A good cost per lead for a roofing company in 2026 depends entirely on the channel — and the raw number matters less than what it costs you per booked job. Here's the range, channel by channel, and the math to check your own.

A good cost per lead for a roofing company in 2026 centers on roughly $90.92 — the 2026 average across the Home & Home Improvement category on Google and Microsoft Ads — though your real number depends heavily on channel mix and market competition, with non-branded search consistently the most expensive way to reach a stranger. The number that actually determines profitability isn't cost per lead — it's cost per booked job, which also depends on how well each channel's leads close.
Across Google and Microsoft Ads, the Home & Home Improvement category — the bucket roofing sits in — averaged a cost per lead of $90.92 in 2026, on an 8.05% conversion rate and an $8.33 cost per click. That's a blended average across a wide category, and it hides real variation by campaign type — branded search, non-branded search, Local Services Ads, and Performance Max all price differently, which the next section breaks down. Pull your own cost per lead by campaign type before assuming the category average applies to your account.
The spread comes down to how much buying intent a channel captures before it charges you. Branded search only bills for people already searching your company by name, so it's typically the cheapest channel available. Non-branded search costs meaningfully more, because you're bidding against every other roofer for a stranger's attention before they've decided who to call. Local Services Ads tend to sit closer to branded search on price, since Google only charges for verified leads rather than clicks that never turn into a contact. Encouragingly, cost per lead fell industry-wide in 2026: the all-industry average across Google and Microsoft Ads dropped to $66.69, the first decline in five years. A roofing company still budgeting off 2025 numbers is likely overpaying regardless of channel mix.
On Google Ads specifically, expect your roofing account to land above the $90.92 Home & Home Improvement average once non-branded, high-competition search makes up most of your spend. Demand moves with the season, and cost per lead tends to track it: contractor sentiment in Q1 2026 leaned positive, with 36% of roofing contractors reporting rising inquiries against just 21% reporting a decline — more roofers competing for the same reroofing demand pushes the bid needed to win the click higher. Track your own account's cost per lead month over month rather than assuming a fixed seasonal pattern; local weather events move this market more than the calendar does.
Generally yes, because of how the two products charge you. Local Services Ads bill per verified lead — a phone call or message Google confirms is real — rather than per click, so you're not paying for traffic that never engages. Standard Google Search Ads charge per click regardless of outcome, so part of every non-branded budget goes to visitors who never reach out. That's why operators who run both typically see Local Services Ads produce a lower cost per lead, though the size of the gap varies by market and Google doesn't publish a roofing-specific benchmark. The trade-off is control — Local Services Ads limit how much you can customize targeting and creative. Most roofing companies run both: Local Services Ads for lower-cost volume, Google Search Ads for control and scale. See the full Local Services Ads vs. Google Ads breakdown for how that trade-off changes once Local Services Ads moves into Performance Max in 2026.
Cost per lead only tells half the story, because cheaper channels often close at lower rates too. Here's an illustrative model — not a benchmark, just the math — showing how the comparison can flip once you follow it to a booked job.
Say a roofing company splits a $6,000 monthly budget evenly between two illustrative channels — Channel A at $150 per lead and Channel B at $75 per lead. At $150 per lead, $3,000 in Channel A buys about 20 leads; closing at an illustrative 25% — reasonable for high-intent search traffic — that's roughly 5 booked jobs, or about $600 per booked job. At $75 per lead, the same $3,000 in Channel B buys about 40 leads; closing at an illustrative 15% — lower, since some of those leads are still comparison shopping — that's roughly 6 booked jobs, or about $500 per booked job.
Channel B's raw cost per lead is 50% lower than Channel A's. Its cost per booked job is only about 17% lower once close rate is factored in. Run this math against your own numbers before shifting budget — a cheap lead that converts poorly can cost more per job than an expensive one that doesn't.
There's no universal percentage-of-revenue rule that fits every roofing company's margin structure, so anchor the budget to the job-value math above. Against a national average roof replacement of $9,606, a cost per booked job of $500 to $600 is roughly 5.2% to 6.2% of ticket value — well inside an acceptable customer acquisition cost. The bigger risk in 2026 isn't overspending; it's under-investing while demand climbs. The industry is now worth $92.5 billion and growing at 5.0% a year, and 36% of contractors already see inquiries rising. Set your budget from your own booked-job math, then scale it as close rate improves.
For the strategy layer above channel-by-channel cost per lead, see our guide to building a profitable paid media strategy, and what changed in roofing marketing in 2026 for why storm-chasing alone stopped filling a pipeline. If roofing is one of several trades you run, the 2026 home services field guide covers the forces reshaping HVAC, plumbing and roofing alike, and what's the smartest way to generate HVAC leads in 2026? breaks down the same category's budget allocation in more depth. The same cost-per-lead-versus-total-transaction-cost logic applies well outside home services too — see cost per lead for mortgage lenders: 2026 benchmarks. And more broadly: businesses that invest in paid media grow three times faster than those relying on organic alone.
WordStream's 2026 benchmarks put a good cost per lead for a roofing company in 2026 at roughly $90.92, the average for the Home & Home Improvement category on Google and Microsoft Ads. Your actual number will move well above or below that depending on channel — branded search and Local Services Ads tend to run cheaper, non-branded search runs more expensive. Judge any number against cost per booked job, not cost per lead alone. The full cross-industry table lives in our marketing benchmarks index.
Non-branded Google Ads bids against every roofer targeting the same buyer intent, while branded search only charges for people already searching your company by name, and Local Services Ads only charge for a Google-verified lead. The gap reflects targeting precision and how much buying intent a channel captures before it bills you, not lead quality on its own.
Cost per lead is ad spend divided by leads generated; cost per booked job divides spend by jobs actually sold, which accounts for close rate. A channel with a low cost per lead but a poor close rate can produce a higher cost per booked job than a pricier channel that converts well.
It tracks demand closely. The NRCA reported 36% of roofing contractors saw customer inquiries increase in Q1 2026 against 21% reporting a decline, and cost per lead typically rises when more roofers compete for the same seasonal reroofing demand. Budget extra for periods when contractor sentiment — and competition for clicks — is on the rise.
There's no fixed percentage that fits every roofing company, since margins and job size vary by market. A more reliable approach is calculating cost per booked job against your average ticket — Angi puts the average roof replacement at $9,606, which can support a cost per booked job of $500 to $600 for most operators.
Performance advertising for B2B and niche B2C brands — turning paid media into a predictable revenue channel.
Follow Refinex Media on LinkedIn.
Thirty minutes. We look at your spend and tell you where the return is hiding — you keep the findings either way.