How to Find Roofing Customers Without Buying Leads

Every roofer hits the same plateau. You start with referrals and Angi leads. Referrals carry the first year. Angi works in years 1-2 then the per-customer cost climbs. By year 3, the question becomes urgent: how do you find roofing customers without paying $40-80 per shared lead that converts at 3-5%?

One free starting point: the US hail map — find the counties hit this month before anyone sells you the lead.

The answer is four channels stacked. Each works in isolation. The shops that scale past $2M revenue use all four. This post is the playbook.

The "buy leads" plateau most shops hit

Quick context. Shared-lead marketplaces work until they don't. The math shifts under you over 2-3 years:

• Year 1: $30/lead, sold to 3 competitors, 8% close rate → $375 CAC
• Year 2: $50/lead, sold to 5 competitors, 5% close rate → $1,000 CAC
• Year 3: $70/lead, sold to 7 competitors, 3% close rate → $2,333 CAC

The CAC tripled in 24 months. Your job size didn't. Margins compress.

The fix isn't to find a better-quality lead marketplace. It's to add channels with structurally different economics. Four to choose from.

Method 1: Local SEO + Google Business Profile

Cost: $0 (if DIY) or $500-1,500/mo (with an agency) Time to first lead: 60-90 days Per-customer cost: $100-400 once ranking

The infrastructure most shops underinvest in.

The 6 things that move local SEO rankings for roofers:

1. Google Business Profile completeness — every field filled, every product/service category claimed, 30+ photos
2. Review velocity — 1-2 new 5-star reviews per week sustains ranking; 5+ per week climbs ranking
3. Citations consistency — your name/address/phone identical across BBB, Yelp, Yellow Pages, industry directories
4. Service-area pages — separate page on your site for each city you serve, each with unique content
5. Customer review responses — respond to every review within 48 hours
6. Google Posts — weekly updates on your GBP (project completions, storm responses, etc.)

The shops doing all 6 of these consistently rank in the top 3 for "[city] roofer" within 6-9 months. That ranking becomes the largest single source of inbound leads for most mature roofing shops.

Method 2: Google Local Service Ads (LSAs)

Cost: $25-80 per qualified lead (paid only when prospect actually contacts you) Time to first lead: Same week as approval (typically 1-2 weeks to get Google Guaranteed) Per-customer cost: $100-400

The "exclusive leads" channel that's actually exclusive — because Google sends them directly to one verified provider.

To run LSAs you need:

1. Google Guaranteed verification — Google does a background check on your business, license, insurance. 1-2 weeks.
2. Google Business Profile in working order (see Method 1)
3. Review velocity because LSAs rank higher with more reviews

Once running, LSAs typically deliver 5-15 leads per month for a typical residential roofing shop at $30-60 per qualified lead.

Pros:

• Genuinely exclusive (Google sends one prospect to one provider)
• Pre-qualified (homeowner clicked from a search intending to hire)
• Pay-per-lead, not pay-per-impression
• Close rates 30-50% (vs 3-5% for shared marketplaces)

Cons:

• Lead volume is capped by search volume in your area
• "Qualified lead" definitions can be lenient — some Google-billed leads aren't actually buyers
• Requires ongoing review-acquisition discipline to maintain ranking

Method 3: Referral programs

Cost: $50-200 per closed referral (depending on incentive) Time to first lead: Same week as program launch (if you have a customer base) Per-customer cost: $50-200

The cheapest channel for shops with at least 100+ past customers.

A program that works:

"Refer a homeowner. If they get a roof from us, you get $200. They get $200 off their job. No limit on referrals — refer 5 neighbors, get $1,000."

Why it works:

• High close rate (referrals close at 40-60% vs 3-5% for cold leads)
• Self-selecting trust — past customer vouches for you
• Compounds with neighborhood cascades — one referral often leads to 2-3 more on the same block
• Net new spend = $0 until a referral actually closes

The execution:

• Email past customers quarterly with the referral offer
• Include in your invoice email signatures
• Print on every door hanger
• Mention at the end of every job: "If you know anyone..."

Most roofers undermarket their referral program because they don't have a system. The shops with referral programs that actually work treat them as a 6-step funnel, not an afterthought.

Method 4: Direct AI prospecting

Cost: $99-299/month flat subscription Time to first lead: 1-2 weeks (depending on canvassing schedule) Per-customer cost: $50-300

The newest channel and the one with the best per-customer economics for shops with sales reps.

The premise: AI satellite imagery scans every roof in your service area, scores each one, and surfaces the top candidates. You door-knock or direct-mail the top scores.

Workflow:

1. Define service area (4-8 zip codes typical)
2. AI scores every property (~5 min)
3. Filter top 100-300 prospects
4. Plan canvassing route by geographic cluster
5. Door-knock 25-30 per session

Per-customer math for a typical DFW shop:

• 100 doors knocked per week
• 15-20 inspections scheduled
• 3-5 jobs closed
• Subscription cost ~$50/week
• Rep labor ~$500/week
• Total: $550/week → $110-180 per closed customer

That's 5-10x better than HomeAdvisor's $800-2,400 CAC.

Here's the part that changes the game in 2026: when Roofbird scores a roof, it doesn't just hand you an address. Every lead now comes with the homeowner's contact details unlocked — owner name, phone, email, and mailing address. Each phone number is DNC-scrubbed so you're not burning calls on litigations, and you know upfront whether the property is owner-occupied or an absentee landlord. That last part matters more than most roofers realize: absentee owners are often the easiest to close via direct mail because they're not emotionally attached to the roof — they just want the asset protected. Owner-occupied? You lead with the hail map and the inspection offer. Either way, you're not knocking blind anymore. You know who lives there, how to reach them, and whether they're the decision-maker before you ever leave the truck.

Roofbird does exactly this. The free trial loads 25 scored leads in your service area — no card. See the DFW sample dashboard first to see what the output looks like.

Building a 4-channel acquisition mix

The shops that scale past $3M revenue use all four channels in a deliberate mix. Here's the typical split for a mature residential roofer in DFW:

Channel	% of total leads	% of closed customers
Local SEO + GBP	25%	30%
Google LSAs	20%	25%
Referrals	15%	30% (highest close rate)
Direct AI prospecting	35%	12% (lowest close rate but highest volume)
Paid marketplaces (Angi etc)	5%	3% (kept only for storm-event surge)

This mix has structural redundancy. If one channel breaks (LSA rates rise, AI tool quality drops, SEO ranking slips), the others carry the pipeline.

The shops that lose are the ones with 80%+ from one channel — typically Angi or HomeAdvisor — and no backup.

Sequence: how to build this from zero

If you're starting today with 0% of these channels active, the sequence:

Month 1-2:

• Google Business Profile completeness audit
• Set up automated review request flow (every closed job triggers a review ask)
• Direct AI prospecting trial (free 25 leads — start here, cheapest)

Month 2-3:

• Apply for Google Guaranteed
• Start running LSAs once approved
• Launch referral program with email blast to past customers

Month 4-6:

• SEO content build (service-area pages, blog posts)
• Citation cleanup across directories
• Increase AI prospecting cadence (2-3 days/week canvassing)

Month 6-9:

• Local SEO rankings start climbing
• LSAs hit volume stride
• Multi-channel mix becomes self-sustaining

Month 9-12:

• Drop Angi/HomeAdvisor if not contributing
• Optimize each channel's per-customer cost weekly
• Add storm-event surge spend (Methods covered in insurance claim post)

By month 12 most shops doing this systematically have lower CAC than year-1 Angi pricing, with no single-channel risk.

What to do this week

If you've only ever bought leads, the cheapest test is direct AI prospecting:

1. Pull a free trial on Roofbird or similar (25 leads, no card)
2. Block one Saturday this week for canvassing
3. Knock the top 25 leads using the door-knocking script
4. Track responses

If your response rate is better than your shared-lead conversion rate, you've validated the channel. Add LSAs and referrals next month.

FAQ

What's the fastest way to get roofing customers without buying leads? Direct AI prospecting is the fastest — you can have 25 scored leads in your service area within a week and start door-knocking that same Saturday. Local SEO and LSAs take 60-90 days to ramp. Referrals work immediately if you have a past customer base, but they're capped by your existing network size.

How much does it cost to get roofing customers without buying leads? The range is $0-300 per closed customer depending on channel. Referrals run $50-200 per closed job. Direct AI prospecting runs $110-180 per closed customer when you factor subscription plus rep labor. Local SEO costs $0 if DIY but takes months. Compare that to $800-2,400 per customer on shared marketplaces.

Do AI roofing lead tools include homeowner contact information? Yes — Roofbird now unlocks the full contact details on every lead: owner name, phone, email, and mailing address. Every phone number is DNC-scrubbed, and you know whether the property is owner-occupied or absentee. That means you can skip the blind door-knock and go straight to a phone call, email, or direct mail piece tailored to who actually owns the roof.

Can I combine AI prospecting with storm chasing? Absolutely. Use the US hail map to identify counties hit this month, then run an AI scan on those specific zip codes to find the roofs most likely to have damage. The combination of storm timing plus roof scoring is the highest-converting prospecting method most roofers aren't using yet.

— Jake