Case study · September to November 2026 (returns for 2026 to Nov 9)

Which Ads Actually Pay for a Chicago HVAC Company

The owner of a Chicago HVAC company spends on Google Ads, Local Services Ads, Microsoft Ads and a few smaller channels to keep the phones busy. The owner's question was simple. Which of that money comes back as paying jobs?

about 2.8× Return on Google Ads and Local Services, 2026 to date about 1.8× for all marketing, counting the costs we can see, our fee included · Jobber revenue, costs as booked, read 2026-11-09

Chicago HVAC company

HVAC · Ads + revenue by channel September to November 2026 (returns for 2026 to Nov 9)

The problem

The owner of a Chicago HVAC company spends on Google Ads, Local Services Ads, Microsoft Ads and a few smaller channels to keep the phones busy. The owner's question was simple. Which of that money comes back as paying jobs? The numbers I'd been showing the owner didn't answer it. My dashboard said marketing brought in a small number of jobs. Then it put the company's whole invoice total under "brought in", like marketing had earned every dollar of it. One number was too low, the other was too high, and neither one helped the owner decide where the next dollar should go. That's on me.

What I found

The office had already done the hard part. Staff had typed a lead source into the CRM's own Lead source field for nearly 6 in 10 of the company's clients, by hand, as new customers came in. My dashboard was reading a separate custom field with the same name. That one was empty, so it missed almost all of their work. Local Services Ads had never been priced against the company's own revenue. When I matched a year of charged leads to the job records in Jobber, the result split hard by job type. Repair leads paid back well. Maintenance leads, the bigger group, brought back less than they cost on the first job. A big share of the Local Services leads were bookings and messages with no name, so there was nothing to match on. I kept those on their own line instead of counting them as losses.

Count from the office's records

A client counts as marketing's now when the website verified where they came from, or when the office tagged them to a marketing channel. I show the two counts side by side. Referrals stay out, even when the note names a platform.

Claim only marketing's share

The revenue I report for marketing is the paid invoices of those clients. The company's total sits next to it, labeled as the total.

Report Local Services by job type

Leads from the Lead Inbox get matched to Jobber on name and the last ten digits of the phone number, and the return is split by the job type Google charged for.

Keep maintenance on

Switching off maintenance job types in Local Services was an option. I kept them on because maintenance customers come back, and put the work into answer rate instead. Of the inputs Google names, that's the one the business fully controls.

Let the owner's page keep itself current

It reads Jobber and QuickBooks on its own, four times a day. It shows what marketing returns by channel and all in, counting the marketing costs we can see, my fee included. Figures from the books are labeled "as booked", next to how far each account has been reconciled.

The numbers

The owner can see on one page which channels bring paying customers and what those customers paid, without waiting on anybody to build a report. The jobs counted as marketing's come from the office's own tags, and the revenue claimed for marketing is only theirs. The jobs counted as marketing's went up more than tenfold. The revenue claimed for marketing came down to about a fifth of the company's paid invoices. For 2026 through November 9, Google Ads and Local Services together returned about 2.8 times their cost. Marketing as a whole returned about 1.8 times once every cost we can see is counted, my fee included. On Local Services, repair leads returned about three and a half times their cost on the first job. Maintenance leads returned less than half. All charged leads together came in around 1.35 times.

What I can't show yet

The return is revenue, not profit. The costs are the company's books as booked, which ran through early November when I read them. Channels with no lead source of their own in Jobber, like Microsoft, get no revenue credited.

Result table

The receipts.

Metric Value Source
Jobs marketing brought Clients verified by the website or tagged by the office. Nearly 6 in 10 clients carried an office tag Jobber, read 2026-10-29
Revenue claimed for marketing Paid invoices of those clients only. The company total shown apart, labeled as the total Jobber paid invoices, read 2026-10-29
Return by channel and all in That revenue against ad spend, then against the marketing costs we can see, my fee included Jobber, QuickBooks and the ad platforms, read four times a day
Local Services by job type Charged leads matched to jobs on name and last ten phone digits. Revenue of the first job Lead Inbox export and Jobber, 2026-09-08

Open your CRM and count how many of last year's customers carry a lead source. If it's most of them, you're already holding the evidence. I'll read that field and your Local Services leads and show you which ads pay. [Send me a note at No Brainer Media](https://nobrainersmma.com/contact/).

Want this kind of result?

Book a discovery call. We'll walk your numbers.

No pitch deck. Just your CRM, your ad accounts, and a straight read on what the next 90 days could look like.