Satellite Imagery for Roofing: The Complete Guide

Every "AI roofing" tool is downstream of one decision: which imagery does it read, and how old is that imagery? The provider determines what the tool can actually see. Different sources vary in resolution by 3-5x and in freshness by years. This post is the breakdown of which source does what, so you can ask the right questions when evaluating any aerial roofing tool.

The four imagery sources (and what each is good for)

Google Maps Static API

The most widely-used source in the AI roofing space because it's the cheapest.

• Resolution: 5-15cm/pixel at maximum zoom (zoom 20)
• Freshness: Wildly variable — 6-24 months typical in major US metros, 2-4 years in suburban and rural areas
• Coverage: Global, US-comprehensive
• Pricing: Priced per thousand image requests, which is why nearly every consumer-scale tool can afford it

What Google gets right:

• Coverage breadth (you can pull imagery for any US address)
• Cost (most AI tools can't afford anything else)
• Reasonable resolution for residential prospecting

What Google gets wrong:

• Freshness is unpredictable — you might be looking at the property's PREVIOUS owner's roof
• Resolution drops in rural areas
• No timestamp on imagery (vendors usually don't know when the image was captured)

Best for: consumer AI tools doing residential prospecting at scale.

Nearmap

The premium alternative most enterprise roofing tools use.

• Resolution: 5-7cm/pixel, sometimes finer
• Freshness: 3-6 month refresh in major US metros (much faster than Google)
• Coverage: US plus a few international markets, but with gaps in rural areas
• Pricing: Enterprise contracts, and Nearmap does not publish a rate card — you get a quote after a call, and it scales with the coverage area and refresh frequency you want

What Nearmap gets right:

• Best-in-class freshness for major US metros
• Consistent quality across the coverage area
• Available as direct API for AI tools to integrate

What Nearmap gets wrong:

• The contract shape makes it impractical for small-shop AI tools
• Coverage gaps in rural and smaller-market areas
• Smaller-shop AI tools that claim "Nearmap powered" often only have it for major metros

Best for: mid-to-large roofing contractors using premium measurement tools.

EagleView / Vexcel aerial imagery

The premium-of-premium tier. Uses planes (not satellites) for highest possible resolution.

• Resolution: under 5cm/pixel — near-survey quality
• Freshness: Variable, often custom-flight-on-demand for large jobs
• Coverage: US-comprehensive with regular flight schedules over major metros
• Pricing: Per-property reports or annual subscriptions plus per-property fees. The bill is shaped per report, which is fine for a job you already won and painful for a list you haven't qualified yet.

What EagleView gets right:

• Insurance-accepted reports
• Measurement accuracy under 1% error
• Custom flights for high-value insurance work

What EagleView gets wrong:

• The per-report shape prohibits scaled prospecting — you can't economically score 500 candidate homes when every report is billed
• Designed for measurement-at-a-known-property, not prospect identification

Best for: insurance restoration, commercial work, post-storm damage documentation.

Drone imagery (on-demand)

Not a "provider" in the same sense, but increasingly used by AI tools for verification.

• Resolution: 1-2cm/pixel typical, finer with specialized equipment
• Freshness: Real-time (flight on request)
• Coverage: Anywhere you can legally fly
• Pricing: Per-flight, contracted. You pay for the flight whether or not the roof turns out to be worth knocking.

What drone gets right:

• Highest resolution available
• Oblique angles satellite can't capture
• On-demand timing (no waiting for satellite refresh)

What drone gets wrong:

• Cost-per-property prohibits scaled prospecting
• Regulatory constraints (FAA Part 107 license, no-fly zones)
• Weather-dependent

Best for: verifying high-priority AI-flagged properties before in-person inspection.

What resolution actually means for roof assessment

Resolution determines what the AI can detect. Practical breakdown:

Resolution	What's detectable	Use case
under 5cm/pixel	Granule loss patterns, individual missing shingles, fine cracking, hail bruise depth	Insurance-grade documentation
5-10cm/pixel	Material classification, damage signal detection, age estimation, condition scoring	Direct prospecting
10-15cm/pixel	Material type, large damage, tarp detection, structural features	Initial screening
15-25cm/pixel	Footprint, slope, basic structural features	Measurement only
over 25cm/pixel	Roof presence/absence	Useless for roofing

Most prospecting AI tools operate at 5-15cm at zoom 20 (Google range). Premium tools operate at finer resolution.

If a vendor doesn't tell you their resolution, ask. If they can't answer, treat the tool as Google-based until proven otherwise. And if you want to see how the two cameras combine in practice, how Roofbird scores a roof from imagery walks through it on real addresses.

The freshness problem (most important variable)

Imagery freshness is the single most important variable nobody asks about.

Why it matters:

If a tool is scoring a roof from a 24-month-old image, you're potentially looking at the previous owner's roof. The current homeowner might have replaced it 18 months ago. You knock, embarrass yourself, lose credibility.

Typical freshness by region (Google-based tools):

• Major US metros (DFW, Phoenix, LA, NYC): 6-18 months
• Mid-sized metros: 12-24 months
• Suburban: 18-30 months
• Rural: 24-48 months

Typical freshness for premium-imagery tools:

• Nearmap in covered metros: 3-6 months
• EagleView in active flight schedules: 6-12 months
• Custom drone: real-time

The question to ask every vendor:

"What's the median image age in [my top zip code]?"

If they can't answer, they're not thinking about freshness — which means you'll be working with whatever Google has. Sometimes fine, sometimes a year and a half old.

There's a second picture most tools ignore entirely: the street-level shot. On about 85% of roofs the street-level imagery is newer than the aerial, by roughly five years, and a contractor in Missoula found that out the hard way — nine doors off his own list, eight of them already re-roofed, because his block had been ranked from a 2016 aerial while a 2025 street shot of the same houses sat unread. The roofs hadn't changed since we read them. The picture had. That's the case for reading both cameras, top-down and ground-level, into one assessment, which is what scoring a roof from Street View or satellite is actually about.

When to trust the imagery, when to verify on ground

The rule of thumb:

Trust the AI screening when:

• The property is in a major metro with recent refresh
• The signal is high-contrast (visible tarp, missing shingle section)
• You're using AI for prospect prioritization, not insurance documentation

Verify on ground before action when:

• Imagery is older than 12 months in your area
• The signal is borderline (mild granule loss, possible curl)
• You're documenting for an insurance claim
• The property is high-value and worth the inspection time anyway

The screening plus verification combo is faster and more accurate than either alone. The honest version of this is a tool that tells you which roofs it has actually seen from the road and which it genuinely cannot judge — behind tree canopy, set too far back from the road — instead of passing off a nine-year-old photograph as current. Every shortlist should be checked against the newer picture before you see it, anything already done moved down and said out loud, and where the aerial and the street photo disagree, the newer one wins and both dates get shown.

Manual vs AI-assisted satellite review

Manual review of satellite imagery (the Google Earth Pro approach) still has a place. It's slow — 5-10 addresses per hour — but it catches signals AI sometimes misses. Most useful for:

• Verifying AI-flagged borderline cases
• Reviewing historical changes (Google Earth Pro has timeline functionality)
• High-value insurance claims where you want human-confirmed signals

AI-assisted review is faster (every property in a service area in minutes) but isn't perfect. The best workflow uses AI for the 95% screening and manual for the 5% verification.

Where to point the tool in the first place

Freshness and resolution tell you what a tool can see. They don't tell you where to look, and that's the question that actually costs a morning. Search the area you work — a town, a suburb, a ZIP, a street — and a good tool reads the neighborhoods in view and ranks them before you draw anything. For each block it should report 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 is bad canvassing even under a perfect storm, because a morning is measured in doors per hour. And when nothing in view clears the bar, it should say so plainly rather than offering you three neighborhoods to skip.

That's the part no imagery provider sells you directly, and it's why the satellite versus drone versus Street View comparison for roof prospecting is worth reading before you buy anything. If you want to see a scored area without signing up, there's a live one you can look at.

Cost models by use case

The pricing models tools use depend on what they're built for:

Per-report measurement (EagleView, HOVER):

• Best for: low-volume measurement on already-qualified prospects
• Billed per property, so it scales poorly to prospecting — you pay for the report whether or not the roof was worth knocking

Subscription with included scoring:

• Best for: high-volume prospecting across a service area
• Billed monthly, whether or not you work the list that month

Enterprise imagery contracts (Nearmap direct, EagleView annual):

• Best for: large multi-state contractors needing imagery API access
• Quoted per territory, and the vendor does not publish a rate card — expect a sales call and a coverage commitment

Credits, spent only when you act (Roofbird):

• Best for: prospecting where you want to look for free and pay only when you open a house or mail a card
• Free and unlimited: asking, searching, sorting, saving lists, and scanning new ground at any size
• Credits buy the work: reading a roof properly (both photographs plus the full property record, 1 credit per 8 roofs), the owner's name, phone and email at 7/8 of a credit a roof, postcards at 3 credits a card for 4x6 or 6x9 and 4 for the big 6x11, and opening a house completely for 1 credit
• One-off packs, no subscription: Starter $49 for 49 credits, Crew $99 for 125, Hunter $199 for 300, Pro $399 for 800. Bought credits never expire and nothing renews.

Match the model to your use case. Don't buy per-report pricing for prospecting; don't buy a subscription for occasional one-off measurements. If you want the arithmetic laid out plainly, here's what a credit actually buys.

Use cases by shop size

1-5 employees:

• Google-based AI prospecting with a free-to-search tool
• Per-report measurement only when needed for a quote
• Skip enterprise imagery entirely

5-15 employees:

• Prospecting tool plus measurement reports
• Consider drone services for high-value insurance work

15-50 employees:

• Multiple tools layered (prospecting plus measurement)
• Begin drone integration for storm work

50+ employees:

• Enterprise imagery contracts with Nearmap or similar
• Premium per-report tier for measurement
• Internal drone team or contracted services
• Custom AI integration

What to ask any "satellite imagery" vendor

A 5-question evaluation framework:

1. "Which imagery do you read?" (Google, Nearmap, EagleView, drone, custom?)
2. "What's the median image age in zip [your top zip]?" (If they can't answer, treat as Google-based)
3. "What resolution does your AI run against?" (Should be 5-15cm or finer for condition assessment)
4. "Have you published accuracy validation data?" (If yes, read it with skepticism; if no, treat the output as heuristic)
5. "Can I see sample output on properties I already know?" (Always pull 3-5 samples on known properties before signing up)

Vendors who answer all 5 confidently are usually the ones worth using. Two more worth asking: does it read the street-level photo as well as the aerial, and does it tell you which roofs it could not judge? A tool that answers those honestly is telling you something about how it handles everything else.

Roofbird's DFW sample dashboard lets you run this evaluation in real-time on actual DFW properties. The unlocked leads show exactly what a dual-camera read at 5-10cm resolution produces. Free trial in your own service area, no card required.

— Jake

FAQ

Q: How does Roofbird handle the freshness problem?

A: Roofbird reads Google-based aerial imagery and ground-level Street View together, and it treats the street-level photo as the newer source, which it is on about 85% of roofs. Every shortlist is checked against the newer picture before you see it, anything already done is moved down and said out loud, and houses that genuinely cannot be judged — behind tree canopy, set too far back from the road — are labelled as exactly that rather than mixed in with the confident ones. Where the two disagree, the newer one wins and both dates are shown. It doesn't pretend the aerial is current. It tells you what it has actually seen and what it hasn't.

Q: Can I trust Roofbird's condition assessment from just satellite imagery?

A: Roofbird combines the top-down overhead tile and a ground-level Street View of the façade into one assessment. The Street View read catches side-visible wear a top-down tile can't: granule loss, algae streaking, curling shingles, patched sections, rusted flashing. It also compares the roof against its immediate neighbors — if the houses either side have visibly been re-roofed and this one hasn't, that's flagged, because a street where everyone has already gone new is the strongest social proof in the trade. Alongside the ranking sits a plain-English verdict in a roofer's words, the specific reasons behind it, the roof's size in squares, how many planes it has and how steep it is. If you want the methodology in more depth, roof age estimation from aerial imagery covers how age is read off the shingles rather than the building's year built.

Q: How does Roofbird decide which homes are worth my time?

A: Every lead carries two scores. NEED is how badly the roof needs replacing — condition and roof age only. NOW is whether anything is forcing a decision on THIS house: visible active damage like a tarp or missing shingles, hail damage on this specific roof, a purchase in the last 18 months, patching already attempted, or neighbors who have re-roofed. Neighborhood hail deliberately does NOT raise the NOW score, because every home in the scan shares it, so it can't separate one door from the next. That separation keeps the ranking honest: most homes in a settled neighborhood have no trigger, and Roofbird says so rather than inventing urgency to look busy.

Q: Where does the homeowner's contact information come from, and can I call it?

A: The full property record shows on every lead for free as soon as it's scanned — owner name, whether they live there or it's an absentee owner, estimated market value, year built, last sale date and price, beds and baths, living square footage, lot size, stories, garage, annual property tax, mortgage lender, and an owner-equity estimate, because equity is what decides whether someone can say yes. Only the phone and email sit behind a one-click unlock. Every 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, verify means the scrub couldn't confirm either way. Manual dialling only, no texts and no auto-dialler. A lookup that fails is refunded; a definitive "nobody listed" answer is charged once and cached so the same address is never charged again. Homeowner contact details and DNC screening explains the whole chain.

Q: Where hail is on record, does Roofbird tell me anything about the claim window?

A: It shows roughly how long the homeowner has left to file, based on the typical notice deadline in that state. Treat it as an estimate of the typical deadline and verify with the carrier before you say anything at a door — it is not a legal deadline and Roofbird does not guarantee a claim will be approved. Neighborhood hail still doesn't raise the NOW score. It belongs in the pitch, not the ranking.