the odds are best.

Be honest about the limits, because overselling this gets contractors burned. A roof-age figure is a range, not a birth certificate, it narrows the field, it doesn't replace an inspection. And storm exposure modeling gives you odds, not proof, it tells you which roofs the storm most likely wore on, not which roofs are definitively damaged. The roof still has to be inspected, the conditions documented, and the decision about coverage still belongs to the insurer with the homeowner owning the claim. What the data buys you is sequencing: you knock the roofs that are most likely due first, instead of knocking doors in map order and hoping.

Used that way, age-plus-exposure data is the cleanest answer to the real question behind "which streets got hit worst," which is actually "which streets will convert." The worst-hit street full of brand-new roofs is a worse target than a moderately-hit street full of roofs aging out. The map alone can't tell you that. The age layer can.

Doing the age layer manually, if you don't have a tool

You can approximate the age layer by hand; it's just slower. County permit records show re-roof permits, so a subdivision with a wall of recent re-roof permits is freshly redone and worth deprioritizing inside the swath. Assessor year-built data gives you the original roof age for houses that have never been re-roofed, useful in older neighborhoods where the original roof is plausibly still up there. Subdivision build-out dates tell you when whole tracts went up, so a tract built in 1999 and never permitted for a re-roof is full of roofs around or past their typical service life. None of this is per-roof precise, a single house in a 1999 tract might have been quietly re-roofed without a permit, but at the street level it's directionally right and far better than ignoring age entirely. The tradeoff is labor: pulling permits and year-built house by house across a nine-mile swath is hours of work, which is exactly the bottleneck per-address tools remove.

Why age beats stone size as a tiebreaker

Here's the counterintuitive part worth sitting with. Between two streets, if one took slightly bigger hail but has newer roofs and the other took slightly smaller hail but has roofs aging out, the older-roof street usually converts better. A near-end-of-life shingle fractures under impacts a healthy shingle survives, so the vulnerability gap can outweigh a modest stone-size gap. This is why a swath map alone, which only shows stone size, systematically mis-ranks streets. It will send you to the biggest-hail block even when the better-converting block is the older neighborhood a half-mile off the peak. Treat stone size and roof age as two factors of roughly equal weight, not stone size first.

Step 5: Sequence the knock

Now you have a ranked map and confirmed blocks. Sequencing the actual canvass is where shops leak the most productivity.

Work the spine in tight loops, not long lines. Assign a canvasser a compact cluster of high-score streets, not a three-mile corridor. Dense, contiguous territory means more doors per hour and shorter walks, and it concentrates your yard signs so the social proof compounds on the streets that matter.

Favor the wind-driven faces. On streets where you know the drive direction, start with the houses presenting their hit slopes to the road and to view. Those are the easiest first conversations because the damage is more apparent, and early easy wins build a canvasser's momentum for the day.

Time of day matters. Late afternoon and early evening on weekdays, and mid-morning to early afternoon on weekends, put you in front of the most people answering doors. Don't waste prime swath streets on a 10 a.m. Tuesday when nobody's home, save the spine for when people answer.

Re-knock the hot streets. A no-answer on your best street is not a no. The spine streets are worth two or three passes at different times before you move on. A single dry pass through a cored street and then abandoning it is one of the most common and most expensive mistakes in canvassing.

Track outcomes by street. Log knocks, contacts, inspections set, and signed jobs per street as you go. After the first day you'll see which blocks are actually converting, and you can double down on the live ones and quietly retire the streets that aren't producing despite a good map score. The map is a hypothesis. Your knock data is the result.

Crew assignment and density

How you split territory among canvassers matters as much as which streets you pick. Carve the swath spine into compact, contiguous chunks and give each canvasser one chunk, not a scattered set of streets across the map. Density does three things: it cuts walking time so doors-per-hour climbs, it concentrates your yard signs so social proof builds on the streets that matter, and it makes re-knocking easy because the canvasser is never far from a no-answer house. A canvasser bouncing across a wide, fragmented territory spends half the day in the truck. Tight loops keep them on porches.

Reconciling canvass pace with install capacity

There's a discipline most growth-minded shops skip: don't sign faster than you can build. Storm work has a freshness window for the homeowner, too. A roof signed today but not started for seven weeks is a roof a competitor can poach with a Monday start date. Before you flood a great swath with canvassers, know your real install throughput for the next few weeks, and pace the sell to it. It feels backward to slow down on a hot storm, but a backlog you can't service quickly bleeds signed jobs and torches your reputation when homeowners wait and wonder. Sell the spine hard, but sell what you can build.

A note on door-to-door etiquette that actually lifts close rates

The mechanics of the knock are outside the scope here, but one thing ties directly to street selection: leading with the specific. A canvasser who can say, accurately, "the storm Tuesday evening drove hail onto the west-facing slopes on this street, and I'd like to take a look at yours" outperforms one reciting a generic script, because it's true and it's specific to that house. All the desk and windshield work, the swath, the aspect, the timing, hands the canvasser those specifics. The better your map work, the more concrete and credible your canvassers can be at the door. Precision upstream pays off as trust at the threshold.

A simple daily structure

Step 6: The inspection that confirms the street

Knocking finds the doors. The inspection confirms whether the street is what the map promised, and the first few inspections on a new block tell you whether to pour resources in or pull back.

When you get on a roof, inspect every slope, including the inconvenient ones. On a wind-driven storm the damage clusters on the windward faces, and an inspector who skips them will under-call a hammered roof. Document what you find plainly, photograph it, measure it, and let the conditions speak for themselves. Your job is to document conditions and provide an estimate, the insurer evaluates coverage and the homeowner owns the claim. Stay in that lane and your reputation, and your standing with adjusters, stays clean.

What functional hail damage actually looks like, so you can tell a real street from a cosmetic one:

• Bruising / mat fracture. A soft spot you can feel, where the impact fractured the shingle mat. This is the functional damage that matters, granule loss alone at a strike point may be cosmetic depending on the product and inspector.
• Random pattern of strikes. Genuine hail hits are randomly distributed and roughly circular, with directionality tied to the wind drive. Patterns that are too regular point to mechanical or foot-traffic marks, not hail.
• Fresh, unweathered impacts. Bright fracture points and clean exposed mat indicate a recent event. Oxidized, dirty, weathered marks are old.
• Soft-metal corroboration. The dents on vents, flashing, and valleys should match the size and direction story the shingles tell. When the soft metals and the shingles agree, you have a coherent, defensible picture of the event.

If your first two or three inspections on a street confirm consistent fresh functional damage, that street is real, surge the crew onto it. If the inspections keep coming back cosmetic or old, the map over-rated that block, and you should reallocate before you sink another day into it.

Documenting so the picture holds up

Good documentation isn't about building a case for the homeowner; it's about recording conditions accurately so everyone downstream is working from reality. Photograph each damaged slope with something for scale, mark and circle representative impacts, and note the slope orientation so the aspect story is on the record. Capture the soft metals, the dented vents and bent fins, in the same set, because they corroborate stone size and direction. Record the date and, if you can establish it, the storm event you believe caused the damage. Then provide a clear estimate of the work the conditions call for. The insurer evaluates coverage against the policy and the homeowner owns the claim, your contribution is an honest, well-documented record of what the roof actually shows. Shops that document cleanly and consistently earn smoother adjuster interactions over time, which compounds across a market you'll storm again.

The test-square habit

A disciplined way to confirm a street is the test square: on a representative roof, mark off a 10-by-10-foot area on a damaged slope and count the impacts. It standardizes your read so you're comparing apples to apples between houses and streets, and it forces you to look at density rather than cherry-pick the one dramatic strike. A slope with numerous fresh functional impacts in a test square is a different animal from a slope with two marginal ones, and the difference should drive whether you commit the crew to that street. The test square also keeps you honest: if you can't find consistent impacts in a square on what should be a hammered roof, the street may be cosmetic, and you've learned that for the cost of ten minutes.

Building a storm response system you can repeat

The shops that win storm after storm don't rediscover this workflow each time. They systematize it. A few elements make the difference between a scramble and a machine.

A pre-built data routine. The desk-phase checklist above should be a standing procedure someone runs the morning after every event, not an improvisation. Whoever holds that role pulls the reports, builds the swath, layers age, and hands the crew a ranked list before the first knock. When that's a habit, your crew is on the spine while competitors are still arguing about where to start.

A standing territory map. Keep your service area mapped with the data you reuse every storm, parcels, build-out dates, known re-roofed tracts, prior storm history. Then each new event is just a swath laid over a map you already understand, not a cold start. Shops that know their territory cold can rank a new storm in an hour.

A feedback loop. Per-street outcome tracking shouldn't die at the end of a storm. The streets that converted, the building stock that responded, the neighborhoods where adjusters were reasonable, all of that is intelligence for the next event. Over a few seasons you build a private map of how your specific market behaves, which is an edge no swath vendor sells.

Capacity-aware pacing. Tie your canvass intensity to install throughput as a standing rule, not a panicked mid-storm realization. Knowing your real weekly build capacity lets you scale canvassers up or down so you sell what you can deliver.

The storm is the same for everyone. The system is what's yours. Two crews under the same swath post different numbers because one runs a repeatable process from data to door to roof, and the other drives toward the dot and hopes.

Common mistakes that cost contractors the best streets

A few patterns show up again and again, and each one quietly hands your best blocks to the competition.

Chasing the radar instead of the swath. The bright radar cell is where hail was aloft. Ground reports, MESH, and your windshield survey tell you where it landed. Crews that knock under the radar dot instead of along the confirmed swath spine waste their best hours on the wrong streets.

Canvassing uniformly across the warning polygon. Spreading the crew evenly over a 40-square-mile warned area guarantees most of your knocks land on undamaged streets. Concentrate on the ribbon.

Ignoring aspect. Writing off streets because the road-facing slopes look clean, on a storm that drove the hail onto the back faces. Read the soft metals and collateral, not only the visible roof.

Ignoring roof age. Pounding a cored but recently-re-roofed subdivision and getting nothing but "my roof's only four years old." Same hail, wrong building stock. Layer age onto the swath so you knock the streets that are both hit and vulnerable.

One-pass canvassing. Driving a great street once at 11 a.m., getting mostly no-answers, and never coming back. The spine streets deserve multiple passes at different times.

Confusing old damage with new. Calling weathered, oxidized marks fresh. It poisons your credibility with adjusters and homeowners and can cross legal and ethical lines. When in doubt, inspect carefully and document honestly rather than overstate.

Outrunning your install capacity. Signing four streets you can't service for six weeks. Storm work has a freshness window for the homeowner too, sell what you can build, or you'll lose signed jobs to a competitor who can start Monday.

Working safely and within the rules

Two guardrails keep a storm operation out of trouble.

Crew safety. Storm roofs are wet, littered with debris, and sometimes structurally compromised. Fall protection, ladder discipline, and not putting people on roofs in lightning or high wind are non-negotiable. OSHA's fall-protection requirements apply on every roof, and a hurried storm push is exactly when shops cut corners and get someone hurt. A injured canvasser or inspector costs you far more than a missed street.

Honest representation. Storm-chasing has a reputation problem for a reason, fly-by-night operators who overpromise, fabricate damage, or pressure homeowners. Don't be that. Document real conditions, give honest estimates, let the insurer make the coverage call, and let the homeowner own their claim. Many states regulate roofing contractors' conduct around insurance work through their Department of Insurance, and consumer-protection rules apply to door-to-door sales. Being the company that plays it straight is also, not coincidentally, the company adjusters trust and homeowners refer. Reputation is the compounding asset in a market you'll storm again.

Putting it together

Finding the worst-hit streets is a repeatable process, not a hunch. Pull the storm data, hail reports, MESH, wind measurements, and storm motion, and lay it over a street map. Score and rank candidate streets by swath intensity, position on the spine, roof-age profile, and aspect exposure. Ground-truth the top zones with a windshield survey, reading soft metals and collateral, not only the road-facing slopes. Knock the spine first, in tight loops, at the times people answer, and re-knock your best streets. Confirm with thorough, every-slope inspections, and let your real per-street outcomes re-rank tomorrow's map.

The layer that ties it together is matching the swath to the building stock. The street that converts is the one that is both hit and due, where aging roofs sat in the high-energy part of the storm. Per-address roof-age ranges and per-roof storm-exposure modeling, the kind RoofPredict produces, let you sequence your canvass toward those doors instead of knocking in map order, as long as you treat the age as a range and the exposure as odds, and confirm everything on the roof. Map first, then knock, then inspect, then re-rank off results. Run that loop with discipline and the same storm that gives a sloppy crew two jobs will give yours eleven.

FAQ

How soon after a storm should I start canvassing?

As fast as you safely can, ideally within the first 72 hours. Damage is fresh and easy to spot, splatter marks haven't weathered away, homeowners are receptive, and the field isn't picked over yet. Spend the first hour or two on desk work and a windshield survey to find the swath spine, then knock. Precision early beats effort late, because every hour on a skipped street is an hour a competitor spends on a cored one.

Where do I get a hail swath map for a specific storm?

Start free with the NOAA Storm Prediction Center storm reports and your local National Weather Service local storm reports for reported stone sizes and times. For continuous swath coverage between sparse ground reports, several commercial vendors sell radar-derived MESH (Maximum Estimated Size of Hail) maps by event. Combine the two: ground reports confirm real stones hit the ground, and MESH fills in the ribbon shape between them.

Why did one street get destroyed and the identical street a mile away get nothing?

Hail falls in a swath, and inside that swath the storm dumps its biggest stones in pulses as the updraft strengthens and weakens. You get hot cells, gaps, then more hot cells. So one block can sit under an intensification burst while a block up the road catches a lull. Roof age compounds it: a street of aging roofs fractures under stones a street of newer roofs would shrug off.

Does radar tell me which streets got hit hardest?

Not directly. Radar shows where hail and rain were aloft, not where damaging hail reached the ground or how big the stones were when they landed. A bright core can melt before it lands, and a modest signature can drop dense, wind-driven stones. Use radar as a clue, then confirm with ground reports, MESH, and your own windshield survey of soft metals and collateral damage.

What should I look for during a windshield survey?

Scan for dented soft metals (gutters, downspouts, roof vents, chimney caps), bent A/C condenser fins, torn window screens, dinged garage doors, granule wash at downspouts, shredded leaves and stripped branches, pocked wood fences, and fresh splatter marks on decks and concrete. Also note competitor signs. These roadside tells confirm the storm reached the ground and hint at stone size and the direction the hail was driven.

Why does roof age matter so much for choosing streets?

As asphalt shingles age, the mat gets brittle and granule adhesion weakens, so older roofs fracture and bruise under impacts newer roofs would survive. Inside one swath, a 6-year-old roof might show only cosmetic marks while a 21-year-old roof one street over has functional damage on multiple slopes. The street that converts is both hit and aging out, so layer roof age onto the swath rather than ranking by stone size alone.

How does per-address roof-age data change my canvass?

It lets you rank streets by how many end-of-life roofs sit inside the swath and stack storm exposure on top of age per roof, so you knock the roofs most likely to be due first instead of knocking in map order. Treat the age as a range, not an exact date, and the storm exposure as odds, not proof. It improves sequencing; it does not replace the on-roof inspection or the insurer's coverage decision.

How do I tell fresh hail damage from old damage?

Fresh impacts show bright, clean fracture points and unweathered splatter marks; old damage looks oxidized, dirty, and weathered. Genuine hail strikes are randomly distributed, roughly circular, and directional with the wind drive, while too-regular patterns suggest foot traffic or mechanical marks. Soft-metal dents should match the size and direction story the shingles tell. When the evidence disagrees or is ambiguous, inspect carefully and document honestly rather than overstate.

How many times should I knock the same street?

Your best swath-spine streets deserve two or three passes at different times before you move on. A no-answer is not a no, it usually just means nobody was home. Hit prime streets during peak answer windows (late afternoon and early evening on weekdays, mid-morning to early afternoon on weekends), and reserve low-answer times for warm-up streets. One dry pass and abandoning a cored street is one of the costliest canvassing mistakes.

What should I avoid so I stay on the right side of the rules?

Document only real, current conditions and provide honest estimates; let the insurer evaluate coverage and let the homeowner own the claim. Don't fabricate or exaggerate damage, don't call weathered old marks fresh, and follow your state's contractor and Department of Insurance rules plus consumer-protection rules for door-to-door sales. Keep crews safe with fall protection and ladder discipline. The straight-shooting company is also the one adjusters trust and homeowners refer.