Two numbers get quoted in every Amazon advertising report, and they answer different questions. ACoS tells you whether a campaign paid for itself. TACoS tells you whether the account is getting healthier. Reading the first as though it were the second is the most common way a good Amazon programme gets optimised into decline.
What ACoS measures, and what it leaves out#
ACoS, or advertising cost of sale, is ad spend divided by the revenue those ads produced, shown as a percentage. Spend $300, take $1,000 in attributed sales, and the ACoS is 30%.
What ACoS leaves out is everything the ads did not directly produce. Organic sales are excluded by definition. So is the shopper who saw a sponsored listing on Monday, thought about it, and searched the brand name on Friday. ACoS is a campaign metric that often gets read as an account metric.
- It answers one question: did this campaign return more than it cost?
- It ignores organic revenue, repeat purchases, and the rank gained from ad-driven sales velocity.
- It penalises launches, new ASINs, and any deliberate push for share.
What TACoS measures#
TACoS, or total advertising cost of sale, is ad spend divided by total revenue, paid and organic together. The denominator becomes the whole business on Amazon, not the slice the ad platform will take credit for.
That single change turns it into an account-health number. When advertising is doing its real job, driving the sales velocity that lifts organic rank, organic revenue grows faster than ad spend. TACoS falls while total revenue rises. That combination is the clearest good signal in Amazon reporting, and it is invisible in ACoS.
Amazon advertising benchmarks in 2026#
Across categories, the average Amazon account ran roughly a 32% ACoS in 2026, with most accounts falling between 25% and 36%. Below 28% is ahead of the market. Above 40% usually points at something structural instead of something that can be bid down.
The rest of the 2026 picture sits around a $1.18 to $1.22 cost per click, a 0.4% to 0.6% click-through rate, and a 10% to 12% conversion rate, with TACoS in the 10% to 15% band for an established brand.
Category spread matters more than any of those averages:
- Food and grocery: around 23% ACoS.
- Cross-category average: around 32%, with most accounts at 25% to 36%.
- Apparel: around 42%.
- Clothing: higher again, past 50% in some subcategories.
Comparing a clothing brand to a cross-category average is comparing it to nothing. The useful comparison is the same account last quarter, and the same subcategory.
Why a falling TACoS is the number to hold an agency to#
A rising TACoS on flat revenue means advertising is buying sales that would have happened anyway, or buying sales that do not stick. A falling TACoS on rising revenue means the paid spend is doing work that outlives the click.
That is why TACoS belongs at the top of a monthly report and ACoS belongs inside it. An agency can improve ACoS this week by turning off everything except branded search. Total revenue will not move, the report will look better, and the account will be worse.
When a high ACoS is the right answer#
Three cases make a high ACoS the correct decision rather than a problem to fix.
- A launch. There is no organic rank to harvest yet, so early spend is buying the sales history that rank is built from.
- Defensive brand terms. The alternative is paying nothing and letting a competitor take a click from someone searching for you by name.
- Inventory that has to move before a long-term storage fee, where the comparison is the cost of holding it, not the margin on it.
In each case the campaign is buying something other than immediate margin, and the report should say which.
How to calculate both from reports you already have#
ACoS equals ad spend divided by ad-attributed sales, times 100. Both figures are in the Amazon Ads campaign report.
TACoS equals the same ad spend divided by total ordered product sales, times 100. Total ordered product sales comes from Seller Central Business Reports, not from the ads console, which is why TACoS gets skipped: it needs two systems, not one.
Run both monthly for the account and for each of your top ten ASINs. An account-level TACoS can look stable while a single ASIN quietly stops carrying its own weight. This is the kind of thing our marketing analytics and reporting work is meant to surface before it becomes a quarter.
Three mistakes worth avoiding#
The first is judging a launch on ACoS. A new ASIN has no rank and no reviews, so every early sale is expensive and every early sale is the point.
The second is chasing a target ACoS across a mixed catalogue. A single number applied to products with different margins tells you to stop selling the ones that are working hardest.
The third is reading attributed revenue as incremental revenue. Amazon attribution credits an ad with sales that a brand-loyal shopper would have made regardless. TACoS is the check on that, because a shopper who would have bought anyway shows up in the denominator whether the ad ran or not.
What belongs in the monthly report#
Four lines, in this order, answer what an owner actually wants to know:
- Total ordered product sales, and the change on last month.
- TACoS, and its direction over the last six months.
- ACoS by campaign type, so branded and non-branded are never averaged together.
- Organic share of revenue, which is the thing all of the above is supposed to be moving.
If a report cannot show those four, it is describing the ad account, not the business. We run Amazon accounts and Amazon SEO against the second one.