Case study · Drain & sewer
2-3 jobs a week, about to quit ads. Then 8 a week for nine straight weeks.
That's the account. Every number below comes off that screen.
Tap to enlargeClick to enlarge
May 19 to July 21 2026, nine weeks. Every job tracked in a shared log. The return improved as it ran - 3.56x over the first thirty days, 3.69x across the full nine weeks. He's paused for August. He works alone and he took the month.
Read this before the numbers
His clicks cost $3.26. Yours will cost more like $25.
This ran in Europe. A drain click there is about $3.26. In the US you're looking at $14 to $30 in most markets, and $45 to $65 in a big metro or on an emergency term. So don't take his cost per lead home with you. That number doesn't travel and I'm not going to pretend it does.
What travels is the two rates underneath it. His page turned 10.55% of clicks into a call. Most agencies are happy at 5 or 6. And of the 95 leads that came in, 74 ended up as a job on the calendar.
Run those at your prices. At $25 a click, a page converting at 6% costs about $417 to produce one lead. The same click at 10.5% costs $237. If three of four book like his did, that's $556 a job against $316. Same ad, same budget, same city.
At his prices that gap was worth a few hundred dollars over nine weeks. At yours it's the difference between the campaign paying for itself and you switching it off in October.
| At $25 a click | Typical page (6%) | This page (10.5%) |
|---|---|---|
| Cost per click | $25 | $25 |
| Cost per lead | $417 | $237 |
| Cost per booked job | $556 | $316 |
May 19 - July 21, 2026
The story
Attila runs a one-man drain and sewer service. Skilled at the work, drowning on the marketing side. His Google Ads were bringing 2 to 3 jobs a week and he was about to shut them off for good.
“It only brought 2-3 jobs a week, that's why I asked for help. I suffered a lot with it.”Client reference available on request.
The ads needed work, but the biggest lever was where the clicks landed. Every click now goes to one simple page built around one exact job - a backed-up main line - not a generic homepage. Everything on that page gets counted.
One thing worth saying plainly, because it cuts the other way from most case studies: the longer this ran, the better it got. A cherry-picked month gets worse when you extend the window. This one went from 3.56x to 3.69x. That's why I'm showing you nine weeks instead of the first thirty days.
Not everything worked first time. The first version was one campaign covering the whole county, and it spent real money without producing a single booked job. Instead of feeding it more budget I broke it into separate campaigns by service town, so each one could be judged on its own. Then it was obvious which areas were paying and which needed watching, and the money went where the jobs were.
Once that was steady he wanted a fourth area. Same build, new territory, and the jobs started coming from it too. That's the part that matters if you cover more than one town: it isn't one lucky city, it's a setup that gets repeated.
Where the 74 comes from
Tap to enlargeClick to enlarge
What's different here
Two things. Both of them are on this page.
Google counted 87. His job log counted 74 paid jobs.
Both numbers are printed above, because they disagree. The ad account counts a conversion when the phone rings. The job log counts it when the invoice gets paid. Every agency reports the first one and calls that revenue tracking. I put it next to your book, and when the two don't match I tell you which one to trust.
One page, one job, written from what his customers actually said.
Not a service menu. Every click lands on a page about one thing, the backed-up main line, built from his reviews and from talking to people who had already paid him. Knowing the drain business is not the same as knowing what someone types at six in the morning with water across the kitchen floor.
Contact
Let's look at your numbers.
Tell me your city and the service you want more of. I'll search what your customers are actually typing, see who's showing up above you and what their ads say, look at what happens when someone lands on your page, and send back what I'd change, in what order, and what it would take.
You'll know from the breakdown whether I'd be any good at this before you've spent anything or booked a call. If you'd rather take the list and hand it to whoever runs your ads now, do that instead.