Case study · 60-day engagement
A Las Vegas garage door company had been spending $5,800 a month on Google Ads. Every search term that converted was an emergency repair (broken spring, busted opener, stuck door) but the agency had been treating all garage door queries the same way. The good leads were buried in noise and the after-hours calls were going to voicemail.
The challenge
The owner had been frustrated for a year. Search volume looked good. Click counts looked good. But the booked-job rate per click was way under his industry benchmarks. The reason became obvious 20 minutes into the audit.
What we found:
The plan
Built two separate campaign trees. One for emergency repair (broken spring, stuck door, opener not working, garage door wont close, after hours, weekend). One for installation and consultation (new garage door, panel replacement, smart opener install). Different bids, different ad copy, different landing pages, different conversion goals.
Cranked emergency-campaign bids 35% between 7 PM and 7 AM, and 20% on weekends. The math holds because emergency callers convert at 78% close rate vs about 22% for daytime shoppers. The CPL goes up but the cost per booked job goes way down.
Every emergency-campaign call now routes through CallRail with on-call cell forwarding. We added an SMS auto-responder that fires within 15 seconds of any missed call: "Saw we missed you. We have a tech on call tonight. Reply YES and we will call back in 5 min." About half the missed calls are recovered before they hit a competitor.
147 negatives added in the first pass to kill "diy," "color," "designs," "ideas," "free quote shopping," and a few competitor brand searches that were never converting.
Single page covering the four most-common emergency calls (broken spring, opener failure, broken cable, stuck door), with a sticky click-to-call button and a "we are open right now" indicator that flips to "leave a callback request" after hours.
AI tools we used
We do not pretend the work happens by hand. Three of our internal tools are the reason this engagement moved as fast as it did. Each of them replaces what used to be days of human time.
Watches the clock + day of week, auto-adjusts bids in 2-hour blocks based on historical conversion patterns. Works without us touching the account.
Detects unanswered calls within 15 seconds, fires a personalized SMS asking the caller to confirm callback. Recovers about 41% of missed jobs.
Auto-splits a service business account into emergency + non-emergency campaigns based on keyword analysis, no human structuring required.
The execution
| Week | What we shipped | What moved |
|---|---|---|
| Week 1 | Account audit, intent-split campaign tree built | Two campaigns deployed by day 6 |
| Week 2 | Time-of-day bid boost active, CallRail routing live | After-hours pickup rate climbs to 87% |
| Week 3 | 147 negatives swept, emergency landing page live | CPL on emergency campaign settles at $24 |
| Week 4 | SMS missed-call recovery deployed | Recovers 11 jobs in first 7 days |
| Week 5 | First weekly review with owner | Emergency calls/wk: 4 → 11 |
| Week 8 | Mid-engagement performance check | Emergency calls/wk stable at 14 |
The numbers
The lift in emergency calls broke down to:
The owner stopped chasing more click volume and started chasing the RIGHT clicks. Same total ad budget, completely different revenue mix. Most months he now refuses installation work to keep the emergency capacity open.
I was getting calls. The right ones were just always the ones I missed. Once we fixed the after-hours piece I was working harder on dispatch than I was on marketing. That is the right problem to have.
Keep reading
Thirty minutes. We pull up your account on screen. We tell you what we would do, what we would not do, and whether we are the right fit for your business.