How to Find Homes With Old Roofs (Without Buying Lead Lists)

The painful truth most roofers eventually accept: the easiest way to find homes with old roofs isn't to buy lead lists. It's to look for them — systematically, in your own service area, before the homeowner files a request with Angi or HomeAdvisor and starts the race-against-five-other-roofers game.

The reason most shops don't do this is execution. "Drive around looking for old roofs" doesn't scale past one rep's daily windshield time. So shops fall back to shared lead marketplaces and accept whatever conversion rate the model produces.

This post is about the four methods that do scale — three of them completely free, the fourth one cheap enough to outperform shared leads on per-customer cost. None of them require buying a list.

Why "buy a list" is structurally broken for roofing

Before we get into the methods, a quick reminder of why we're avoiding the easy path:

• Shared lead marketplaces sell each lead to multiple roofers. The homeowner has been contacted before you finish dialing.
• Exclusive lead marketplaces charge more per lead and aren't always exclusive in practice.
• Bought lead lists (Reonomy, Datafiniti, county auctions) give you raw data but no condition signals — you're guessing which homes actually need work.

Direct prospecting flips this. You decide which homes to pitch before the homeowner enters the marketplace. The roofers I know with the best margins do this consistently — about an hour a day on direct prospecting beats two reps cold-calling shared leads all day. For a deeper look at how this fits together, see how roofing contractors find homeowners with old roofs without door-knocking.

The four methods below stack. None is a complete solution by itself. The roofers who win do 2-3 of them every week.

Method 1: Drive-by + Google Street View

Cost: free. Effort: high. Effectiveness: high if disciplined.

This is the manual baseline every other method tries to scale. The premise: a trained eye can identify a 20+ year old asphalt roof from the sidewalk in about 4 seconds. Two driveways down, a 5-year-old roof is just as obvious.

How to actually do this:

1. Pick a zip code or neighborhood you want to work this month. Drive it once with a clipboard or a notes app — write down 30-50 addresses where the roof looks past its prime.
2. Cross-check in Google Street View when you get back. Modern Street View images are usually 1-2 years old. If the roof looked rough on the drive and the SV image shows the same roof, that's a high-confidence prospect.
3. Look for the four signals:
   • Granule loss — patchy color, especially on south-facing slopes
   • Curl and lift — shingle edges no longer flat
   • Algae streaking — dark vertical streaks
   • Patch-and-repair signatures — different-colored sections

The roofers who get good at this can clear 50 addresses an hour from a clipboard + Street View pass. The hard part is keeping it up week after week.

When this beats every other method: when you have a sales rep with windshield time. The drive itself is the cost — Street View is just verification.

Method 2: County property records (last roof permit lookup)

Cost: free. Effort: medium. Effectiveness: high for permit-cooperative counties.

Every roof replacement in most U.S. counties requires a permit. Those permits are public records. Most counties have a search interface; some don't, but you can usually get the data via a public records request or by visiting the building department in person.

What you're looking for: the date of the last roof permit at each address. Combine that with typical asphalt lifespan (20-25 years), and you have a rough roof age. Plus the absence of a recent permit tells you the roof has likely never been replaced.

How to make this scale:

1. Build a spreadsheet of every address in a zip code (export from county GIS or buy a basic property data dump)
2. Cross-reference roof permits — most counties let you export permits by date range
3. Flag addresses with no roof permit in the last 20+ years OR a permit older than 22 years
4. Use that as your prospect list — these are almost always candidates

When this beats Method 1: when you have a large enough service area that windshield-driving every neighborhood doesn't make sense. Permit data scales geographically.

The downside: some counties have garbage permit data. Some homeowners replace without pulling permits. And the data is always 30-90 days stale.

Method 3: Satellite signals you can spot yourself

Cost: free. Effort: medium. Effectiveness: high for trained eyes.

Google Earth Pro (the desktop app) has historical satellite imagery for most metros going back 10-15 years. You can pull up an address, scroll through the years, and see the roof change over time.

What you can spot from satellite:

• Granule loss appears as color variation between slopes — the worn slopes look lighter
• Algae streaking is dark vertical streaks visible from above
• Missing tabs and tarps are very visible — bright blue tarps especially
• Neighborhood replacement clusters — when 3-5 houses in a 400ft radius all have fresh roof colors, you're looking at a cascade that hasn't finished

This is where the AI tools (next method) get most of their value — they automate this exact visual check. But you can do it manually for high-value prospects.

A practical workflow:

1. From your Method 2 spreadsheet (homes with 20+ year old roofs), filter for high-value zips
2. Pull each address in Google Earth Pro, scroll back 5-10 years, then come forward
3. Anything where the roof has visibly degraded in the last few years → priority prospect
4. Anything with a visible tarp, missing shingle section, or 3+ neighbors with fresh roofs → drop everything and knock today

When this method shines: post-storm. Two weeks after a hail event, the satellite imagery will catch the damaged roofs before insurance has caught up. Speed matters, and the US hail map is a good way to see which storms actually crossed your area.

Method 4: AI-screened prospect lists

Cost: low (subscription). Effort: low. Effectiveness: highest of the four.

This is what Methods 1-3 collapse into when automated. AI-powered satellite analysis tools scan every roof in a service area, score each one against the visual signals roofers care about (the same signals you'd look for in Method 3), and surface the highest-priority candidates.

What modern AI roof-scoring tools detect:

• Granule loss patterns (mid-stage vs. advanced)
• Shingle curl and lifting
• Algae streaking severity
• Missing tabs and visible tarps
• Roof age estimation band (within ±3-5 years for asphalt)
• Neighborhood replacement cascade effect
• Hail-impact bruise signatures (post-storm)
• Patch repair signatures (high-intent signal — homeowner already aware)

The best of them rank each roof against its immediate neighbours instead of handing you a number nobody can repeat at a door. A roof only means something next to the roofs either side of it, and "everyone on this street went new except you" is the strongest social proof in the trade.

Leads also carry two separate scores. NEED is how badly the roof needs replacing, condition and roof age only. NOW is whether anything is forcing a decision on this specific house: visible active damage like a tarp or missing shingles, hail damage on this roof, a purchase in the last 18 months, patching already attempted, or neighbours who have re-roofed. Neighbourhood hail deliberately does not raise NOW, because every home in the scan shares it, it belongs in the pitch, not the ranking. Ranking is free, but a ranked roof shows no address or owner. The full property record and an owner-equity estimate come with the roof read (1 credit per 8 roofs); the DNC-scrubbed phone and email come with the owner lookup (1 credit per 2 roofs), or open the house (1 credit) for everything. Where hail is on record, it also shows roughly how long the owner has left to file a claim, based on the typical state notice deadline, an estimate to verify with the carrier, not a legal deadline.

The economics beat any shared-lead source because you're approaching the homeowner before they're in shopping mode. You're not racing six other roofers — you're the first person at their door with a credible reason for being there. If you want the structural comparison rather than a rate card, the pay-per-lead comparison page lays out who each model actually suits.

Roofbird does this automatically — scans every roof in your service area from satellite and street-level imagery, ranks each roof against its neighbours, and gives you the candidates worth knocking. See a live scored area without signing up. The free trial includes 10 credits, no card, no expiry.

And here's where Roofbird goes further: the owner lookup (1 credit per 2 roofs) gives you the homeowner's contact details (owner name, phone, email, mailing address, each phone DNC-scrubbed, plus owner-occupied vs absentee status), or open the house (1 credit) for everything. No separate skip-tracing tool. You get a complete prospect profile ready for a manual, DNC-compliant call or direct mail piece. No cold texting or robocalling, just a clean data set you can act on immediately. The skip-tracing page explains exactly how the DNC screening works.

Combining the four into a weekly prospecting routine

The roofers I know who do this best run a weekly routine that takes 5-7 hours total but generates 30-50 high-quality knock candidates:

Monday (1 hour): Pull this week's AI-flagged list (Method 4). Cross-reference top 100 against your service area map.

Tuesday (1 hour): Spot-check top 20 in Google Earth and Street View. Drop anything that's obviously not what the AI thought it was (commercial buildings, multi-unit complexes, newer-than-flagged roofs).

Wednesday (1 hour): Pull county permit records (Method 2) for the surviving 80 addresses. Cross-reference. Drop anything with a recent permit.

Thursday + Friday (3 hours): Drive-by + door-knock the top 30. Leave door hangers at every house regardless of whether someone answers.

Saturday morning (1 hour): Review the week's responses, schedule callbacks and inspections.

If you'd rather skip the manual zip-canvassing step entirely, AI canvassing lets you search the area you work and get told which blocks are worth your morning before you draw anything. For each block it reports the housing era, what share of roofs read as weathered, how many homes are actually there (counted from building footprints, not estimated from area), whether hail has hit in the last 12 months, and how walkable the block is — scattered acreage on long driveways gets called out as bad canvassing even under a perfect storm, because a morning is measured in doors per hour. When nothing in view clears the bar it says so plainly rather than offering three neighbourhoods you should skip. That's the honest version of "where should I knock today," and no lead marketplace can do it, because they don't know your territory — they just sell you whoever filled in a form.

The output of the full routine: 30 door knocks → typically 6-10 inspections scheduled → 1-3 jobs closed. That math beats any shared-lead source by a wide margin and is the reason this is the channel I'd build a roofing business around.

What to do this week

If you've never done structured direct prospecting before, the lowest-friction starting point:

1. Pick one zip code in your service area
2. Run Method 1 (drive-by + Street View) on that zip this Saturday — 2 hours, no cost
3. Make a list of 30 candidates
4. Knock 10 of them next week
5. Track the response rate

If the response rate is better than your shared-lead conversion rate, the case is made. If it isn't, the diagnostic is in either your script (try our door knocking script post) or your service area — both are fixable. And if you want to see how Roofbird scores a roof from imagery before you commit, how it works walks through the whole read.

FAQ

Q: Does Roofbird provide the homeowner's phone number and email so I can call them directly?

A: Yes. The owner lookup (1 credit per 2 roofs) gives you the owner's name, phone, email, and mailing address, and opening a house (1 credit) gives you everything. Each phone number is DNC-scrubbed and labelled, "clear" means it was checked against the federal Do Not Call registry and is safe for a manual sales call, "DNC" means don't call it, and "verify" means the scrub couldn't confirm either way, so treat it as unknown. The system also flags owner-occupied vs. absentee properties. You get a complete contact profile with no separate skip-tracing tool needed. Use this data for manual, DNC-compliant calls or direct mail. Roofbird does not support cold texting or robocalling, and the contact details are intended solely for one-to-one outreach. A failed lookup is refunded; a nobody-listed answer is charged.

Q: How does Roofbird decide which blocks I should scan first?

A: Search the area you work — a town, a suburb, a ZIP, or a street — and Roofbird reads the neighbourhoods in view and ranks them, so your first scan isn't a guess. For each block it reports the housing era, what share of roofs read as weathered, how many homes are actually there (counted from building footprints, not estimated from area), whether hail has hit in the last 12 months, and how walkable the block is. Click a suggestion and it draws the area for you; you can still drag the corners or draw your own polygon from scratch. Scanning new ground is free at any size, so you can look before you spend anything.

---

Want to skip the manual zip-canvassing step? Roofbird scans every roof in your service area and gives you the candidates ranked by need and by what's forcing a decision now. Free trial with 10 credits — no card.

— Jake