If I had to delete one Amazon Ads metric from every dashboard in the world, it would be ACoS. Not because it is wrong. Because it is incomplete, it is constantly misread as a business metric, and it routinely hides the one number that actually predicts whether the account is making money.
That number is TACoS. Total Advertising Cost of Sales. Ad spend divided by total revenue – not just ad-attributable revenue. The metric that ties Amazon ads directly to the P&L instead of to Amazon’s attribution model.
Across India, US, UK, UAE and KSA over seven years, the pattern is identical in every market. Teams who run an account by ACoS post a healthier-looking number and a sicker-looking P&L. Teams who run an account by TACoS post a number that takes longer to explain to a CFO – and a P&L that holds up.
What ACoS actually measures
ACoS is ad spend divided by ad-attributable revenue. If you spend $10 on a campaign and Amazon credits $50 of sales to that campaign, your ACoS is 20%. That is a useful number for one thing: comparing the efficiency of two campaigns inside the same account.
It is not useful for anything else.
It does not tell you whether the campaign cannibalised organic sales. It does not tell you whether the customers who bought would have bought anyway. It does not tell you what happened to the SKUs that sit in the same brand bucket. It does not tell you whether your overall ad-to-revenue ratio is sustainable. It only tells you the ratio of dollars in to attributed dollars out, in the narrow slice of the catalogue that the campaign touched.
When a brand says “our ACoS is 22%,” what they mean is “the slice of our sales that Amazon credited to ads cost us 22 cents on the dollar.” It says nothing about whether the 22 cents was money well spent.
What TACoS measures
TACoS is total ad spend divided by total revenue, including organic. The number that ties ads to the P&L instead of to Amazon’s attribution model.
This is the number that sits next to CM1 (pre-ads margin) and CM2 (net profit) in any honest P&L. If CM1 is 34% and TACoS is 21%, you have 13 points of CM2 left. If TACoS climbs to 26%, you have 8 points left. The number is comparable to other line items in the business: COGS, fees, freight. ACoS is not. ACoS is an internal ad ratio that does not show up anywhere on the P&L.
TACoS also catches what ACoS cannot. Cannibalisation. Hijacking. Stockouts. New competitors. Anything that shifts the balance of organic to paid shows up in TACoS within a week. ACoS is happy as long as the dollars-in-to-dollars-out ratio inside the campaigns holds.
The trap of the ACoS target
The most common operating mistake I see is the flat ACoS target across an account. “Hold everything at 25% ACoS.” It sounds disciplined. It is actually a recipe for value destruction.
Here is why. A SKU at 25% ACoS with 50% gross margin is generating contribution. The same 25% ACoS on a SKU with 22% gross margin is a loss. A flat ACoS target treats them the same. The campaign manager optimises both to the same number, and the high-margin SKU gets under-spent while the low-margin SKU keeps draining the account.
Worse, a flat ACoS target encourages the wrong reaction to a successful campaign. If a discovery campaign hits 35% ACoS, the team pulls back. But that 35% ACoS campaign might have been the only thing keeping organic share alive on a SKU where the brand has 12% category penetration. Pulling spend collapses the organic, and three months later the SKU is in trouble for reasons nobody can trace back to that one decision.
This is what happens on stuck accounts where two ASINs are running at 80% TACoS. The team’s instinct is to cut ad spend to push ACoS down. The actual move is almost always to restructure the campaigns and rebuild keyword targeting against the max paid ACoS formula, then increase spend in the right places. ACoS in some campaigns goes UP. TACoS comes down. The business gets healthier.
Where ACoS still has a use
I am not arguing for deleting ACoS from the platform. It still has a job. ACoS is the right metric for comparing the efficiency of two campaigns inside the same SKU bucket. If campaign A is at 18% ACoS and campaign B is at 28% ACoS on the same product with the same audience, the gap is meaningful. If they are both above the SKU’s CM1 ceiling, both campaigns need work. If they are both below, both are fine.
ACoS is also the right metric for short-window decisions. A new keyword bid test, a placement modifier test, a creative refresh. The ACoS movement tells you whether the change worked at the campaign level. The TACoS movement on a one-week window is too noisy to read.
The mistake is using ACoS as the business metric. As the number you report to the CFO. As the target on the quarterly plan. It is not built for that.
How TACoS forces honest conversations
When you put TACoS at the top of the dashboard and bury ACoS, three things change in how the team operates.
One, the ads team stops being able to claim wins that the business is not seeing. If campaigns are at 18% ACoS but TACoS is at 32%, the team has to explain where the gap is. Is the ad-attributed revenue cannibalising organic? Are there off-Amazon paid sources adding to revenue but not to ad spend? Is the catalogue mix shifting? The TACoS number forces those questions. ACoS lets the team hide behind a clean campaign-level dashboard.
Two, the conversation with finance gets simpler. Finance does not care about ACoS. Finance cares about the P&L. TACoS shows up on the P&L next to CM1 and CM2 (net contribution after ads). The CFO can read it without translation. The ads team gets taken seriously by finance for the first time, which is usually overdue.
Three, the spend ceiling stops being abstract. Once TACoS is the headline number, you can put a ceiling on it that ties to CM1. The spend ceiling math has its own post, but the short version is: max TACoS equals CM1 minus your target CM2 (net contribution). With TACoS as the headline, that ceiling sits on the dashboard every day. With ACoS as the headline, the ceiling is a memo nobody reads.
How to actually run an account on TACoS
The mechanics are not complicated.
Track TACoS at the SKU bucket level weekly. Roll up to the account level monthly. Set a TACoS ceiling per bucket based on CM1. Run to a floor 3 to 5 points below the ceiling so you have headroom. Use ACoS for campaign-level decisions inside each bucket, but never as the business target.
When TACoS drifts above the ceiling, do not start with bids. Start with the diagnosis. Is the drift because ad spend went up? Or because organic revenue went down? Those have different fixes. If organic is dropping, the listing has a problem, or a competitor is gaining share, or a stockout pulled the SKU off page one. None of those are bid problems.
When TACoS sits 5 points or more below the ceiling for two months in a row, that is a signal to invest. The bucket has room to spend more. Either expand the keyword set, push into a new ad type like Sponsored Display, or test creative variations. Sitting comfortably below the ceiling is not a win. It is unused capacity.
Pull TACoS on a cadence that matches the account, not on a generic Monday rhythm. On a stable mature account with strong organic share, a weekly account-level pull and a daily SKU-bucket pull is usually right. On a stuck account, a recovery account, or an account running active promos and lightning deals, look at TACoS daily – sometimes twice a day inside a deal window. The cadence flexes with team size, account size, season and promo state. There is no universal answer.
TACoS discipline matters even more on marketplace expansion work, because the organic baseline is lower and small TACoS drifts compound faster.
The reporting cadence that surfaces problems early
The dashboard is only half the discipline. The other half is the cadence at which you read it.
In a high-demand window like December, seasonality does much of the work – a lower CM1% can still produce a healthy CM2% because demand carries the volume and less ad spend is needed. The same SKU in January cannot lean on that. The cadence flexes with team size, account size, season and promo state. There is no universal weekly rule.
The reporting itself is not a time problem. Account-level TACoS takes roughly two minutes to pull. ASIN-level TACoS takes about twenty-five minutes if you are reading it properly. Brand and Sponsored Display layered on top brings the total to roughly thirty minutes – and most brands at any meaningful scale have an MIS analyst who runs the reports anyway. The discipline is in what you do with the numbers, not in the minutes spent pulling them.
The teams I see struggling are not the ones reading numbers too often. They are the ones reading the wrong numbers. Daily ACoS at the campaign level on a steady-state account is genuinely noise. Daily TACoS at the account level, especially during promotional windows, is the opposite – it is the signal that catches a Lightning Deal eroding margin in real time, or a Saturday US conversion drop the weekly average smooths away.
The ACoS-trap categories
There are categories where the ACoS trap is worse than average. Three I keep running into.
Categories with high organic search volume on brand terms. If you sell a branded product where 60% of conversions come from people typing your brand name into search, your ad-attributed ACoS will look fantastic. You are paying ads to show up for people who would have bought anyway. TACoS exposes this. ACoS does not.
Categories with strong seasonality. ACoS in November on a Christmas-driven SKU looks great because organic is up. ACoS in February on the same SKU looks terrible because organic dropped. The ads did not get better in November and worse in February. The denominator moved. TACoS is also seasonal, but at least it ties to the actual revenue you are earning, not to a slice of it.
Categories with subscription overlap. If your category sells partly through Subscribe and Save, those subscribers buy without seeing ads. The ad-attributed sales become a smaller share of total sales, and ACoS looks better than the ad spend justifies. TACoS catches it.
If you want a second read on where ACoS is hiding problems on your account, the Amazon Account Audit reviews it directly.
Bottom line – what to put at the top of your dashboard from tomorrow
ACoS is a campaign-level diagnostic. TACoS is a business metric. Stop putting ACoS at the top of the dashboard. Stop setting flat ACoS targets across the catalogue. Put TACoS at the top, anchor it to CM1 via the max paid ACoS formula, and let ACoS go back to being a tactical signal – the way it was originally designed.
If you want to walk through what switching the headline metric looks like on your specific account, message me on WhatsApp. No deck. Just the question.
Frequently asked questions
What is the difference between ACoS and TACoS on Amazon?
ACoS measures ad spend as a share of ad-attributed sales only. TACoS measures ad spend as a share of total brand sales (paid + organic). ACoS can look healthy while TACoS is quietly bleeding – and vice versa. ACoS is a campaign-level signal; TACoS is the business metric that belongs on the P&L.
Is a low ACoS always good on Amazon?
No. A low ACoS often means you are under-spending and missing impression share, or that ads are only converting buyers who would have bought organically anyway. A 100% ACoS on a launch campaign can be the right number. A 15% ACoS on a mature SKU can be too low. ACoS in isolation tells you almost nothing.
What is a healthy TACoS for an Amazon brand?
TACoS should leave at least 8-10 points of CM2 after ads, and the target shifts by season. In a high-demand window like December, seasonality does much of the work – you can run a lower CM1% and still make a healthy CM2% because demand carries the volume and less ad spend is needed. The same SKU in January cannot lean on that demand.