Every roofing company has the same complaint: not enough good leads. Usually the problem is not volume, it is that nobody knows which channel actually produces sold jobs at what cost. This guide walks every meaningful lead channel, what each really costs, and the follow-up math that decides whether any of them pay.
The only metric that matters: cost per sold job
Cost per lead flatters bad channels and hides good ones. A referral costs nothing and closes half the time; a shared internet lead costs real money and closes rarely. Divide what you spend on a channel by the number of jobs it actually sold and rank everything by that number. You cannot do this from memory, which is why lead source tracking belongs in your CRM and not in your head.
The channels, ranked honestly
1. Referrals and repeat customers
Highest close rate, lowest cost, chronically under-worked. Referrals do not just happen; they get asked for. The mechanics: do clean work, ask at the moment of peak happiness (final walkthrough, not invoice day), make it effortless to refer, and thank people visibly. A review-and-referral engine that triggers automatically after job completion turns this from a personality trait into a system.
2. Storm response and canvassing
In restoration markets, the storm is the season. The shops that win it move fast and knock smart: verified hail and wind data tells you which streets were actually hit, canvassing maps keep crews from overlapping, and every knock gets logged with an outcome. RoofGrid pulls NOAA storm data and turns a verified swath into a canvassing territory, so the truck goes where the hail went instead of where the driver guessed.
3. Google Business Profile and reviews
When a homeowner searches "roofer near me," the map pack wins the click. Ranking there is mostly three things: review count, review recency and review replies. Fifty recent, replied-to reviews beat a decade-old profile with eighty stale ones. Ask for the review the day the job closes, by text, with the direct link. This channel is free and compounds.
4. Your website with an instant-quote widget
A website that says "call for a free estimate" converts a fraction of its visitors. A website that answers the question the visitor actually has ("roughly what will this cost?") converts multiples of that. An embedded instant roof estimate widget gives the visitor a ballpark from aerial measurement in seconds and gives you their address and contact info as a warm lead with the measurement already attached.
5. Local SEO and content
Slower but durable: service-area pages, before/after photos, straight answers to the questions homeowners search. Costs time instead of money, and unlike ads it does not stop producing the day you stop paying.
6. Paid search and social ads
Works when the math is watched weekly and the landing page gives the visitor something (an instant estimate beats a contact form). Dangerous when set-and-forgotten: storm-market CPCs spike exactly when everyone wants leads, which is exactly when discipline matters.
7. Purchased leads
Shared leads are a speed contest you pay to enter. They can fill gaps in a slow month, but treat them as perishable inventory: answer in minutes, track close rate per source, and cut any source that does not produce sold jobs at an acceptable cost. Never build the business on them.
Follow-up is where leads actually die
Most shops do not have a lead problem, they have a leak problem. The lead came in, someone meant to call, the week happened. Three fixes, in order of impact:
- Speed to lead: minutes, not hours. First real conversation usually wins.
- A pipeline, not an inbox:every lead lands on a board with a stage and an owner, so "waiting on us" is visible instead of silent.
- Automatic follow-up sequences: the polite second and third touch that humans forget and software does not. Most jobs close on a follow-up, not the first contact.
Common questions
What is a good cost per lead for roofing?
It varies wildly by channel and market, which is why cost per lead is the wrong number to manage. Track cost per SOLD JOB instead: a $150 purchased lead that closes 1 in 10 costs $1,500 per job, while a $0 referral that closes 1 in 2 costs a gift card. Rank your channels by cost per job and the budget allocates itself.
Are purchased roofing leads worth it?
Sometimes, as a supplement, never as the plan. Shared leads sell to three or more contractors, so you are paying for a race: whoever calls first wins a disproportionate share. If you buy leads, answer inside five minutes and track close rate per source ruthlessly. The shops that lose money on purchased leads are the ones that never measured.
Does door knocking still work for roofing?
After a verified storm, yes, it is still one of the highest-ROI channels in restoration. The difference between canvassing and wandering is data: knock the streets the hail swath actually hit, log every knock so nobody hits the same door twice, and hand the homeowner something concrete like an instant roof measurement instead of a pitch.
How fast should I respond to a roofing lead?
Minutes, not hours. Speed to lead is the single highest-leverage variable in residential sales: the first contractor to have a real conversation usually wins the inspection slot, and the inspection slot usually wins the job. This is a systems problem: leads must land somewhere that alerts a human immediately, not an inbox that gets checked at dinner.