We brought Southern Slumber’s Meta cost per acquisition to $38.24.

The founder had taken the ads in-house after a mentorship course and lost money doing it, then switched everything off. October 2025 was the clearest month: $9,000 of Meta spend produced $5,800 in sales. Her husband had pulled thirteen months of store data and found the blended return at roughly 1.0x, a large loss after cost of goods and fulfilment.

“This October… was the 9,000 ad spend to generate 5,800 in sales. So we actually have a negative ROAS for this October… So when you put in our cost of goods sold and our fulfillment costs and all our variable costs, it’s a huge loss.” Tanner Powell, co-founder.

The obvious read was that the ads were bad. They were not the constraint. Site conversion had fallen from above 2% to below 1%, and a recent Shopify theme change was the suspected trigger. Mobile page speed measured 37 out of 100 against a 70 target. The homepage led on features rather than benefits and carried two competing calls to action. The minicart quoted two different free-shipping thresholds, $49 in one place and $29 in another. Three separate 15% discount codes were live at once, from email signup, a thank-you card and a first purchase, all cannibalising each other. Every dollar of Meta spend was landing on that.

Step #1. Find the constraint before touching the budget.

We diagnosed the website rather than the ad account as the primary bottleneck and told them to cut ad spend to fund the fixes. That is the opposite of what a retainer-maximising agency does.

Step #2. Build the audit and give it away.

Simon Kristensen built a full site and speed audit unpaid, delivered in mid-November 2025 and walked through with the founder. It was written for a developer to act on: one benefit-led call to action, a founder photo and signed message in place of the stock brand-story image, image zoom and a short product video on product pages, the Shop Pay button removed, and one consistent $29 shipping threshold. The client took it to a freelance developer at $50 an hour and rolled it out in phases.

Step #3. Simplify the offer, then run the account properly.

We recommended making the standing 15% offer visible up front instead of gating it behind an email signup, and retiring the duplicate codes. Simon changed his own default position on always-on discounting for this brand after studying how it actually sells. The Meta account ran as a CBO campaign with dynamic product ads and separate prospecting and retargeting lanes, seasonal creative batched around the Mardi Gras collection, and a December gifting set of Buy 3 Get 1 Free plus a free $19 lovey on orders over about $120. It hit Meta’s 50-ad ceiling and we pruned low performers to make room.