Glossary · 8 min read
ACOS vs TACoS: Formulas, Definitions and an Example
ACOS measures how efficient ads are on the sales they get credit for. TACoS measures ad spend against all sales. Formulas, a worked example with made-up numbers, and how attribution windows and halo sales change both.
By The Vexio team Published Updated
ACOS vs TACoS: the two formulas
ACOS = ad spend / ad-attributed sales x 100
TACoS = ad spend / total sales (ad plus organic) x 100
The numerator is the same. The denominator is what separates ACOS vs TACoS. ACOS divides by the sales Amazon credits to your ads, so it judges how efficiently the ads convert. TACoS divides by every sale, paid and organic, so it shows how heavy the ad bill is against the whole business. Neither number is good or bad until you set it against your margin and your goal. The glossary has short definitions of both terms.
What do ACOS, ROAS and TACoS mean?
ACOS
Amazon’s ACOS guide describes it as a metric for pay-per-click campaigns, calculated by dividing ad spend by ad revenue and converting to a percentage. Its example is $50 of spend producing $100 of sales, a 50% ACOS. The guide is written around Sponsored Products, so check the column definition for any other ad type. The denominator contains only sales credited to the ad. Organic sales the ad may have helped are not in it.
ROAS
Return on ad spend is ad-attributed sales divided by ad spend. Amazon’s guide calls it the inverse of ACOS: in the example above, ROAS is 2. By the same arithmetic a 25% ACOS is a ROAS of 4. The two carry identical information. ROAS grows when things improve, while ACOS compares directly with a margin percentage.
TACoS
TACoS is a seller-side metric, not an Amazon report column. Amazon’s ACOS and ROAS guides do not define it. Two other Amazon Ads guides mention it in passing: the author’s guide describes it as ad spend as a percentage of total sales, and the holiday guide says it shows whether total sales, organic and paid, grow relative to ad spend (author’s guide, holiday guide). Tools phrase it slightly differently, so write down your own definition.
Advertising reports only carry attributed sales, so the denominator comes from another source. For sellers, the Selling Partner API offers a Sales and Traffic report that Amazon describes as sales and traffic metrics aggregated by date and ASIN, available to sellers only (report types). Vendors and multichannel sellers use whatever their order system provides. Amazon SP-API vs Amazon Ads API shows where each number comes from.
| Item | ACOS | TACoS |
|---|---|---|
| Denominator | Sales attributed to ads | All sales |
| Source | Advertising console or API | Your sales data |
| Answers | Is this campaign or keyword efficient? | How dependent is the business on ads? |
| Level | Campaign, ad group, keyword, product | Business, brand, product line, listing |
| Distorted by | Spending only on branded, high-intent terms | Cutting ads, which can look good until organic rank slips |
A worked example with made-up numbers
These figures are invented to show the arithmetic. They are not benchmarks and describe no real account. Assume a 30% margin before ad costs.
| Item | Month 1 | Month 2 |
|---|---|---|
| Ad spend | $2,000 | $3,000 |
| Ad-attributed sales | $8,000 | $10,000 |
| Total sales | $20,000 | $26,000 |
| Organic sales (total minus attributed) | $12,000 | $16,000 |
| ACOS | 25% | 30% |
| ROAS | 4.0 | 3.33 |
| TACoS | 10% | about 11.5% |
| Profit after ads (30% of total sales minus spend) | $4,000 | $4,800 |
ACOS rose from 25% to 30%, which equals the 30% margin, so the ads break even on their own attributed sales. TACoS barely moved. ACOS alone says cut the extra spend. TACoS alone says relax. Look at the increment: the extra $1,000 of spend produced $2,000 of extra attributed sales, a marginal ACOS of 50%, above the margin. Month 2 still earns more because organic sales rose by $4,000.
Whether the ads caused that rise is the question neither ratio answers. If organic sales had stayed at $12,000 for unrelated reasons, total sales would be $22,000 and profit after ads would be 30% of $22,000 minus $3,000, which is $3,600, below month 1. Same spend, same ACOS, opposite conclusion. Treat organic lift as a hypothesis to test, not credit to assume.
When should you use ACOS and when TACoS?
Use ACOS for decisions inside the advertising account:
- Bids for a campaign, ad group or keyword.
- Comparing two campaigns that target the same product.
- Finding targets that spend without selling.
- Checking break-even. Amazon’s guide says ACOS has to be lower than your profit margin to stay profitable. Use a margin that already removes fees and product cost, measured before ad spend.
Use TACoS for decisions about the business:
- Whether ads are a growing or shrinking share of revenue.
- Whether a launch investment turns into organic rank and sales, which should pull TACoS down as organic sales grow.
- Setting an overall ad budget as a share of sales.
- Reading what pausing ads does to the whole listing.
There is no universal target for either. Amazon’s guide says there is no single good ACOS, that it depends on industry, company size and goals, and that new campaigns often start high.
How do attribution windows and halo sales change ACOS?
Last touch is the standard
Amazon’s multi-touch announcement describes its traditional attribution as last touch, crediting a conversion to one ad click or view, usually the last ad encountered on the path to purchase. The same page, checked on 2026-10-14, describes multi-touch attribution as a beta for US advertisers that splits credit across several ads and shows multi-touch versions of ACoS to the right of the last-touch columns (announcement). Record which version each dashboard uses, because the two will not match.
The window decides which sales are counted
Attributed sales count only purchases inside a look-back window after the click or view. Amazon’s Reporting API v1 overview states that Sponsored Products base metrics use a 7-day look-back for sellers and 14 days for vendors, and that other campaign types use 14 days, except Amazon DSP campaigns with advertiser-selected windows (overview). Version 3 reports expose separate 1, 7, 14 and 30-day columns, so a figure’s column name says which window it uses.
Two consequences follow:
- Recent days are incomplete. Purchases that will be credited to a click have not arrived, so ACOS for the last few days looks worse than it will. Compare complete periods. Amazon’s holiday guide makes the same point: sales during peak events often occur days after the click.
- ACOS values built on different windows are not comparable.
Attributed data is also restated. Amazon says conversion data can change for up to 60 days back (reporting overview). A pipeline that stores daily snapshots must overwrite recent days. Amazon Ads API reporting explained covers the endpoints, and idempotent ad reporting pipelines covers the load pattern.
Halo sales
A shopper who clicks an ad for one product may buy another from the same brand. Amazon’s Reporting API v1 overview separates conversions for promoted products from conversions for halo products, which it defines as highly relevant products set by each campaign’s expansion rules. For Amazon DSP, it says base metrics now include both. Amazon Attribution, a separate tool for non-Amazon traffic, defines total conversions as promoted plus brand halo (attribution guide). Whether the sales behind an ACOS include halo therefore depends on the ad product and report, so read the definition before comparing across ad types. The Amazon DSP explainer covers how DSP reporting differs.
Halo sales also reach TACoS regardless, since they are part of total sales. A campaign that sells a related product shows up in TACoS even when the ad-attributed figure does not credit it.
Where the ratios mislead
- A low ACOS on branded search terms mostly reflects demand that already existed.
- ACOS compared with revenue margin instead of margin after fees and product cost.
- TACoS used to judge a single campaign. It is a business-level ratio.
- Yesterday’s ACOS read as final, ignoring the window.
- Mixed definitions: last-touch against multi-touch, promoted-only against total sales, ordered against shipped. Pick one and label every chart.
- Mismatched periods between ad sales, attributed by when the purchase happened, and a sales export by order date or ship date.
Amazon Ads is one of several retail media networks; what retail media advertising is explains the category.
Where Vexio fits
Vexio builds advertising data pipelines that bring ad spend and total sales into one warehouse, so ACOS and TACoS come from one stated definition every day. See the advertising data pipelines page for what an engagement covers. The campaign builder starts with Google Ads. For bulk changes to targets and budgets, see the campaign builder page, or request a demo to talk through your setup.
Sources
Checked on 2026-10-14.
- https://advertising.amazon.com/library/guides/acos-advertising-cost-of-sales (ACOS formula, ROAS as inverse, break-even, no single good ACOS; does not mention TACoS)
- https://advertising.amazon.com/library/guides/authors-guide-to-sponsored-products (passing mention of TACoS as ad spend over total sales)
- https://advertising.amazon.com/en-gb/library/guides/common-holiday-advertising-myths (TACoS mention, sales arriving days after the click)
- https://advertising.amazon.com/resources/whats-new/uncover-ad-value-across-funnel-with-multi-touch-attribution (last-touch standard, multi-touch beta)
- https://advertising.amazon.com/API/docs/en-us/guides/reporting/ads-v1/overview (look-back defaults, promoted and halo metrics)
- https://advertising.amazon.com/API/docs/en-us/guides/reporting/overview (conversion data changes for up to 60 days)
- https://advertising.amazon.com/library/guides/basics-of-amazon-attribution (Amazon Attribution total versus promoted and brand halo)
- https://developer-docs.amazon.com/sp-api/docs/report-type-values-analytics (Sales and Traffic report)
