A roofing contractor was paying for ads it could not measure, so every campaign was learning from the wrong numbers.
The Client and the Problem
A regional roofing replacement contractor was running Facebook ads to bring in estimate requests from homeowners. The ads were live, but the tracking behind them was not reliable: tag and pixel errors meant some form submissions were counted twice, so the campaigns were optimizing toward numbers that were not real.
The creative was thin too, with too few angles to find out what actually moved a homeowner to ask for an estimate. And the budget was fixed at under $3,000, so every fix had to come from spending better, not spending more.
1. Fix the Tracking First
We corrected the Google Tag Manager events and the Facebook Pixel mapping so that each estimate request was recorded once, in the right place.
- Every lead credited to the ad that produced it
- Duplicate form submissions removed
- Clean conversion data, so the ad platform learned from real leads
Every lead counted once, in the right campaign
2. Rebuild the Estimate Page
Homeowners were arriving on a page that made asking for an estimate harder than it needed to be. We:
- Simplified the residential roofing landing page layout
- Made it load faster on phones
- Put trust badges and warranties where visitors could see them
- Restructured the form fields so requesting an estimate took less effort
3. Test Four Creative Angles
We built and scaled four ad angles, each written for a homeowner who is ready to replace a roof:
- Lifestyle videos: satisfied homeowners and finished craftsmanship
- Before and after: the property transformation and the curb appeal upgrade
- Seasonal urgency: storm readiness and limited-time inspection offers
- Local presence: community trust, regional expertise and licensed local crews
The Results
Month over month, with ad spend held under $3,000:
| Metric | Change | What it means |
|---|---|---|
| Leads | +42.9% | More estimate requests from the same budget |
| Conversion rate | +66.1% | More of the people who clicked asked for an estimate |
| Cost per lead | -6.7% | $87.07 average per lead |
| Click-through rate | +21.0% | More people who saw an ad clicked it |
| Unique reach | +18.8% | More homeowners saw the ads |
Key Takeaway
The biggest gain came from fixing what the ads were measuring. Once the platform could see real leads, better creative and an easier estimate page turned the same budget into 42.9% more leads at a lower cost per lead.
More on how we work as a home services marketing agency.
Not sure whether your ads are learning from real leads?