Case study · Auto protection · Minneapolis, MN
$483 a job here.
$20 a job over there.
Same account.
A paint protection film and window tint shop with two locations. Five Google Search campaigns, 60 days, $7,328. Business name withheld; numbers unchanged.
What the dashboard showed
On cost per lead, every campaign looked healthy: between $4 and $26. That's the number most reports stop at, and on that number nothing needed fixing.
What the receipt showed
The shop's job software already reported completed jobs back to Google Ads. Matching those to campaigns changed the picture:
| Campaign | Spend | Leads | Jobs | Cost / lead | Cost / job |
|---|---|---|---|---|---|
| Paint protection · city | $2,417 | 102 | 5 | $24 | $483 |
| Paint protection · suburb | $1,207 | 47 | 6 | $26 | $201 |
| Window tint · city | $1,140 | 111 | 7.5 | $10 | $152 |
| Window tint · suburb | $1,383 | 134 | 21 | $10 | $66 |
| Brand-name searches | $1,180 | 289 | 58.5 | $4 | $20 |
How we got these numbers: spend and leads from the Google Ads API, daily, Aug 10 – Oct 8, 2026. Leads = form submissions plus calls from ads and website. Jobs = completed jobs uploaded from the shop's job software and CRM as offline conversions; half-jobs occur when credit is split between campaigns.
Three findings
1. One campaign took a third of the budget and closed 5% of the jobs.
The city paint protection campaign spent $2,417 (33%) and produced 5 of 98 jobs. Its leads were real people, but at $483 per paid job it cost six times the account average.
2. The campaigns that closed best were capped by budget.
Every campaign lost between 34% and 49% of its possible impressions because the daily budget ran out, including brand searches at $20 per job. The account was turning away its cheapest customers.
3. Lead counts hid the difference.
Window tint in the city and in the suburb cost the same per lead ($10) but $152 versus $66 per job. Without job data, they'd be treated as equals.
The plan
- Cut the city paint protection budget by half and narrow it to the search terms that produced its 5 jobs.
- Move the freed budget to brand searches and suburban window tint until their lost-to-budget share falls under 10%.
- Add negative keywords for research and DIY searches (“PPF vs ceramic coating”, “how to install clear bra”, “vehicle wrap DIY”).
- Checkpoints: week 1 (spend shifted, no tracking breaks), week 3 (cost per lead in receiving campaigns), day 90 (cost per job, with 90 days for jobs to complete).
Projection
Same $7,328. About 17 more jobs.
Moving $1,200 out of the weakest campaign, and assuming every moved dollar performs at half the efficiency of the campaign it moves to.
This is a projection, labelled as one. We'll update this page with measured results after the 90-day checkpoint.
What does your receipt look like?
Grant read-only access to your Google Ads account. In 2–3 business days you get a plain-English audit showing the dollars that produced customers and the dollars that didn't. Free, no obligation.