High ACoS: what causes it and how to diagnose the campaign that bleeds

A high ACoS almost never has a single cause: it comes from paying too much per click, from bringing in traffic that doesn’t buy, or from a listing that converts poorly once the click lands. In other words, a high ACoS is a symptom, not a diagnosis. Before you cut the budget or pause the campaign —the first thing every seller wants to do— you need to know which of those three links is where the spend is bleeding out. Lowering the bid on a campaign whose real problem is a listing with no photos just costs you profitable sales without fixing anything.

The mental rule is simple: ACoS = CPC ÷ (sale price × conversion rate). Anything that raises the CPC, lowers conversion, or reduces your effective price pushes ACoS up. Diagnosing is, literally, going term by term asking which of those three levers moved. An ACoS that jumped from 18% to 40% in two weeks is not explained by “the campaign is expensive”; it’s explained by a specific number that changed, and that number is in the report.

The problem for the multichannel seller is that this number lives scattered. Amazon gives you its search term report with its ACoS; MercadoLibre reports Product Ads with a different logic; your product cost sits in a separate catalog and your stock in another tab. To find out why a campaign bleeds you end up, once again, downloading reports at eleven at night and cross-referencing them by hand in Excel. And by the time the table is built, the bids have run another day with the same problem. This article gives you the diagnostic order so you stop guessing.

first define “high”: high compared to what

Before you call an ACoS “high,” you need something to compare it to. A 30% ACoS is a disaster on a product with 25% margin and a bargain on one with 55%. The only threshold that matters is your break-even ACoS: the percentage of ad spend at which you stop making money. If after commission, fulfillment, product cost and taxes you keep 35% margin, any ACoS below 35% leaves profit and any above it burns it.

So the first step is not to look at the campaign, it’s to look at the margin. A 28% ACoS can be perfectly healthy and require no action; it just looks “high” against the feeling that “it should be 15%.” Without the real margin per SKU in view, every diagnosis starts crooked. And note: break-even is not the same on Amazon as on MercadoLibre for the same product, because commissions and fulfillment differ. A campaign can bleed on one channel and be fine on the other at the same nominal ACoS.

Glossary: real net margin is what’s left after ALL costs —product, fees, shipping, tax and advertising—, not just price minus cost.

cause 1: the CPC spiked

The first lever is CPC: how much you pay per click. If your CPC went up and everything else stayed the same, your ACoS rises by the same proportion. CPC rises for two typical reasons: more competition entered bidding on your same terms, or you raised the bid chasing impressions and the system charged you for it. A CPC that went from $6 to $11 on a keyword doubles its contribution to ACoS even if conversion hasn’t moved a millimeter.

To diagnose this you open the term report and sort by CPC, high to low. The expensive terms that also convert poorly are the first suspects. The classic trap is bidding high on very competitive generic keywords —“bluetooth headphones”— where you fight against everyone and overpay for cold traffic. Lowering those bids or moving them to exact match usually trims CPC without sacrificing the good sales. If you want to understand deeply how that price per click is formed, the pillar on CPC explains what pushes it and how to control it.

cause 2: irrelevant terms bringing clicks that don’t buy

The second lever is traffic relevance. Even if your CPC is reasonable, if the campaign brings clicks from people searching for something else, you pay for visits that were never going to convert. This happens a lot in automatic campaigns and in broad match: the algorithm pairs your product with tangential searches. You sell cases for the iPhone 15 and end up paying clicks from someone who searched “iPhone 13 case.” The click gets charged, the sale doesn’t come, and each of those empty clicks inflates the numerator of your ACoS without adding anything to the denominator.

The diagnosis here is the search terms report, not the keyword report. There you see the actual phrases people typed before clicking. You sort by clicks and filter the terms with many clicks and zero or almost zero sales: those are pure leaks. The action is to turn them into negative keywords. A handful of well-negated irrelevant terms usually lowers ACoS faster than any bid adjustment, because you’re not cutting good traffic, you’re just closing the tap on the bad.

cause 3: the click lands but the listing doesn’t convert

The third lever —and the most underestimated— is conversion once the click is already on your page. If two people out of a hundred buy instead of three, your cost per sale rises 50% even if CPC doesn’t change. And here the problem is almost never the campaign: it’s the listing. Poor photos, price out of market, not enough reviews, or —the silent cause that fools everyone most— being out of stock on the best-selling color or size.

This last case deserves attention because it breaks every diagnosis that only looks at advertising. You ran out of black, which was 60% of your sales. The clicks keep coming in the same because the campaign doesn’t know there’s no stock, but people land, don’t find what they wanted, and leave. Your conversion collapses, your ACoS spikes, and you review bids and terms for days without finding anything odd… because the problem is in the warehouse, not in the campaign. That’s why reading advertising next to real-time inventory is not a luxury: it’s the only way to rule out the cause that doesn’t live in the ads panel.

Glossary: real available stock is the sellable stock net of reservations and transit; if it drops, your conversion falls and your ACoS rises without the campaign having changed.

the right order to diagnose

With the three causes clear, diagnosis stops being trial and error. The order that saves money is this: first confirm the ACoS really exceeds your break-even (if not, there’s no problem). Second, check availability: do you have stock of the SKU and its main variants? If not, the fix is to restock, not to touch the campaign. Third, open the search terms report and negate the leaks from irrelevant clicks. Fourth, sort by CPC and lower bids on expensive terms that don’t convert. Fifth, if traffic is relevant and CPC reasonable but it still doesn’t convert, the problem is the listing: photos, price, reviews.

Notice that three of those five steps aren’t even in the campaign interface. Availability lives in your inventory; margin, in your costing; market price, in your competitor monitoring. The click-through rate (CTR) tells you whether the problem starts before the click, in the ad itself. Diagnosing a high ACoS is, at heart, an exercise in cross-referencing signals that normally sit in different tabs. When they all live in the same dashboard, the cause jumps out in minutes instead of hiding among three manually downloaded reports.

why real time changes the diagnosis

The hidden cost of diagnosing by hand isn’t just the time you lose building the Excel: it’s that the campaign keeps bleeding while you investigate. Every day you spend cross-referencing yesterday’s reports is another day of bids running with the same problem. If your conversion dropped from lack of stock on Monday and you find out Thursday, you paid three days of clicks that were never going to convert. In advertising, deciding with old data has a direct, daily price.

A single source of truth in real time changes the question from “what happened last week?” to “what’s bleeding right now and why?”. Seeing the ACoS, CPC, conversion, real margin and available stock of the SKU on the same screen, updated, turns a three-hour diagnosis into a three-minute read. It’s not about having more charts: it’s about the cause and the metric living together, so the campaign that bleeds stops bleeding the same day you catch it, not the following week.

Glossary: ACoS is ad spend divided by the sales attributed to those ads; a high ACoS is a symptom, not a cause.