ACoS vs TACoS: What Each Metric Tells You About Amazon Ad Profitability

ACoS is a campaign metric Amazon defines. TACoS is a business metric you have to calculate, because Amazon does not publish it as a report. Teams that manage only the first number will cut the campaigns that are building the second, or they will celebrate a low ACoS on a brand that cannot sell unless it pays for the click.

AEG reports contribution, not a single ratio. The stance is the advertising essay. This page is the arithmetic, and the decisions the arithmetic does not make for you.

What ACoS is

Amazon Ads defines advertising cost of sales as ad spend divided by the sales attributed to the ads, expressed as a percentage.

ACoS = (ad spend ÷ ad-attributed sales) × 100

Amazon’s own example: $50 of spend and $100 of ad-attributed sales is a 50 percent ACoS. Return on ad spend is the inverse, ad-attributed sales divided by spend. In that example, ROAS is 2. Same facts, different presentation. Amazon’s guide is advertising.amazon.com/library/guides/acos-advertising-cost-of-sales.

Two limits sit on the number.

Attribution. The sales in the denominator are the ones Amazon’s ad console attributes to a click, inside the window you selected. Change the window and the ACoS changes without a single bid changing. Say which window you are using. Do not change it mid-quarter to improve a chart.

Delay. Sponsored Products sales do not all post the hour the order happens. Do not judge a campaign on “today.”

ACoS says nothing, by itself, about whether the order made money. A 20 percent ACoS on a product with 15 percent contribution after fees and cost is a loss. A 40 percent ACoS on a launch, against a contribution that can carry it for a defined period, can be a decision. Amazon’s older ACoS guidance says the same thing in its own words: a sensible target depends on your margin, and there is no universal good ACoS. We will not publish one.

What TACoS is

Total advertising cost of sales uses the same spend and a larger denominator: all sales of the product or the account, not only the sales the ad console claims.

TACoS = (ad spend ÷ total sales) × 100

Amazon does not offer a report named TACoS. You build it from ad spend on one side and ordered product sales on the other, for the same dates and the same ASINs. If the date ranges do not match, the ratio is a story.

Read the gap between the two numbers.

• ACoS 25 percent and TACoS 10 percent means most of the sales did not need a click the console could see. Advertising is a tax on a larger business. Watch whether the tax is shrinking as the page ranks for its own name.
• ACoS 25 percent and TACoS 24 percent means almost every sale is an ad sale. The page is rented. A “healthy” ACoS in that situation is the business paying a quarter of revenue to stay visible, which is a fact about the brand, not a fact about the bids.
• A falling ACoS and a rising TACoS can mean you cut spend so hard that total sales fell faster. The campaign looks efficient because it got small.

TACoS moves slowly. It is a monthly chart. ACoS moves when you negate a term or change a bid. Do not use the slow chart to manage a keyword, or the fast chart to judge whether the brand is becoming less dependent on ads.

Where each number belongs

Use ACoS, or ROAS if your team thinks in multiples, inside a campaign review.

• Exact and phrase terms that name the product should be judged against contribution after referral fees, fulfillment, returns, and cost of goods. The fee schedule is on sell.amazon.com/pricing. The 2026 fee letter is the update. The pricing essay is the floor those fees have to clear.
• Branded queries are often cheap. A low ACoS on the brand name is not evidence that advertising is “working.” It is evidence that people who already searched for you bought you. Defend the name. Do not present that campaign as the growth engine.
• Discovery terms are allowed a worse ACoS for a period you write down. When the period ends, the term either earns a place or it gets negated. An open-ended “investment ACoS” is how spend escapes.

Use TACoS in the monthly brand review, next to three other lines.

1. The share of total sales that came from queries containing the brand name. Amazon Brand Analytics is where that search share is visible.
2. In-stock rate on the advertised ASINs. A stockout makes TACoS look however the remaining days look. It is not a marketing result.
3. The price the customer paid versus the floor. A low TACoS achieved by discounting is the brand getting smaller.

If you need path and audience questions that neither ratio answers, that is what Amazon Marketing Cloud is for. It does not replace TACoS. It does not know your cost of goods unless you bring it.

1P and 3P break the comparison

On Vendor Central, retail revenue is Amazon’s. A vendor-funded ad dollar is not coming out of the same P&L as a seller’s ad dollar. An ACoS target copied from a vendor business review into a seller account is the wrong denominator twice. The models are compared in Vendor Central vs Seller Central.

On a seller account, total sales for TACoS should be the sales you recognize, net of the returns you are willing to count, for the ASINs in the ad account. Do not include a wholesale purchase order in the denominator. That mix makes TACoS look artificially low and teaches the team nothing.

What to do on Monday

Pull last month’s spend, ad-attributed sales, and total sales for the hero ASINs. Compute both ratios. Then do only this:

• Negate search terms that spent and did not belong to the product.
• Pause any campaign whose ACoS, after fees and cost, cannot be explained as a launch you still believe in.
• Leave branded defense on, at a bid that does not chase the generic terms.
• Write the TACoS down. Look at it again next month. Do not restructure the account to move it in a week.

Ongoing brand growth is the review in which these two numbers sit beside price and stock. They are not the review.

Questions

Should we optimize to a target ACoS?

Optimize to contribution. Let ACoS be the campaign’s efficiency against the sales it is allowed to claim. A target with no margin under it is a slogan. Amazon says the target varies by seller. That is the correct amount of precision.

Why did ACoS improve and the business get worse?

Because ACoS ignores every sale the ads did not get, and every sale the ads chased away by stocking you out or training the customer to wait for a coupon. TACoS, in-stock rate, and the price versus the floor catch that. ACoS does not.

Is TACoS an Amazon metric we can export?

No. Build it. Keep the ASIN set and the dates stable so the line is comparable. A metric you recompute with a new formula every month is not a trend.

Sources

• Amazon Ads: what ACoS means, with the formula. Spend divided by ad-attributed sales. ROAS as the inverse.
• Amazon selling fees. The costs an ACoS target still has to survive.
• 2026 US referral and FBA fee update. Why last year’s unit economics are not this year’s.

Book a strategy call

If the ad account is busy and you still cannot say whether the brand is less dependent on it than last quarter, send a brief or Book a strategy call.