Brand's ACoS looked fine at 28% — right in line with category averages. The audit revealed that three of twelve ASINs were running PPC at a net loss after FBA fees and COGS, masked by two high-margin products that were subsidizing the rest.
This is a case about a brand whose PPC looked perfectly healthy to their previous agency, and was in fact quietly destroying margin.
A 12-ASIN supplement brand, strong category presence, growing year-over-year. Their previous agency reported a 28% blended ACoS — roughly in line with the supplement category average. Revenue was up. Ad spend was up. Everything on the dashboard looked fine. The founder came to us because net profit wasn’t growing in proportion to revenue and he couldn’t figure out why.
The answer was hiding in the ASIN-level breakdown, which nobody had been looking at.
Three of the twelve ASINs were running PPC at a net loss. Not a small loss — a structural loss. After Amazon’s FBA fees, COGS, referral fee, and advertising cost, the brand was losing money on every ad-driven sale of those three products. Two high-margin hero ASINs were generating enough profit to mask the losses across the rest of the catalog, which is why the blended ACoS looked acceptable.
This is the number-one failure mode of generalist PPC management in the supplement category. Blended ACoS is a useful top-line metric, but it’s not actually a profitability metric. A 28% ACoS on a product with 55% gross margin is profitable. A 28% ACoS on a product with 22% gross margin after fees is a structural loss. Most agencies don’t build their reporting at the ASIN-margin level, so they never see the distinction.
We also found that search term harvesting — the work of reviewing actual search terms driving ad spend and adding irrelevant ones as negatives — hadn’t been done in over five months. The same wasted keywords had been running across multiple campaigns for most of the year. Negative keyword lists weren’t synchronized across Sponsored Products and Sponsored Brands, which meant the brand was effectively paying Amazon to compete with itself on certain terms. And dayparting data showed 23% of spend going to time windows with conversion rates well below the account average — spend nobody had audited.
First, we paused PPC entirely on the three margin-negative ASINs and repriced them. Two of the three could be made margin-positive with a modest price increase and FBA fee category recode. The third needed to exit the catalog — it was structurally unprofitable regardless of advertising, and the brand had been subsidizing it without realizing.
Second, we rebuilt the PPC structure from the ground up. New campaign architecture organized by ASIN-level margin tier, search term harvest against 5 months of accumulated data, synchronized negative keyword lists across all ad types, dayparting bid adjustments.
Third, we installed margin-level reporting. Every weekly update now shows net margin by ASIN, not blended ACoS. The founder can see at a glance which products are actually making the brand money and which are consuming subsidies from the rest of the catalog.
Net margin recovered 11 points within 90 days. Wasted spend dropped 38%. Two of the three margin-negative ASINs returned to the catalog as profitable products after repricing. The third was sunset. PPC spend went down slightly in absolute terms — and net profit went up significantly.
The more important outcome: the founder now understands what questions to ask his agency. “What’s our ACoS?” was the wrong question — it got an acceptable answer while profit leaked. “What’s our net margin by ASIN after all fees?” is the right question. Any agency that can’t answer it in 30 seconds doesn’t have the data set up to actually optimize for profit.
ACoS is a comforting metric. It’s easy to measure, easy to benchmark, and easy to report on. It’s also an incomplete picture of whether your Amazon business is making money. If you’re a supplement brand on Amazon doing six figures a month and your agency reports on blended ACoS, there’s a reasonable chance the same pattern exists in your account — a few high-margin heroes subsidizing a few structural losses, with the blended number hiding the problem. The audit is how we find out which.
Every audit surfaces findings this specific. The only question is what's hiding in your account.