A US camper builder came to us on one unit sold in the month, with Google as its only channel and Google no longer working. We fixed the measurement, rebuilt Google, then opened Meta, a platform the founder had ruled out. It produced roughly $60,000 in its first month, and the lead capture on top of it returned about 70 leads inside days at a reported cost of about $1 each, against a product averaging around $11,000.
A channel the client had never used produced roughly $60,000 in its first month, and filled a phone pipeline at about a dollar a lead.
The founder said the business was in trouble. At the setup call he was open to significant spend to “get the business back on track”. Days later he told Brian where he was: “I’m literally on one unit this month, Brian.” His goal was 80 units a month, and we came on in his trough. Brian: “This is his quieter time. He has massive months in like February, January, February, March.”
Google had been the whole business and Google had stopped working. Brian: “they were always Google predominantly, Google crashed, and then we did Google and Meta with them.” Meta had never been run and was not on his list: “Like Meta for him was never even a consideration. like, why would I do that? I was like, you should do it.”
The reporting he steered by did not agree with itself. On 2025-10-16 Triple Whale showed 148,000 in sales while Shopify, which it was linked to, showed roughly double. Average order value had the same problem the other way. He reported it at around $11,000 while the store data showed 800 in September and falling.
Step #1. Fix the measurement, then cut Google back before adding to it.
He wanted to open at $750 a day on Google and $500 a day on Meta. We took the account apart first. Brian at the setup call: “let’s cut back first and fix your campaigns from the ground up.” In week two Simon found the Triple Whale gap and stopped the scaling conversation until it closed. The spend gate then came off the client’s own economics: on a reported 55% net margin, Simon’s rule was “we need to scale based on net profit right here”, which put profit break-even at roughly 2.0x rather than at a ROAS number someone had picked.
Step #2. Open Meta without moving the Google budget.
Meta went live as five lanes: website leads, USA broad Advantage+, retention add-to-cart, 180-day visitor retargeting, and geo-targeted. Google spend stayed where it was. Brian: “he kept his spend the same at what he had spent previously on Google, and he threw money at Meta.” Creative ran at 300 tests a month, the highest volume in our book at the time, built with the client’s own in-house videographer. Our copy ran in the same campaign as his own, so the comparison was live.
Step #3. Build the lead capture, kill the test that lost.
Brian designed an incentivised form capture inside Meta, with the client’s own person phoning every lead that week. About 70 leads came in within days. We tested geo-targeting, because customers collect these units in person. We recommended a 100 mile radius; the client pushed it to 300. It lost and we said so: “this did nothing for him, right? Geo-targeting actually wasn’t better for him when it came for his product”. Targeting went back to the whole of the US. Outside the platforms we pushed him on off-season demand, abandoned cart email flows, and hiring a rep to work the phone leads.