Scaled ad spend from $30k to $130k per month and dropped CPL from $220 to $120 while improving lead quality. Revenue went from $700k to $2.4M. Their highest month ever.
This is a roofing company in Texas. When I was onboarded, they were spending around $30k/month on ads and generating about $700k in monthly revenue. Their cost per lead was sitting at $220, which isn't terrible for roofing, but I knew we could do better.
Important context: I started working with them in winter and we scaled through spring into summer. Roofing is seasonal—summer is peak season. So yes, I had a time advantage. But even accounting for that, what we pulled off was significant.
Their account was a mess. Like seriously. They had 3 campaigns with 20 different ad sets all trying to do the same thing. The only difference was the audiences they were targeting, which is hilarious because audience targeting has been dead since 2021. Broad targeting is where it's at now - just let the creatives do the heavy lifting.
Speaking of creatives, they had ZERO diversity. Almost everything that was working for them was the same format: picture of the business owner standing next to a happy homeowner, tons of text slapped on top. Sure, it worked for a bit, but it'd fatigue so fucking fast.
Their landing page was also a disaster. Too many buttons, too much text, way too many distractions. And their form had them asking unnecessary questions thinking they were "qualifying" leads, but really they were just scaring people away.
The worst part though was their tracking. They were only tracking leads. They weren't tracking appointments or sales. So they had NO IDEA which ads were actually making them money.
First, I cleaned up their media buying. Consolidated everything, went broad with targeting, and let the algorithm do its thing.
Then I started pumping out creative concepts. We were testing 5 new creative concepts every single week. That's how you beat fatigue - you stay ahead of it.
I rebuilt their landing page with ONE goal in mind: get people to fill out the form. That's it. No distractions, no extra buttons, no walls of texts to read. Literally a landing page with a headline, subheadline, and a multi-step form ONLY. And I shortened their form by removing those intrusive questions they thought were helping.
That alone got the conversion rate from ~3% to ~7%. This is more than double the leads and half the cost per lead just by changing the landing page. To be fair, their lead to appointment and appointment to sale conversion ratios DID take a negative hit, but it was still worth it because the overall cost per sale went down dramatically.
The biggest thing I did was fix their conversion tracking. I built this whole system that connects their 2 different CRMs (yeah, they use two) to both Facebook and Google Ads. Now every appointment and sale shows up in the ads manager with proper attribution to the right campaign, adset, ad, placement, keyword, etc.
The trick is that my form collects everything Meta and Google need to match people back to their ad clicks: Facebook Click ID, Google Click ID, IP address, User agent (browser, resolution, all that), Email, name, phone. All of this data gets sent to both platforms, and THAT'S what makes the attribution actually work.
You'd be shocked how many ads they thought were winners because of low CPL, but they weren't converting to sales AT ALL. Now we optimize for actual sales instead of just leads.
At the peak, we were spending $130k/month in ad spend. Revenue hit $2.4M in a single month—their highest ever. Cost per lead dropped from $220 to $120, and lead quality actually improved (better lead-to-appointment and appointment-to-sale ratios).
Full transparency: part of this was seasonal. We went from winter (slow season) into summer (peak season). But even their previous summer peaks never hit $2.4M. The combination of better media buying, creative diversity, conversion tracking, and yes, favorable timing, got them to numbers they'd never seen before.
Media buying (killed the fragmented campaign structure). Ad creatives (5 new concepts tested weekly). Landing page and form optimization. CRM backend and conversion tracking/attribution. Right now I'm working on implementing an AI bot that automatically texts and emails leads to book appointments. Based on what I've seen with other clients, this usually improves lead to appointment rates by about 30%. I'll update this when I have the data, but yeah, it's gonna be good.