Doubling ad spend without the cost per lead going up

A local patio and outdoor living contractor came to me for Google Ads management, already running ads but not sure the leads coming in were worth what they were paying. Before touching the ad spend, I connected CallRail so every lead got tracked back to an actual phone call, not just a click.

Here's what changed over the next 13 weeks:

Normally, doubling a budget that fast drives the cost per lead up, you run out of the cheap, high-intent searches first and start paying more for the rest. Here, the cost per lead actually went down slightly while lead volume doubled.

What made that possible

• Cut the junk searches before they cost anything. Added negative keywords to block people searching for DIY guides, generic pricing questions, and roofing-cost comparisons, plus a separate list blocking clicks from people specifically searching for named competitors.
• Stopped spreading the budget thin. Paused two underperforming test campaigns and put the full budget behind the one that was already converting, doubling its daily budget.
• Tightened how keywords were matched. Swapped broad-match keywords for exact-match versions of the same terms, so ads only showed for the specific searches that were actually converting, and capped how much was bid on each.
• Rewrote the ads around local trust. New headlines led with decades of local experience and a limited-time discount offer, instead of generic "get a quote" language.

The pattern holds for most local contractors: the budget isn't the problem, where it's going is. Fix that first, and scaling stops being risky.