How to Find Homeowners With Old Roofs Who Haven't Been Pitched Yet

Every roofer has felt it. You drive a street you have worked for years, knock a few doors out of habit, and a competitor's yard sign is already staked in the lawn of the one house you should have hit first. The roof was twenty years old. You knew it was old. You just got there late.

The problem is almost never that there are no old roofs near you. In most metro and suburban markets, somewhere between fifteen and thirty percent of single-family homes are sitting on a roof past the practical replacement point for an asphalt shingle. The problem is that you cannot see roof age from the curb on most homes, the good prospects are scattered one or two per block instead of clustered, and the handful you can spot are the same ones every other crew in town can spot. So you end up fighting over the obvious roofs and walking past the worn-out ones that nobody has touched.

This is a sorting problem, not a volume problem. The homeowners you want share three traits: their roof is genuinely near end of life, they have a reason to act soon (age, a storm, or both), and no one has gotten in front of them yet. Find the overlap of those three and you are talking to people who need you and have heard from nobody. Below is the operational playbook for doing that, from free public-records work you can do tonight to imagery-based age signals to the canvassing and mail mechanics that turn a list into signed contracts. No fluff, real workflows, and the edge cases that trip up crews who try this without a system.

What "old roof" actually means before you build a list

If you are going to target by age, you have to be honest about what age you can know and what you cannot. Three numbers get confused constantly, and confusing them is how roofers waste a season chasing the wrong houses.

Year the house was built. This is what Zillow, the county assessor, and most public data show you. It is a hard fact and it is free. It is also the weakest proxy for roof age, because any roof installed in the last 20 to 30 years has likely been replaced at least once. A 1968 ranch may be on its third roof. A 2015 build is almost certainly on its original roof. Year built tells you the ceiling on roof age, not the roof age.

Date of the last reroof. This is the number you actually want, and it is the hardest to get. Permits capture some of it (more on that below), but a large share of residential reroofs are done without a pulled permit, especially smaller jobs and storm work, so permit absence does not prove the roof is original. Where a permit exists, you have gold: a dated, address-specific reroof.

Estimated roof age from imagery. Modern aerial and satellite imagery, compared across years, lets you estimate how long the current roof surface has been on the house and what condition it is in. This does not give you an exact install date. It gives you a range, and a range is enough to sort a street. A roof estimated at 18 to 22 years old is a knock; a roof estimated at 4 to 8 years old is a skip. You are not trying to be a county recorder. You are trying to rank doors.

The practical takeaway: build your list from the strongest signal you can get per address, fall back to weaker signals where the strong one is missing, and never treat year-built as roof age. The sections that follow go strongest to weakest, and then show how to stack them.

Why "old" is a range, not a date, and why that is fine

Asphalt three-tab and architectural shingles do not fail on a fixed clock. A south-facing slope in Phoenix bakes; the north slope on the same house can last years longer. Ventilation, attic heat, the quality of the original install, and storm history all move the number. Two roofs installed the same week can be five years apart in real remaining life.

So the right mental model is a probability, not a deadline. "This roof is 18 to 22 years old and has taken two hail events" is a far better targeting signal than "this roof was installed on March 4, 2004," even though the second sounds more precise, because the second number is usually unknowable and the first is enough to decide whether to knock. Build your whole program around ranges and odds. It keeps you honest with homeowners, and it stops you from skipping a worn-out roof just because you could not find an exact date.

Tier 1: Public records you can mine for free this week

Before you pay for anything, there is real signal sitting in public data. It is tedious to assemble by hand, which is exactly why most of your competitors have not done it. That tedium is your moat for the first few weeks.

County assessor and parcel data

Nearly every county publishes parcel records: owner name, mailing address (which tells you owner-occupied versus rental or absentee), year built, square footage, and often the last sale date and price. Most counties have a searchable property portal; many publish bulk parcel files or have an open-data GIS portal. Some let you export.

What to pull and why it matters:

• Year built — your roof-age ceiling. Filter to homes old enough that an original roof would be due, but remember reroofs hide here.
• Owner-occupied flag — compare the situs (property) address to the mailing address. Same address means owner lives there, which is who you want for a retail or insurance-restoration job. A different mailing address flags a rental or an out-of-area owner; those convert differently and some you will skip.
• Last sale date — a home that sold 8 to 12 years ago, was built decades earlier, and shows no recent reroof permit is a strong original-roof candidate. New buyers also reroof more readily.
• Assessed value and square footage — rough job-size and affordability sorting.

Building permits: the single most underused free signal

Permit data is the closest thing to a free reroof date that exists, and most roofers never touch it. A pulled reroof permit gives you an address and a date. That does two things at once: it tells you which roofs are new (skip them) and, by absence over a long-enough window on an old house, hints at which roofs may still be original.

Where to get it: many jurisdictions post permit records on their building-department site or an open-data portal (Socrata-style city data portals are common in larger cities). Search permit type for "reroof," "roof," "roofing," or "residential reroof." Pull a multi-year window.

How to use it, concretely:

1. Pull every reroof permit in your target ZIPs for the last 12 to 15 years.
2. Tag those addresses as recently reroofed — do not pitch. This list alone saves real money in mail and gas.
3. Take your old-housing-stock list (year built old enough to be due) and subtract the recently-permitted addresses.
4. What remains is your strongest free candidate pool: old houses with no recent reroof on record.

The honest caveat, stated plainly so you do not get burned: permit data has holes. A meaningful share of reroofs never get permitted, so a missing permit does not guarantee an old roof. Treat "no permit on an old house" as a positive signal, not proof. You will still misfire on some doors that were quietly reroofed; you confirm at the curb or on the ladder. The point of the list is to raise your hit rate, not to make every door perfect.

MLS and recent-sale data

If you have an agent relationship or access to listing history, old listing photos and remarks frequently mention roof age or condition ("newer roof," "roof replaced 2019," "roof at end of life — priced accordingly"). A home that sold years ago with a roof already called "aging" in the listing is now years older. This is slower to mine, but on higher-ticket homes it is worth the dig.

A second angle most roofers miss: pending and recently closed sales create a window. A buyer who just closed on a 1990s home, where the inspection report flagged the roof, is primed to act in the first year of ownership. Inspection reports are not public, but listing remarks and a quick conversation with the buyer's agent often surface the same information. If you can build a referral relationship with two or three agents who sell in your target neighborhoods, they will hand you the aging-roof homes before you ever knock, because a clean roof estimate helps them close the sale.

Tax and homestead records as a tenure signal

Homestead exemptions and long stretches of stable ownership tell you a homeowner has been in the house long enough to be on the original roof and long enough to have the equity and intent to fix it. A homeowner of fifteen years in a home built twenty-five years ago, with no reroof permit, is a textbook original-roof prospect. Pair that tenure read with the assessor's year-built and you have sharpened the list before you have touched any imagery.

A note on doing this legally and cleanly

Public records are public, but how you contact people off them is regulated. Mail is generally the most permissive channel. Phone and text fall under federal and state telemarketing and do-not-call rules, so if you call or text off a list, scrub against the National Do Not Call Registry and follow your state's rules and the FTC's Telemarketing Sales Rule. Door-knocking is governed by local solicitation ordinances and posted no-soliciting signage. None of this stops you from using public-records targeting; it just shapes which channel you use to reach each address. Know your rules before you dial.

Tier 2: Reading roof age and condition from imagery

Public records get you old houses. Imagery gets you old roofs — the actual surface, and whether it has been replaced since the house was built. This is the step that separates a year-built mailing list from a genuine due-roof list.

What you can see from the sky, and what you cannot

With decent aerial or satellite imagery, an experienced eye can read a surprising amount: granule loss showing up as color mottling and shine differences, patched or mismatched slopes, moss and streaking on north-facing aspects, sagging ridgelines, and obvious damage. Crucially, if you compare imagery from different years, you can often see when a roof changed color and texture — the signature of a reroof. A slope that was dark and uniform in a 2016 image and worn and streaked today has aged in place; a slope that went from worn to crisp-and-new between two image dates was reroofed, and you can bracket roughly when.

What you cannot reliably get from imagery: the exact shingle product, the precise install date, the number of layers, or hidden deck rot. Imagery is an age-and-condition estimate, not an inspection. Anyone who tells you they can read exact roof age or material off a satellite is overselling. The realistic output is a defensible range plus a condition read, which is exactly what you need to rank a street.

Free and low-cost ways to do this by hand

You can do a version of this for free, address by address, today:

• Aerial map services with a historical imagery slider let you scrub a single roof across years and watch for the reroof color-change or the slow aging.
• Street-level imagery sometimes catches a roof edge or a low slope from the road, and date-stamped street images give you another time anchor.

The limitation is obvious: this is one-address-at-a-time work. It is excellent for qualifying a hot prospect or prepping a specific street before you walk it. It does not scale to scanning ten thousand homes to find the few hundred worth your time. For that you need the data done in bulk.

Curbside condition cues that confirm the imagery

Once imagery flags a roof as old, a thirty-second look from the truck confirms or kills it before you commit a canvasser's afternoon. Train your reps to read these from the street: granule loss showing as bald, shiny patches and dark streaks; cupped, curled, or clawing shingle tabs throwing irregular shadows in low sun; a roofline that dips or waves where the deck has started to give; missing or lifted tabs along ridges and rakes after a wind event; and dark algae streaking down north-facing slopes, which signals an older surface even when it is not itself a defect. None of these is proof of failure, and none of them is a claim. They are the cheap confirmation step between a data flag and a ladder.

A practical rule for crews: imagery and records decide which streets and which doors; the curb decides the order you knock them once you are there. The two together are far stronger than either alone, and they keep you from knocking a roof the data missed as freshly reroofed.

Tier 3: Storm exposure — the second axis that finds the worn-out ones

Age alone is half the picture. The other half is what the weather has actually done to each roof. A 14-year-old roof that has sat through two real hail events and a high-wind day can be in worse shape than a 20-year-old roof in a calm pocket. If you only sort by age, you miss the storm-aged roofs; if you only chase storms, you miss the ones simply aging out. The homeowners worth the most are where those two axes overlap.

Where the public storm data lives

You can pull genuine storm history for free, and you should know how, both to target and to keep yourself honest:

• NOAA's Storm Prediction Center publishes storm reports including hail and wind, with sizes and locations.
• The National Weather Service and NOAA's storm events database give you a searchable, dated record of severe weather by county and date.
• Hail-swath and wind products show roughly where hail fell and how big, by date.

These sources tell you a storm passed through an area on a date and roughly how severe it was. That is real and useful. What they do not tell you, on their own, is which specific roof got hit hard enough to matter, because hail and wind are wildly uneven street to street and roof to roof. A hail map shows you where it hailed. It does not show you which roofs it wore out.

The gap between "it hailed here" and "this roof is damaged"

This is the single biggest mistake in storm-driven targeting: treating a hail polygon as a list of damaged homes. Two houses across the street from each other can have completely different damage from the same storm because of roof pitch, orientation to the storm's path, shingle age and brittleness, and the simple physics of how a given hailstone hit a given slope. An old, brittle roof takes damage from hail that a fresh roof shrugs off. Aspect matters: the slope facing into the wind-driven hail gets hammered while the lee slope is fine.

So the useful question is not "did it hail in this ZIP" — it nearly always did somewhere — but "given this roof's age, pitch, orientation, and the actual storm that passed over it, how likely is real damage?" That is a per-roof physics question, and it is where age and storm exposure finally combine into a single rank.

Stacking the signals: building one ranked list instead of three

The individual signals are useful. Stacked, they are decisive. Here is the model that consistently produces a list crews actually work instead of argue about.

Score each address on the signals you have. A simple, transparent point system beats a black box you cannot explain to your canvassers:

The hard excludes matter as much as the positives. Most wasted spend in roofing outreach goes to roofs that do not need replacing. Cutting the obvious-new roofs out of your mailing and walking lists is often the fastest profit improvement available, before you add a single new prospect.

A worked example

Say you target three ZIPs with roughly 18,000 single-family homes. A realistic funnel:

• Start: 18,000 homes.
• Filter to year-built old enough to be due, owner-occupied: down to about 9,000.
• Subtract recent reroof permits: removes maybe 1,500, leaving 7,500.
• Apply imagery age estimate, keep 15+ years: down to roughly 3,000 genuinely old roofs.
• Overlay storm exposure, flag the 16+ year roofs that also took two or more events: about 900 "age plus storm" priority homes.

You have just turned 18,000 cold addresses into 900 doors where the roof is genuinely worn out and the homeowner has a real reason to act — and, if you got there before the swarm, nobody has pitched them. That 900 is a season of work for a crew, and you mail and walk it in priority order instead of blanketing 18,000.

Run the cost the other way to see why the sort matters. A blanket mail drop to all 18,000 at a dollar a piece is eighteen thousand dollars, most of it landing on new roofs and renters. The same eighteen thousand dollars spread across your 900 priority homes funds twenty touches each, or a tighter design with a follow-up wave, hitting only people who could plausibly buy. Same budget, a completely different return, because the spend is no longer diluted by roofs that did not need you. The exclusions did as much work as the targeting.

Keep the scoring transparent

Resist the urge to over-engineer the score. A canvasser who can see why a home ranks high — "19-year roof, two hail events, owner lives there" — works the list with conviction and can answer the homeowner honestly. A black-box score nobody understands gets second-guessed in the field and quietly ignored. Start with the simple weighted table, run it for a few weeks, and let your own conversion data tell you which signals actually predict signed jobs in your market. In a calm market, age and tenure will carry most of the weight; in a hail belt, storm exposure will. Tune to where you work.

Where RoofPredict fits

Everything above is doable by hand. The catch is that the highest-value version of it — imagery-based roof-age estimation across every home, combined with per-roof storm modeling, refreshed and ranked — is exactly the part that does not scale one address at a time. That is the gap RoofPredict is built for.

RoofPredict scans the roofs in an area and gives you, house by house, a roof-age range estimated from aerial imagery plus a storm read modeled on each individual roof, rather than the ZIP it sits in. Instead of "it hailed in this area," it models the hail and wind against each roof's pitch and orientation and scores the likelihood that that roof was worn out, then pairs it with the age estimate. The output is the ranked street you would have built by hand if you had unlimited time: which roofs are due, in priority order, with the new ones already filtered out.

It is not a lead service and it does not sell you the same homeowner it sold five competitors. It does not hand you a contact who filled out a form. It sharpens the outbound you already do — your knock list, your mail, and your own old customer book — by telling you which doors are worth the gas and which to skip. You can also feed it your existing list or CRM and have it enriched with roof-age and storm signals, so the past estimates and old customers sitting in your book get re-sorted by who is now actually due.

The honest limits, because a tight trade compares notes: roof age comes back as a range, not an exact install date, because that is what imagery can truthfully support. The storm read is odds, not proof of damage — it tells you which roofs likely qualify for a closer look, never that a specific roof is damaged or that a claim will be approved. You still confirm at the curb and on the ladder. What it removes is the part that never scaled by hand: looking at every roof, every storm, and ranking them so your crew spends the day on the worn-out houses instead of the whole street.

Getting there first: working the list before the swarm

A perfect list is worthless if you knock it the same week as everyone else. The whole advantage of finding the un-pitched homeowner is timing. Two situations:

Aging-out roofs (no recent storm). This is the steady, controllable pipeline most roofers neglect because there is no event forcing urgency. There is also almost no competition, because nobody else is working it. A 19-year-old roof in a calm market has likely never been pitched, because the only roofers in that neighborhood are chasing storms elsewhere. Work this list patiently, year-round, and you own it. This is your hedge against feast-or-famine storm cycles.

Storm-aged roofs (recent event). Here speed matters, but not in the reckless storm-chaser way. Out-of-town crews flood a hailed area and blanket every door. Your edge is precision: while they knock the whole ZIP, you walk straight to the old, hard-hit roofs that are most likely to have real damage, with the new roofs already filtered out. You have a better conversation at fewer doors and you are not competing for the homeowner whose three-year-old roof a chaser is trying to talk into a claim.

The aging-out list deserves more attention than it gets, because it is the part of the business a roofer can actually control. A roofer who depends on storms lives feast-or-famine: a fat season when hail hits, a starvation season when it does not, and an out-of-town swarm fighting them for every door when it does. A roofer who works their own streets' aging roofs year-round, and their own old customer book, builds a pipeline that does not care whether it hailed. The thesis is simple: own your next job instead of renting it from a lead site or waiting on a storm for it. The aging-out list is how you own it. It is slower to start, because there is no event manufacturing urgency, but it compounds, it has almost no competition, and it is yours. Treat storm work as the spike on top of that baseline, not the whole business.

Turning the list into knocked doors and signed jobs

A ranked list changes how you canvass, mail, and re-work your own book. Here is the field mechanics.

Door-knocking the right doors

Walk in priority order, not geographic order. Hit the highest-scored roofs on a street first; if you only get through half a block before dark, you got the half that matters. Carry the per-home reason with you — "your roof's in the range where they start failing, and this area took hail in [month]" — so a green canvasser sounds like a fifteen-year vet without having been on the roof.

The opener that works on an age-targeted door is specific and low-pressure: you noticed the roof looks like it is getting up there in age, you are working the neighborhood, and you offer to take a look and document the condition so they know where they stand. You are not promising anything. You are offering information about their own house. That is a very different conversation from the generic "we're doing roofs in the area" knock, and it converts because it is true and specific to them.

A few field habits separate crews that close these doors from crews that burn them. Leave a branded door hanger or a one-page condition note on every no-answer, because half your priority homes will not answer the first knock and the leave-behind earns the callback. Knock the same priority block two or three times across different days and times before you write it off; a Tuesday-afternoon no-answer is a Saturday-morning yes. Log every door's outcome against its score so you learn which score bands convert in your market and stop wasting reps on bands that do not. And give a green canvasser the per-home talking point in their pocket so they walk up sounding like they already know the house, because they do — the roof age and storm history are right there. That single thing, a new hire knocking the right door with a real reason, is what turns rep churn around: reps who knock worn-out roofs close, make money, and stay.

Direct mail to old roofs

Mail is the most forgiving channel legally and the easiest to target off public records. The economics only work when the list is tight: a piece to an old, owner-occupied, storm-exposed roof is a real prospect; the same piece to a three-year-old roof is money in the trash. Lead the piece with the specific reason ("homes in your area built around [era] are reaching the age where the original roof typically needs replacing") rather than a generic discount. Send in waves to your top-scored addresses and track which scores convert so you can refine weighting.

Re-working your own CRM and old estimates

The cheapest old roofs you will ever find are already in your book. Every estimate you wrote and lost three, five, seven years ago was on a roof that is now three, five, seven years older — and many of those homeowners said "not yet" for reasons that have since changed. Pull your old declined estimates and past customers, re-sort them by current roof age and any storms that have passed over them since, and you have a warm, already-pitched-by-you list that no competitor can see. This is the highest-conversion list in the building and most roofers never touch it.

Work it in tiers. First, the estimates you wrote on a roof that was already borderline and is now several years deeper into its decline — those homeowners knew you, got a number, and stalled; the stall reason has usually expired. Second, past repair customers, where you fixed a leak or a section years ago on a roof that was already aging; the patch bought time that is now running out. Third, anyone in your book whose home has since taken a storm you can date. A short, honest re-touch — "we did your estimate back in 2021, your roof's a few years older now and your area saw hail this spring, want me to come re-check it?" — outperforms any cold list you can buy, because the trust is already there. The reason most roofers skip this is simply that the book is messy and nobody has re-sorted it by current age. Sorting it is the whole job.

The legally safe way to talk about storms, damage, and insurance

If any part of your old-roof program touches storm damage and insurance — and for storm-exposed roofs it will — you have to stay on the right side of a bright legal line, because crossing it is unlicensed public adjusting in most states and it can end a roofer's business. Here is the safe frame, stated as the compliance rule it is.

What you may do. You may inspect a roof, thoroughly document its condition with dated photos, and prepare an accurate, itemized repair estimate for the work you would perform — ideally aligned to the standard estimating line items carriers use. You may state facts about your own scope of work to the homeowner and, where appropriate, to the carrier about your own scope. You hand that documentation and estimate to the homeowner. The homeowner files their own claim. The insurer decides coverage. You are the documenter and the contractor, full stop.

What you may not do, for a fee. Do not negotiate, adjust, or "handle" the claim for the homeowner. Do not interpret their policy or tell them what is and is not covered. Do not promise a specific payout, approval, or that a claim will go through. Do not tell a homeowner their deductible will be waived, absorbed, eaten, or disappear — that is illegal in many states and it is a fast way to lose your license and get sued. Do not advertise a "free roof." Do not represent the homeowner against their insurer. Any of those, done for compensation, is acting as an unlicensed public adjuster.

The do-not-say list, for your canvassers

Teach your crew these phrases as the things to never say at the door, because a green rep will reach for them naturally and they create real legal exposure:

• "We'll get your roof approved."
• "We'll get your deductible waived" / "You won't pay your deductible" / "We'll eat the deductible."
• "You're getting a free roof."
• "We'll handle the insurance company for you" / "We'll deal with your adjuster."
• "That's definitely covered" / "Your policy covers this."
• "We'll make sure you get a full replacement paid for."

The compliant version of every one of those is the same move: "We'll document the condition thoroughly and write you an accurate estimate. You file the claim, and your insurer decides. We'll be here to do the work if it's approved." That captures the exact same homeowner intent — "is my storm-worn roof something I can do something about?" — without you ever stepping into claims handling. RoofPredict's role in this is upstream of all of it: which roofs are likely worth documenting (age plus storm), never anything about claims, coverage, or payout.

A documentation workflow that stays on the safe side

The documentation itself is where you add real value without crossing the line, so do it well. A thorough, defensible roof file on a storm-exposed home looks like this:

1. Date and identify everything. Capture the address, the date of inspection, and the date of the storm event you can tie to public records. A photo with no date and no context is worth little later.
2. Wide-to-tight photos. Shoot the full elevation of each slope, then move in to the specific conditions: granule loss, fractured or bruised shingles, lifted or creased tabs, damaged metals, and collateral marks on soft metals like vents, gutters, and downspouts that corroborate hail size. Collateral is often the clearest evidence a storm of a given severity passed.
3. Mark and measure. Use chalk circles and a reference object for scale so the photos read clearly. Note slope, pitch, and orientation, since aspect explains why one slope is hit and another is clean.
4. Write the estimate to standard line items. Build an itemized repair estimate using the line-item conventions carriers recognize, priced to your real cost to do the work. Keep it to your scope of work and the facts of the roof. Hand it to the homeowner as their document.
5. Hand it over and step back. The homeowner files. The insurer decides coverage. You are available to do the work if it is approved. You never tell them what is covered, never promise an outcome, and never touch the deductible.

That file makes you the most useful, most credible contractor the homeowner talked to, and every piece of it sits firmly on the documentation-and-estimate side of the line. The targeting upstream (age plus storm) decides whose roof is worth building that file for in the first place.

Edge cases and what pros get wrong

Treating year built as roof age. Covered above, and it is the number-one error. A list filtered only on old year-built is half new roofs in disguise. Always layer a reroof-exclusion (permits) and an imagery age estimate on top.

Chasing the whole hail polygon. Mailing or knocking every house in a hail swath burns money on roofs that were too new or too sheltered to be damaged. Sort the swath by roof age and per-roof exposure first.

Ignoring absentee owners — and over-ignoring them. Rentals convert differently and many you will skip for retail. But a long-held rental with a 20-year-old roof and an out-of-state owner is sometimes the easiest sale you will make, because the owner has no emotional attachment and just wants the asset maintained. Flag them separately, do not blanket-delete them.

Forgetting north slopes and aspect. Moss, streaking, and the worst hail damage cluster on specific slopes. A roof that looks fine from the street side can be failing on the back. When you qualify from imagery or at the curb, look at all aspects before you write a roof off as fine.

Letting the list go stale. Roofs age, storms hit, and permits get pulled continuously. A list built last spring is wrong by this spring. Refresh roof-age estimates, re-pull permits to remove newly reroofed homes, and overlay any storms that have passed. A living list compounds; a static one decays.

Skipping a worn roof because you could not find an exact date. You will never have a perfect install date for most homes. If the range says old and the roof looks worn, knock it. Demanding certainty before you act means walking past the exact homeowners you are trying to find.

A 30-day plan to stand up an old-roof targeting program

Week 1 — Pull the free data. Export parcel data for your top ZIPs (year built, owner-occupied flag, last sale). Pull 12 to 15 years of reroof permits. Pull storm history (NOAA storm events) for the same area. You now have raw material with zero spend.

Week 2 — Build the exclusion and the base list. Subtract recently permitted reroofs. Filter to owner-occupied, old-enough year built. You have a base list of likely-old roofs with the obvious-new ones removed.

Week 3 — Add the roof-age and storm layer. Either qualify your hottest streets by hand with historical imagery, or run the area through bulk roof-age estimation and per-roof storm modeling so every home gets a range and a storm read, then rank. Pull the top "age plus storm" homes to the front.

Week 4 — Work it and measure. Mail your top addresses in waves, walk the top-scored doors in priority order, and pull your old declined estimates and past customers to re-sort by current age. Track which score bands convert. Feed that back into your weighting. Refresh monthly.

The bottom line

The homeowners with old roofs who haven't been pitched are not hiding. They are scattered one or two per block, invisible from the curb, mixed in with new roofs, and overlooked by every crew that only chases the obvious or waits on a storm. Finding them is a sorting problem with a clear solution: layer year built, reroof permits, imagery-based roof age, and per-roof storm exposure into one ranked list, exclude the new roofs ruthlessly, and work the survivors in priority order before anyone else gets there.

Do it by hand and the tedium is your advantage, because almost nobody else will. Do it at scale — every roof, every storm, ranked and refreshed — and you trade the season of grunt work for a living list of due roofs you can mail, knock, and re-work year-round. Either way, the move is the same: stop working the whole street, and go straight to the roofs that are actually worn out. That is where the un-pitched homeowner has been the whole time.

If you want the at-scale version — a ranked, house-by-house read on which roofs in your area are due, with roof age as a range and the storm modeled on each individual roof — that is what RoofPredict does. Hand us a roof you already know the age of and judge the call for yourself; that is the honest way to find out whether the rank holds up on your streets.

FAQ

How can I find out the age of a roof by address for free?

Stack three free signals. County assessor or parcel data gives you year built (the ceiling on roof age). Building-permit records show dated reroofs, so you can spot which roofs are new and, by absence on an old house, which may still be original. Historical aerial imagery lets you scrub a single roof across years to see when it changed color and texture, which brackets roughly when it was last replaced. None of these gives an exact install date, but together they give you a defensible age range, which is all you need to decide whether to pitch the home.

Is the year a house was built a good way to find old roofs?

Only as a starting filter. Year built tells you the oldest the roof could be, not how old it is. Most homes built more than 20 to 30 years ago have been reroofed at least once, so an old-year-built list is full of new roofs hiding in plain sight. Always layer a reroof-permit exclusion and an imagery-based age estimate on top of year built before you mail or knock, or you will waste a large share of your spend on roofs that were replaced recently.

What is the best way to find homeowners with old roofs nobody else has pitched?

Target the overlap of three things: a roof genuinely near end of life, a real reason to act soon (age, a storm, or both), and no prior outreach. In practice that means aging-out roofs in calm markets that storm chasers ignore, plus old, hard-hit roofs in storm areas that the swarm blankets but does not target precisely. Build a ranked list from roof-age estimates and per-roof storm exposure, exclude new roofs, and work it before competitors arrive. The un-pitched homeowner is usually the worn-out roof scattered one or two per block, not the obvious one everyone can see.

Can you really tell roof age from satellite or aerial imagery?

You can estimate it, as a range, not pin an exact install date. By comparing imagery from different years you can see granule loss, streaking, patches, and the color-and-texture change that marks a reroof, which lets you bracket roughly how long the current surface has been on the house and read its condition. What imagery cannot reliably give you is the exact shingle product, the precise install date, the number of layers, or hidden deck damage. Anyone claiming exact age or material from a satellite is overselling. A range plus a condition read is enough to rank a street.

How do I use storm data to find damaged roofs without guessing?

Use public NOAA and National Weather Service storm records to confirm which areas took hail or high wind, on which dates, and how severe it was. Then do not treat the whole storm area as damaged homes. Hail and wind are wildly uneven street to street, and an old brittle roof takes damage that a fresh roof shrugs off. Sort the storm area by roof age, pitch, and orientation so you focus on the older, hard-hit roofs most likely to have real damage, and confirm condition at the curb or on the ladder. A hail map shows where it hailed, not which roofs it wore out.

What can a roofer legally say to a homeowner about storm damage and insurance?

You may inspect, thoroughly document the roof condition with dated photos, and prepare an accurate itemized repair estimate for your own scope of work, then hand that documentation to the homeowner. The homeowner files their own claim and the insurer decides coverage. You may not, for a fee, negotiate or handle the claim, interpret their policy or what is covered, promise a payout or approval, tell them their deductible will be waived or absorbed, advertise a free roof, or represent them against their insurer. Those cross into unlicensed public adjusting in most states. Stay strictly on the documentation and estimate side.

Are building permits a reliable way to know which roofs are new?

They are the best free reroof-date signal, but they have holes. A pulled reroof permit reliably tells you a roof is new, so it is excellent for excluding recently replaced homes from your list. The catch is that a meaningful share of reroofs are done without a permit, so a missing permit does not prove a roof is old. Treat a recent permit as a hard exclude and treat no permit on an old house as a positive signal to investigate further, not as proof. You confirm the uncertain ones at the curb.

How is this different from buying roofing leads?

Buying leads means paying for a homeowner who filled out a form, often resold to several competitors at once, so you compete on speed and price for someone who already raised their hand. Targeting old roofs is outbound: you find homeowners who have not been pitched at all, using roof age and storm exposure to decide which of your own doors, mailers, and past customers are worth the effort. It is your list, not a rented one, and the homeowner has heard from nobody. RoofPredict supports the second approach by ranking which roofs are due, not by selling you contacts.

What does RoofPredict actually give me, and what are its limits?

It scans the roofs in an area and returns, house by house, a roof-age range estimated from aerial imagery plus a storm read modeled on each individual roof rather than just the ZIP, then ranks which roofs are due with new roofs filtered out. You can also feed it your own list or CRM to enrich it with roof-age and storm signals. The honest limits: roof age is a range, not an exact install date, and the storm read is odds that a roof likely qualifies for a closer look, never proof of damage or that a claim will be approved. You still confirm on the ladder.

Should I bother contacting absentee or rental property owners with old roofs?

Flag them separately rather than deleting them. Many rentals convert differently than owner-occupied homes and some you will skip for retail work. But a long-held rental with a 20-plus-year-old roof and an out-of-area owner can be one of the easiest sales you make, because the owner has no emotional attachment and simply wants the asset maintained. Compare the property address to the mailing address in parcel data to identify absentee owners, then treat them as their own segment with its own pitch.