The Best Way to Find Homeowners Who Need a New Roof

Ask ten roofing owners how they find work and you will get ten versions of the same answer: knock more doors, mail more postcards, buy more leads, chase more storms. All of it boils down to volume. Throw enough effort at a big enough pile of houses and some of them will need a roof.

That works, in the sense that a bucket with a hole in it still holds some water. But it is the most expensive way to grow a roofing company, and it burns out good salespeople faster than anything else in the business. The crew that knocks 200 doors to find two real opportunities does not quit because the work is hard. They quit because 198 of those doors made them feel like a nuisance, and most of those homeowners had a six-year-old roof that was never going to buy from anyone.

The roofers who grow without that grind are doing something different at the front of the funnel. Before they spend a stamp, a gallon of gas, or an hour of a rep's time, they have already narrowed the universe of houses down to the ones with a real reason to need them: roofs that are old enough to be at the end of their service life, and roofs that a storm actually worked over. Everything downstream — the script, the mailer, the follow-up — gets easier when the list underneath it is right.

This is a working playbook for building that list. It covers what actually tells you a roof is due, how to read it from the curb and from the air, how storms change the math house by house, the channels that turn a good list into booked jobs, and the mistakes that quietly waste most roofers' marketing money. There are worked numbers, checklists, and scripts you can lift. Read it as a field manual, not a pep talk.

What "needs a new roof" actually means

Before you can find these homeowners, you have to be honest about what you are looking for. "Needs a new roof" is not one thing. It is three different situations that look different from the street, sell differently, and require different lists.

1. Age-out roofs. The covering has reached or passed the end of its useful service life. Nothing dramatic happened. The asphalt shingles are just old, the granules have shed, the mat is brittle, and the roof is at the point where a smart homeowner replaces it before it starts leaking. This is the biggest and most predictable category, and almost nobody markets to it well because it is invisible from a moving truck.

2. Storm-damaged roofs. Hail bruised the mat or wind lifted and creased shingles, and the roof now has accelerated failure ahead of it even if it looks fine from the ground. The homeowner may have an insurance path to a replacement. This category is lumpy — it shows up in concentrated bursts after a hail or wind event — and it is the most competitive because every restoration crew in three states converges on the same ZIP codes.

3. Failed or failing roofs. Active leaks, visible sag, missing sections, tarps on the deck. These homeowners already know they have a problem. They are searching online, calling, and getting bids. You find them through inbound (search, referrals, your reputation), not through outbound targeting, and they are a small share of the total.

Most "find homeowners who need a roof" advice is really about category 3 — the people already raising their hand — or category 2 in the two weeks after a storm. The money that compounds is in getting good at category 1, the age-out roofs, and in being early and accurate on category 2 instead of late and scattershot. The rest of this is built around those two.

Why roof age is the master variable

Asphalt shingle roofs dominate the U.S. residential market, and they have a finite service life. A standard 3-tab roof installed a couple of decades ago, and many builder-grade architectural roofs, run into the 15-to-25-year range depending on the product, the ventilation, the slope, the install quality, and the climate. The National Roofing Contractors Association and shingle manufacturers publish service-life guidance; the point is not a magic number but a window. Once a roof crosses into that window, the odds it gets replaced in the next few years climb sharply, and they climb whether or not a storm ever hits it.

That single fact reorders your whole prospecting strategy. If you could sort every house in your service area by the likely age of its current roof, you would knock the 18-to-24-year-old roofs first and skip the 6-year-old ones entirely. You would mail the old roofs and not waste postage on the new ones. You would hand a green canvasser a route where most doors have a real reason to talk.

The problem is that roof age is genuinely hard to see, and almost every tool roofers reach for gives them the wrong number. We will fix that next.

The signals that actually tell you a roof is due

There are three layers of signal: what you can read from the air or a database before you ever drive, what you read from the curb in a drive-by, and what you confirm up close. Pros stack all three. Amateurs rely on the last one, which means they have already spent the gas before they learn anything.

The year-built trap

Start with the most common mistake, because almost everyone makes it. Roofers pull a list of homes by year built — from the county assessor, a list broker, or a property data site — and treat "older house" as "older roof." It is the easy data to get, so it is the data people use.

It is wrong often enough to wreck a campaign. Year built tells you when the house went up, not when the roof was last replaced. A 1995 house may be on its third roof. A 2003 house may have had a hail-driven replacement in 2011 and another in 2019. Re-roofs are invisible in the year-built field. So when you mail "older homes," a big slice of your list already has a newer roof than your data implies, and you are paying to reach houses that cannot buy from you.

Permit records help a little — many jurisdictions require a permit for a re-roof — but coverage is wildly uneven. Plenty of re-roofs happen without a pulled permit, small towns may not digitize records, and matching permits cleanly to addresses across a metro is its own project. Permits are a useful supplement, not a foundation.

Reading roof age from aerial imagery

The most reliable signal that scales is the roof itself, seen from above. High-resolution aerial and satellite imagery now covers nearly every populated address, and a roof's surface changes in visible ways as it ages. From the air you can often pick up:

• Granule loss and color fade. New asphalt shingles read as deep, even color. As granules shed, the surface lightens, goes blotchy, and the underlying mat shows through in patches — especially on south- and west-facing slopes that take the most sun.
• Streaking and biological growth. Long dark streaks (often algae) and green tinting collect on older, weathered roofs and on slopes that stay damp.
• Surface texture and unevenness. Cupped, curled, or lifting shingles change how light scatters off the roof. An old roof looks rougher and less uniform from above than a fresh one.
• Repairs and patches. Mismatched shingle color on one slope or around a penetration signals a roof that has already been worked on and is likely near the end.
• Tarps, missing sections, exposed deck. Obvious distress, visible even in a quick scan.

No single frame is proof of an exact age. But across a whole neighborhood, these visual cues sort roofs into believable bands — new, mid-life, aging, end-of-life — far better than year built does, because you are looking at the actual covering instead of guessing from the house's birthday.

The honest limit: imagery gives you a range, not a date. You are estimating that a roof reads as roughly 15-to-20 years old, not stamping it "installed March 2007." Treat anyone who promises an exact install date from a photo with suspicion. A tight, defensible range is what you actually need to prioritize a route, and it is what good aerial analysis delivers.

Curb-side signals on the drive-by

Once a rep is on the street, the read gets richer. Train your people to clock these in the first ten seconds, because they sharpen the air-read and tell you who to talk to first:

• Granule wash-out in gutters and at downspout splash blocks (dark grit = shingle erosion).
• Curling, cupping, or clawing shingle edges visible against the sky line.
• Bald, shiny spots where the mat is exposed.
• Sagging ridge or rolling deck lines.
• Rusty or worn flashing, pipe boots cracked and gaping, exposed nail heads.
• Mismatched patches and prior repairs.
• Daylight-obvious storm signs: lifted or missing shingles, debris, dented gutters, fascia damage.

Up-close confirmation

The last layer is the inspection itself: hands on the roof, photos of mat bruising from hail, creased shingle tabs from wind uplift, granule loss in the field, soft or spongy decking, and the condition of flashing, valleys, and penetrations. This is where you confirm whether a roof is genuinely failing or storm-impacted versus merely cosmetically tired. It is also the most expensive signal to collect, which is exactly why you want the first two layers to filter hard before you spend a rep's time getting on a ladder.

Stack the layers and the funnel inverts in your favor: the air-read and database eliminate most of the street, the drive-by ranks who is left, and the inspection closes. You stop discovering the 6-year-old roofs after you have already knocked them.

Storms change the math — house by house, not ZIP by ZIP

Age is the steady engine. Storms are the spikes. And storms are where the most money is won and lost, because the standard way roofers target storm work is far too coarse.

The usual move after a hail event is to grab a hail map — a colored polygon showing where reported hail of a given size fell — and canvass everything inside the blob. That is better than nothing, but it treats a whole swath of homes as identical when they are not. A hail map answers "where did it hail?" It does not answer the question that actually pays you: "which roofs did it wear out?"

Those are different questions, and the gap between them is large.

Why a hail polygon over-counts and under-counts

Hail damage to a specific roof depends on more than whether a stone fell somewhere on the block. It depends on:

• Stone size and density at that exact location — hail cores are narrow and patchy, and a reported size for a storm is a maximum somewhere in the area, not a uniform value across it.
• Wind direction and speed during the storm, which drives hail in at an angle and concentrates impacts on the windward slopes while sparing the leeward ones. Two identical houses one street apart can take very different damage from the same cell.
• Roof slope and orientation — steeper and differently-facing slopes catch stones differently.
• The roof's age and product. An older, brittle shingle bruises and fractures from a stone that a fresh, pliable shingle shrugs off. The same hail that does nothing to a 3-year-old roof can total a 17-year-old one next door.

So the polygon both over-counts (lots of homes inside it took little real damage, especially newer roofs on sheltered slopes) and under-counts (it ignores that an older roof at the edge of the swath may have meaningful damage while a new roof dead-center does not). Canvass the raw polygon and your crew spends days on doors that will not turn into approved work, while better opportunities at the margins go unworked because they were the wrong color on the map.

Wind is its own story

Wind damage does not even come as a tidy blob. Straight-line winds, downbursts, and gusts on a roof's exposure cause shingle uplift and creasing in patterns driven by direction, the building's height and surroundings, and the roof's existing condition. A gust map and a hail map rarely line up, and a roof can have a real wind claim with no hail at all. National Weather Service and Storm Prediction Center storm reports give you the event; they do not give you the per-roof consequence.

Modeling the storm on each roof

The sharper approach is to model what the storm did to each individual roof, not merely where the storm passed. That means combining, for a specific address: the storm's actual track and intensity, the likely hail size and wind vector at that point, the roof's geometry and which slopes were exposed, and the roof's estimated age and condition. The output is not "this ZIP got hail." It is a per-roof read on which homes likely have storm wear worth inspecting — and crucially, it pairs that with age, so you can tell the difference between a new roof that took a glancing hit and an old roof the storm pushed over the edge.

That pairing is the whole game. A roof that was already aging out and then took a real storm is your strongest opportunity in the area. A new roof in the dead center of the hail core is often a waste of a knock. The polygon cannot tell them apart. A per-roof model can.

A fair caveat: modeling produces odds, not proof. It tells you which roofs are most likely worth a look so you knock and inspect the right ones first. The inspection is still what establishes actual damage, and the homeowner's insurer is the one who decides coverage. Model output is a prioritization tool, never a substitute for getting on the roof and never a claim that damage exists. Used honestly, it means your crew spends its storm window on the roofs most likely to convert instead of grinding the whole polygon.

How RoofPredict fits into this

Everything above describes work most roofers cannot do at scale by hand. You can drive a few streets and eyeball roofs. You cannot personally age-read every roof in your service area, cross-reference storm tracks, and re-rank the list every time a cell moves through. That is the specific gap RoofPredict is built to close.

RoofPredict scans the homes in your area and, for each address, produces a read on roof age as a range (estimated from aerial imagery, not pulled from year-built) and a per-roof storm read that models hail and wind on that specific roof rather than dropping it inside a polygon. The two are combined into a simple priority view: which roofs are old enough to be due, which roofs a storm actually worked over, and especially which roofs are both. You get a ranked picture of your own streets — knock and mail the worn-out ones, skip the new ones.

Where this earns its keep:

• Targeting before you spend. Instead of mailing every "older home" or canvassing a whole hail polygon, you work a list already sorted by the two things that predict a re-roof: age and storm wear. The gas, the postage, and the rep hours land on doors with a real reason to talk.
• Mining your own book. Many roofers are sitting on years of old estimates and past customers whose roofs have quietly crossed into the age-out window. Run those addresses through the same age read and you find a season of jobs in money you already spent to acquire — no ad spend, no rented leads.
• Making green reps sound like veterans. A new canvasser who walks up knowing this roof reads as roughly 18-to-22 years old and took two notable hail events has a real, specific reason to be at the door. That rep closes more, gets fewer slammed doors, and stays — which fixes the rep-churn problem that quietly eats canvassing-heavy companies.

The honest limits, because a tight trade compares notes. RoofPredict gives you a range, not an exact install date, and a likelihood a storm wore a roof, not proof of damage — the inspection is still what establishes condition. It does not measure the roof or identify materials for an estimate; that is a different category of tool (EagleView, Roofr, HOVER measure a roof you have already chosen — RoofPredict tells you which roof to choose). And it does not buy or resell leads — there is no shared homeowner being sold to five of your competitors. It sharpens the outbound you already do on your own streets and your own customer list. If you want to pressure-test it, hand it a roof you already know the truth about and see whether the age read lands.

The rest of this is how to turn a list like that — however you build it — into booked jobs.

Turning a good list into booked jobs: the channels

A list is potential energy. The channel is how you convert it. Each channel has a different cost structure and a different best-fit for the three categories of "needs a roof." Match them deliberately instead of defaulting to whatever you have always done.

Door knocking and canvassing

Canvassing is still the highest-conversion outbound channel in roofing when the list is good, because you are in front of the homeowner, on their roof in many cases, the same day. It is also the channel most destroyed by a bad list, because nothing burns a rep out like rejection at doors that were never going to buy.

The entire economics of canvassing hinge on hit rate. Walk a route where one door in fifty has a real reason, and your rep gets demoralized, sloppy, and gone in a month. Walk a route pre-sorted so that a large share of doors are aging-out or storm-worn roofs, and the same rep has real conversations all day, books inspections, makes money, and stays. Targeting is not a nicety on top of canvassing; it is what makes canvassing survivable.

A canvassing workflow that respects the list:

1. Assign routes by roof, not by geography. Hand reps streets where the density of due roofs is high, rather than streets that are merely convenient to cluster. A tighter, lower-density route of good doors beats a fat route of random ones.
2. Lead with the specific, not the generic. "We're doing roof inspections in the neighborhood" is noise. "I work this area — your roof is reading at the older end and I noticed [specific curb signal] — mind if I take a quick look?" is a reason. Specificity is the difference between a pitch and a conversation.
3. Offer the inspection, not the sale. The ask at the door is a free inspection, not a contract. You are buying the right to get on the roof and show them what is actually there.
4. Document on the roof. Photos of granule loss, mat bruising, creased tabs, worn flashing. The homeowner who sees the pictures of their own roof is a different prospect than one who hears a sales claim.
5. Book the next step before you leave. Inspection booked, or a follow-up time set. A maybe with no calendar entry is a no.
6. Log every door. Knocked, talked, inspected, not-home, no. The not-homes are a follow-up list, not a dead end — a huge share of bookings come from the second or third pass.

A canvassing script that uses age and storm data

"Hi, I'm [name] with [company] — I work roofs here in [neighborhood]. I'm not selling anything at the door, but your roof is reading at the older end of its life [or: this street took a couple of real hail events the last few years], and from the street I can see [granule loss in the gutters / some curling up top]. I'm offering free inspections while I'm out here — it takes me about fifteen minutes on the roof and I'll show you photos of exactly what's up there, good or bad. Worst case you find out you've got years left. Want me to take a look?"

Notice it makes a specific, honest observation, asks for an inspection rather than a sale, and gives the homeowner an out where the answer might be good news. That is what gets you on the roof.

Direct mail

Mail is the workhorse for age-out roofs, the category canvassing reaches slowly and inbound never touches. Its whole economics live or die on the list, because mail cost is roughly fixed per piece while the value of a re-roof is large.

Work the numbers. Say a mail piece costs you about a dollar all-in (print, postage, design amortized) and a closed residential re-roof nets you several thousand dollars in margin. Mail 5,000 pieces to a generic "older homes" list and maybe a thin fraction respond, fewer convert, and a big chunk of your spend hit roofs that were replaced two years ago and could not buy if they wanted to. Now mail 2,500 pieces to a list sorted so most recipients have a roof in the age-out window. Fewer pieces, lower spend, and the response and close rates climb because the offer finally matches the roof. You spent less and booked more — purely from a better list.

Mail that works for due roofs:

• Speak to the roof's situation, not a discount. "Roofs in [neighborhood] built around [era] are reaching the age where they start to fail — get a free inspection before the leaks start" beats "$500 off!" because it gives a reason rooted in their actual situation.
• Make the inspection the offer. Same as the door — sell the free look, not the contract.
• Use neighborhood proof. "We just replaced three roofs on [nearby street]" is the strongest social proof a homeowner has, because they can drive past and see it.
• Sequence, don't blast. Two or three touches to a tight list beats one blast to a fat one. Most response comes after the first piece.
• Track by route. Code each mailing so you know which streets respond, and reinvest there.

Mining your own CRM and old estimates

The cheapest roof to find is one you already paid to find. Almost every established roofer is sitting on a list that outperforms anything they can buy: old estimates that never closed, and past customers from years back.

Two plays here, and they are unanimous favorites among the owners who try them because the cost is basically zero:

Old estimates that went cold. Every bid you wrote two, three, five years ago was a homeowner with a roof old enough that you quoted it. Those roofs are now older. Many of the people who said "not yet" or went with a cheaper bid are now genuinely due. Re-running those addresses for current age and any storm activity since tells you which cold estimates are now hot. This is found money sitting in your own files.

Past customers and their neighbors. A roof you installed a long time ago may be approaching its own end of life, and your former customers are your warmest possible audience. Their neighbors, who watched you work and have similar-age roofs, are nearly as warm. A simple check-in on those addresses turns a quiet customer list into a referral and re-roof engine.

The operational move is the same in both cases: take the addresses you already own, get a current read on age and storm exposure, sort by who is now due, and reach out. No ad spend, no rented leads, no competition for the contact — these are your people.

Inbound: be there when they search

The failing-roof category (and a chunk of storm work) finds you, if you are findable. A homeowner with a leak searches, asks a neighbor, or calls the company whose sign they saw down the street. To capture it:

• Local search presence. A claimed, accurate Google Business Profile with real photos and reviews, and a website that says where you work and what you do. This is table stakes and many roofers still get it wrong.
• Reviews as the trust layer. Homeowners choosing a roofer they did not seek out lean heavily on reviews. A steady habit of asking every satisfied customer compounds over years.
• Yard signs and wrapped trucks. Old-fashioned and still effective — a roof you are working is a billboard to the whole street, many of whom have a roof the same age.
• Referrals, asked for deliberately. The highest-trust lead there is. Build the ask into your closeout process instead of hoping for it.

Inbound is slower to build and you control the timing less, but it is the cheapest per job once it is running and it pairs perfectly with outbound: the homes your canvassing and mail warm up often convert later through inbound when they are finally ready.

What about buying leads?

Bought leads from aggregators have a place as fill, but understand what you are getting: a homeowner who is usually sold to several roofers at once, so you are racing competitors on price for a contact you do not own. The conversion math is tough and the margin pressure is real. Bought leads are a supplement when your pipeline is thin, not a foundation to build on. The whole thrust of everything here is the opposite model — finding the homeowners on your own streets and in your own book, where nobody else is calling the same person, so the work is yours.

Worked example: the same week, two ways

Numbers make the difference concrete. Say a crew has one rep, five working days, and a goal of booked inspections. Two versions of the same week, same effort, different list. The figures are illustrative, not a promise — your real rates depend on your market, your reps, and your offer.

Version A — volume, no targeting. The rep knocks a convenient subdivision, roughly 40 doors a day, 200 for the week. Maybe a quarter are home and willing to talk. Most of those roofs are a random mix of ages — plenty too new to matter. A small fraction of conversations turn into a booked inspection. The rep ends the week tired, discouraged by hours of "no," and the booked inspections are few and scattered.

Version B — same rep, list sorted by age and storm wear. The rep works a shorter, denser route of doors pre-identified as aging-out or storm-worn roofs — fewer total doors because the bad ones were stripped out. A much larger share of the people who talk have a real reason to. The conversation-to-inspection rate climbs because the rep is opening with a specific, true observation about that roof instead of a generic pitch. The rep books materially more inspections from fewer doors, and — the part that does not show up in a single week's tally — comes back Monday instead of putting in notice.

The lesson is not that targeting adds a few points of conversion. It is that targeting changes the rep's daily experience from rejection to real conversations, which is what determines whether your canvassing program exists at all in six months. The list is upstream of everything: morale, retention, close rate, cost per booked job.

A 30-day plan to rebuild your prospecting

If you want to move from volume-and-hope to targeted, here is a sequence that does not require blowing up your operation.

Week 1 — Audit what you have.

• Pull your old estimates from the last three to five years into one list of addresses.
• Pull your past-customer list.
• Note any storm events that hit your service area in recent years (NWS / Storm Prediction Center storm reports are public).
• Honestly assess your current targeting: are you mailing/knocking off year-built or off actual roof condition? Almost everyone is off year-built.

Week 2 — Get real roof reads.

• Get an age read on your old-estimate and past-customer addresses. This is the fastest win because the contacts are warmest.
• Pick one or two target neighborhoods and get them age-read and storm-read so you know which streets are dense with due roofs.
• Build a ranked list: due roofs first, new roofs cut.

Week 3 — Match channel to list.

• Re-run cold estimates that are now due as a personal outreach (call or knock — these are warm).
• Send a tight, sequenced mail drop to the age-out roofs in your target neighborhood.
• Route your canvasser to the densest streets of due roofs, with a script that uses the specific reads.

Week 4 — Measure and double down.

• Track per channel and per route: doors/pieces, conversations, inspections booked, jobs closed.
• Compare the targeted route's hit rate to whatever you were getting before.
• Cut the streets that did not respond, pour effort into the ones that did, and set the cadence to repeat monthly — because roofs keep aging and storms keep coming, so the list refreshes itself.

Edge cases and what the pros get wrong

The playbook is clean. Reality has corners. Here are the ones that separate the operators from the order-takers.

Don't only chase storms

Storm work is seductive because it is concentrated and urgent. But storm-only companies live in feast and famine, fight an out-of-town swarm every time a cell hits, and have nothing to do in a quiet season. The age-out roofs are there storm or no storm. A company that works both has a floor under it: steady age-driven work between storms, and the surge capacity to capitalize when one hits — on the right roofs, because you already know which ones were old and exposed.

Don't mistake "old house" for "old roof"

We covered the year-built trap, but it bears repeating because it is the single most common, most expensive error in roofing prospecting. Every campaign built on year built is paying to reach re-roofed homes. Read the roof, not the house's birthday.

Don't over-promise on storm damage at the door

This is both an ethics issue and a legal one, and it is where a lot of roofers get themselves in trouble. Your job at the door is to offer an inspection and document what is actually on the roof. It is not to tell a homeowner their roof is damaged before you have looked, to promise their claim will be approved, or to make any promise about their deductible. The roofer documents conditions and provides an estimate; the insurer decides coverage; the homeowner owns the claim. Stay on the right side of that line — both because unlicensed adjusting rules are real and enforced, and because the trust you build by being straight is worth more than any one job. Model and imagery output tells you which roofs are worth inspecting. It never tells a homeowner that damage exists. Only the inspection does that, and only the carrier decides what it means.

Don't let a good list rot

A list is a snapshot. Roofs age, storms hit, homes sell, roofs get replaced — by you or by a competitor. A list built once and worked forever decays. The companies that win treat prospecting as a recurring cycle: refresh the reads, re-rank, and re-work on a cadence. The roofs that were "almost due" last year are due now.

Don't ignore the not-homes

A huge share of canvassing bookings come on the second or third pass, not the first. "Not home" is the most common door outcome and the most wasted, because most reps never go back. Build the not-home list deliberately and re-walk it. The roof is still old; you just have not caught the owner yet.

Don't confuse measurement tools with targeting tools

Measurement platforms (EagleView, Roofr, HOVER) are excellent at what they do — measuring a roof you have already decided to bid. They do not tell you which roof to bid in the first place. Age-and-storm targeting answers "which house," measurement answers "how big is this house's roof." You need both, at different stages. Confusing them leads roofers to think they already have targeting when they only have measurement.

Respect contact rules

Mail, door knocking, and calling all sit under different rules. Honor Do-Not-Knock registries and local solicitation ordinances, mind do-not-call and texting regulations for any phone outreach, and keep your claims truthful — the FTC and state regulators care about how you market. Being targeted actually helps here: fewer, more relevant contacts mean fewer annoyed homeowners and fewer complaints than a spray-and-pray blast.

A field checklist you can hand a crew

Print this. It is the whole thing on one page.

Before you spend anything:

• Is this list sorted by actual roof age, or by year built? (If year built, fix it.)
• Have we layered in storm history per roof, rather than a polygon?
• Have we mined our own old estimates and past customers first?
• Are new roofs cut from the list so we don't pay to reach them?

At the door / in the mailbox:

• Are we leading with a specific, true observation about that roof?
• Is the offer a free inspection, not a contract or a discount?
• Are we using neighborhood proof (jobs nearby they can see)?
• Are we documenting with photos of the actual roof?
• Did we book the next step before leaving / set the next mail touch?

On storm damage:

• Are we inspecting before we claim anything about damage?
• Are we staying off the homeowner's claim and deductible entirely?
• Are we documenting conditions and providing an estimate — and letting the carrier and homeowner do their parts?

On the back end:

• Are we logging every door / piece / call by route?
• Are we re-walking not-homes?
• Are we refreshing the list on a cadence so it doesn't rot?

The short version

The best way to find homeowners who need a new roof is to stop treating every house the same and start with the two facts that actually predict a re-roof: how old the roof is, and what storms have done to it. Read the roof from the air and the curb instead of guessing from the year the house was built. Model the storm on each roof instead of canvassing a colored blob. Mine your own old estimates and past customers before you buy a single new contact. Then match the channel to the list — canvass the dense streets of due roofs, mail the age-out homes, work your own book, and be findable for the ones already searching.

Do that and the whole business changes shape. Your reps have real conversations instead of rejection. Your mail hits roofs that can buy. Your gas and postage land on doors with a reason. And you stop renting your next job from a lead site or waiting on a storm for it — you find it on your own streets, where it was the whole time.

That is the work RoofPredict exists to make scalable: a read on which roofs are due, house by house, with the storm modeled on each one — so you knock and mail the right doors and skip the rest. If you want to see whether the read holds up, the fairest test is to hand it a roof you already know and check the call yourself.

FAQ

How can I tell how old a roof is without climbing on it?

You estimate a range, not an exact date, by stacking signals. From aerial imagery you read granule loss, color fade, streaking, surface unevenness, patches, and any distress. From a drive-by you add gutter granule wash-out, curling or cupping edges, bald mat spots, worn flashing, and sagging lines. Year built from the assessor is the worst signal because it misses every re-roof. The most reliable scalable read is the roof's own surface from above, which sorts homes into believable age bands far better than the house's birthday.

Why is year built a bad way to target homeowners who need a roof?

Year built tells you when the house was constructed, not when the roof was last replaced. A house from the 1990s may be on its third roof; a newer house may have been re-roofed after a hail event. Re-roofs are invisible in the year-built field, so any list built on it pays to reach homes that already have a newer roof and cannot buy. Read the actual roof condition instead, supplemented by permit records where they exist.

What's wrong with using a hail map to find storm-damage jobs?

A hail map shows where hail was reported, not which roofs it actually damaged. Damage to a specific roof depends on stone size at that exact spot, wind direction during the storm, slope orientation, and the roof's age and brittleness. So a polygon over-counts newer roofs on sheltered slopes and under-counts older roofs at the swath's edge. Modeling the storm on each individual roof, paired with that roof's age, points your crew at the homes most likely worth inspecting instead of the whole blob.

Is door knocking or direct mail better for finding these homeowners?

They serve different categories. Door knocking has the highest conversion when the route is sorted by due roofs, because you can get on the roof the same day, but a bad list burns reps out fast. Direct mail is the workhorse for age-out roofs that canvassing reaches slowly, and its economics depend entirely on list quality since cost per piece is roughly fixed. Most strong companies run both, matched to the list, plus mining their own old estimates and past customers, which is the cheapest source of all.

How do I find roofing customers without buying leads?

Build your own list instead of renting one. Get an age read on your own old estimates and past customers — those are warm contacts you already paid to acquire, and many of their roofs are now due. Get target neighborhoods read for roof age and storm wear, then canvass the densest streets of due roofs and mail the age-out homes. Stay findable through a strong local search presence and reviews. Bought leads are a thin-pipeline supplement, not a foundation, because the same homeowner is usually sold to several roofers at once.

What should a canvasser say at the door to a homeowner with an old roof?

Lead with a specific, true observation and offer an inspection rather than a sale: something like, 'I work roofs here in the neighborhood — your roof is reading at the older end of its life and I can see some granule loss in the gutters. I'm offering free inspections while I'm out here; it takes about fifteen minutes and I'll show you photos of exactly what's up there, good or bad.' Specificity turns a pitch into a conversation, and offering the free look — with the chance it's good news — is what gets you on the roof.

How is RoofPredict different from EagleView, Roofr, or HOVER?

Those are measurement tools — they measure a roof you have already decided to bid. RoofPredict answers the earlier question: which roof to bid at all. It scans your area and gives each address a roof-age range estimated from aerial imagery plus a per-roof storm read, so you can sort streets by which roofs are due and which a storm wore out. It is a targeting tool, not a measurement tool, and you typically use both at different stages — targeting to choose the house, measurement to size the job.

Does roof-age and storm data tell me a homeowner has insurance-claimable damage?

No, and it's important to be clear about this. Age and storm modeling give you odds — which roofs are most likely worth inspecting — not proof that damage exists. The inspection is what establishes the actual condition of a roof, and only the homeowner's insurer decides coverage. Your role is to document conditions and provide an estimate. Never tell a homeowner their roof is damaged before you've looked, and stay entirely off their claim and deductible; unlicensed adjusting rules are real and enforced.

How accurate is a roof-age estimate from imagery?

It's a range, not a date. Aerial imagery reads the actual surface of the roof — granule loss, fade, streaking, texture, repairs — which sorts roofs into believable bands like new, mid-life, aging, and end-of-life far better than year built. What it won't do is stamp an exact install date, and you should be wary of anyone who claims it can. A tight, defensible range is exactly what you need to prioritize a route; the up-close inspection confirms the rest.

Should I focus on storm work or steady age-based prospecting?

Both, deliberately. Storm work is concentrated and urgent but comes in feast-and-famine bursts with heavy out-of-town competition. Age-out roofs are there storm or no storm and give your company a steady floor of work between events. Companies that work both have year-round volume from aging roofs plus surge capacity for storms — and because they already know which roofs were old and exposed, they capitalize on the right homes when a storm hits instead of grinding the whole polygon.