HyperPop Results Jolly
Metro Vancouver · My own company
Jolly Heating & Cooling
These were Jolly's first four months in operation, not just its first four months running Meta ads. No reviews, no referrals. January to April 2025, and every play I run for a client runs here first.
A note on how these are counted. A sale is booked against the month its estimate was run, not the month the job was invoiced, so each month's cost per sale reflects the leads that month's ad spend actually produced.
Every figure here is the Meta ad account and nothing else. We rebuilt Jolly's website in July 2026, fifteen months after this window.
Figures cover January – April 2025 · Metro Vancouver
Cumulative · Meta ad account
CA$297,938
Revenue
13
Sales
CA$20,467
Ad spend
- Reporting period
- Jan – Apr 2025
- Cost per lead
- CA$97.93
- Average ticket
- CA$22,918
The progression · Cost per sale by month
Cost per sale, month by month.
It started at CA$3,094 the first month a job closed, and reached CA$600 by month four. Still falling, not bottomed out.
No sales
CA$3,094
CA$2,869
CA$600
Bars are cost per sale. January closed no jobs, so it has none.
| Month | Leads | Estimates | Sales | Revenue | Ad spend |
|---|---|---|---|---|---|
| January | 34 | 10 | 0 | CA$0 | CA$5,632 |
| February | 31 | 7 | 1 | CA$54,550 | CA$3,094 |
| March | 63 | 26 | 2 | CA$29,100 | CA$5,737 |
| April | 81 | 47 | 10 | CA$214,288 | CA$6,004 |
| Total | 209 | 90 | 13 | CA$297,938 | CA$20,467 |
February's CA$54,550 was a single job: a gas boiler replaced with 9 ductless heads. One sale, not a typical month.
April, on its own
CA$600 in ads to book a CA$21,429 install.
- Average ticket
- CA$21,429
- Cost per sale
- CA$600
- Sales per CA$1 of ad spend
- CA$35.69
In April, 10 sales on CA$6,004 of Meta spend. That is CA$35.69 of installed work for every dollar we put in, at an average ticket of CA$21,429 and CA$600 to book each sale.
What moved the bars
The selling, not the ads.
The channel worked from month one. January put up 34 leads and 10 estimates for a company nobody had heard of. Getting in front of homeowners was never the problem.
What was wasting those leads was everything after one landed. We had not worked out our positioning: who our customers actually are, and what they want in a system, a quote and a price. We did not know how our pricing stacked up against the other companies quoting the same jobs. And we did not have the leave-behinds and sales materials that answer a homeowner's questions before they ask them.
The ads did not get better, the selling did. Buying leads is not the whole job. A contractor who turns on a lead source and changes nothing else gets January's numbers, not April's.
-
−55%
Cost per lead
-
2×
Lead to estimate
-
−81%
Cost per sale
Cost per lead and lead to estimate compare January with April 2025. Cost per sale compares February with April, because January produced no sale to measure against.
The website
Built to close, not to catch leads.
We relaunched Jolly's website in July 2026, so there are no website numbers to show yet. I am not going to invent them. What I can tell you is what it is for, because a site for an install business is not a lead machine.
Jolly's leads come from Meta ads. The website's job comes after that: it is where a homeowner checks us out the night before they let a stranger into their home, and it is what I hand them during the estimate. The blog is sales material, used at the kitchen table, not bait for a search engine.
- Guide posts
- Used in the sale. A homeowner deciding between systems or working through rebates gets sent the post that answers the question, instead of a phone call that goes to voicemail.
- Published starting prices
- A homeowner can see roughly what a job costs before they call, so the estimate starts from an informed place instead of the price being the first surprise.
- Case studies
- Coming. This page and the ones like it, handed over during the estimate as proof, so the pitch is not just me saying the work is good.
- Location pages
- Coming, city by city across Metro Vancouver, so a homeowner searching their own town finds a page written for it.
The search numbers go on this page once there are three to six months of post-launch data behind them. A number with no before and after is not worth putting up.
Want this run on your own account?
Everything on this page ran on my own company before it ran on a client's. If you own an HVAC business and want to see what a lead costs in your market, what it converts to, and what I would run, book thirty minutes and I will walk you through it.
Book a 30-minute call