The challenge
Sterling runs eleven roofing and exteriors crews across a fast-growing metro, and their paid search account was doing exactly what it was told to — generating leads at a steady, growing volume, every month, on target. The problem showed up downstream: quote requests sat for a day or more before anyone called back, crews were scheduled against jobs that fell through because the homeowner had already hired someone else, and the close rate had been sliding for two straight quarters while the marketing report kept saying the funnel was healthy. Nobody had connected the ad account to the operations calendar, so the campaign kept optimising for the one number it could see.
What we did
Found the real constraint before touching the ads
A short capacity audit showed Sterling could physically quote and close around 40 jobs a week at full crew utilisation. The account was generating 85. The gap wasn't a sales problem — it was pure volume overrun, and every lead beyond capacity was either going stale or displacing a better-fit one.
Rebuilt the qualification layer before the ad click
A short pre-quote form — roof age, damage type, timeline, insurance claim status — sat between the ad and the phone call, so a crew's time went to jobs that were actually ready to move rather than research-stage browsing. This is normally a conversion-rate cost; here it was the entire point.
Paced spend to crew capacity, not to a flat monthly budget
Campaigns now throttle up when a crew is between jobs and pull back when the schedule is full, instead of spending the same amount every day regardless of whether anyone is available to quote the work. Media spend became a scheduling lever, not a fixed line item.
Cut response time from a day to under two hours
The stale-quote problem was solved by automated instant scheduling for pre-qualified leads rather than a manual callback queue — homeowners who have already decided to get a quote lose interest fast, and speed to first contact turned out to matter more than any ad creative change tested.
The result
Lead volume fell 52% on purpose and close rate rose 64%, because the leads remaining were the ones a crew could actually reach while the homeowner was still deciding. Revenue per crew-week rose 38% with no change in headcount — the same eleven crews closing more of what they quoted. Media spend dropped alongside lead volume, so cost per booked job fell even further than the headline numbers suggest.
“We spent a year proud of our lead numbers while our close rate quietly fell apart, and nobody thought to ask whether we could actually service what we were buying. Cutting lead volume on purpose felt insane until the crew-week numbers came in.”