Read the roofing number with its fine print intact
LocaliQ's report covers 3,211 U.S. Search campaigns from April 1, 2024 through March 21, 2025. It combines Google and Microsoft advertising and calls its technical medians "averages." Roofing & Gutters is listed at $10.70 CPC and $228.15 CPL. LocaliQ reported those from its own past customer campaigns. They are not a price every roofer will pay.
The category includes gutters and does not isolate your exact roof system, market or retail offer. Its July 2026 page-update date is not evidence that campaigns ran in 2026. These limits are important when a salesperson presents a precise number as the amount your replacement-estimate campaign should pay.
The benchmark can still be useful as context for a test. It warns against assuming that roofing inquiries will be inexpensive, and it cannot set your allowable acquisition cost. Your inspection process, close rate, contribution, and sales lag decide whether a campaign supports the business. If your cost per lead is higher than the sample, that is a question to investigate, not a final verdict.
Choose the inspection you want to sell
A homeowner with an active leak, one planning a replacement and one asking about a commercial membrane are not interchangeable prospects. The inspection, expertise and proposal process differ. Decide which work the campaign buys before choosing keywords or a budget.
For planned residential replacement, explain what the assessment establishes and the actual next step. For repair, avoid implying that every roof can be fixed without inspection. For commercial work, consider access, decision-makers and a longer procurement process. A generic "free roof" message can attract confusion the office must later unwind.
Keep the offer honest. Before you mention insurance coverage, claim approval, waived obligations, or a financing result, you need the authority and the terms in writing. Have the relevant professionals review regulated claims. The ad should invite an honest assessment. The insurer and the lender decide on their own.
| Demand | Appropriate conversion path | Main reporting risk |
|---|---|---|
| Leak repair | Qualified contact to assessment and repair | Small scope mixed with replacement value |
| Retail replacement | Assessment to proposal to accepted scope | Longer decisions counted as immediate losses |
| Storm-related request | Safe assessment and documented scope | Unsupported insurance expectations |
| Commercial roofing | Access and purchasing qualification | Procurement lag hidden by short reports |
| Gutter-only work | Supported scope and estimate | Category benchmark mistaken for reroof economics |
Worked example: follow fifty inquiries through the funnel
Here is a planning cohort, not a LocalPro result. $10,000 of media spend generates fifty inquiries, thirty qualified inspection appointments, twenty-four attended inspections, twelve proposals, and four completed jobs. The raw inquiry cost is $200, attended-inspection cost is about $416.67, and media acquisition per completed job is $2,500.
Suppose each completed job contributes an assumed $5,000 after direct production costs. Four jobs contribute $20,000 before acquisition, leaving $10,000 after the stated media spend and before other overhead. If actual contribution is $3,000 each, only $2,000 remains after media. The same platform result has a very different economic meaning.
Inspect the largest drop before prescribing more spend. If scheduled inspections do not attend, improve confirmation and qualification. If attended assessments rarely produce proposals, investigate scope fit and estimating. If proposals stall, examine follow-up and value communication. "The CPL is high" does not identify which step is broken.
Use lead cohorts so the calendar stays honest
A sale completed in November may come from a September inquiry. Dividing November spend by all November sales mixes acquisition periods and can fake a trend. Keep the original lead date, proposal date, acceptance, and completed-job date connected in your CRM.
Mark recent cohorts incomplete, and compare them with earlier cohorts at a similar age as jobs move forward. That matters for planned replacements, financing decisions, and commercial procurement, and it shows which old proposals need an owner.
Do not claim causation merely because revenue rose after a campaign launched. Weather, referrals, backlog and capacity can change the same month. Preserve the evidence you have, explain attribution limits and distinguish projected pipeline from completed work. A forecast can inform planning without being presented as earned revenue.
Query control: remove wrong intent without blocking buyers
Employment, training, wholesale supplies, and DIY instructions are common exclusions for a homeowner-service campaign. A person asking about replacement cost or shingle options may still be a buyer, so read the search and the conversation before you block a whole theme.
Use service-specific pages for repair, replacement or supported systems where those distinctions matter. Explain inspection scope and avoid fabricated project photos or testimonials. If the business does not do a particular roof type, do not advertise it simply because the click appears cheaper.
Look at brand intent on purpose. An existing customer who searches the business name is different from someone searching the category for the first time. Competitor-related searches can carry extra cost and legal questions. Write the choice down for each group. A dashboard that mixes easy branded contacts with new replacement jobs will overstate how well you are acquiring new work.
- ☐ Separate repair and replacement intent where meaningful
- ☐ Review employment and supply-only searches
- ☐ Retain cost research when it fits the offer
- ☐ Exclude unsupported roof systems thoughtfully
- ☐ Check brand versus new-demand reporting
- ☐ Match the page to the promised assessment
- ☐ Validate forms, call routing and source capture
A storm budget needs more than a weather alert
Before increasing spend during a demand event, confirm safe inspection capacity, geographic coverage and a clear intake policy. A company can run out of estimator appointments long before it runs out of advertising impressions. Do not keep promising prompt assessment when the backlog makes that promise untrue.
Route urgent safety concerns appropriately and avoid technical instructions that encourage a homeowner onto a roof. Ask for information that can be provided safely. Claims about damage, coverage or required replacement must follow qualified assessment; an AI intake summary does not establish any of them.
Preserve the event period in reporting. Storm-related demand can differ from ordinary retail work in scope, documentation and timing. Do not use a short surge's apparent conversion rate as the permanent budget assumption for the rest of the year. Review completed contribution after the event has matured.
Teach bidding the milestone you can confirm
Google's qualified-lead and converted-lead definitions support deeper offline events. Choose a milestone the business can confirm reliably and use consistent stage definitions. A form fill, an attended inspection and a completed sale should not all become indistinguishable primary successes merely because they are easy to track.
Check that duplicate contacts are merged, and check the corrections that show up later. The same homeowner may call, submit a form, and reply to a proposal. Those can be one opportunity. Upload only data you can support, through a workflow someone approved, and keep the privacy and consent steps in place. Hashing the fields does not clear those obligations.
For AI Max or PMax tests, inspect which campaign product and controls apply. Standard cross-channel PMax is different from a local pay-per-lead campaign migrated from LSA. Broader automation can amplify weak event definitions. Repair measurement before interpreting an inexpensive automated conversion as evidence of a valuable inspection.
The monthly review should end with a production-aware decision
Reconcile spend, qualified inspections, attendance, proposals, accepted scopes and completed contribution. Show unresolved pipeline and data gaps. Include the costs of estimation, management and creative consistently when discussing total acquisition, while keeping media-only metrics labeled separately.
Then decide whether to expand a work family, improve a handoff or stop an unfit offer. Do not increase every campaign because one large sale made a small sample look excellent. Check whether the next additional job would fit estimator and installation capacity and whether the pricing still supports it.
A useful roofing PPC budget stays tied to your inspection pipeline, with the historical benchmark as context. Buy work you can assess and complete, described the way the crew does it, at a contribution the business can support.