Key points
- ACoS and ROAS measure the efficiency of the ads themselves; TACoS measures how much the whole business depends on them. They need to be read side by side. [1]
- There is no industry answer for a good ACoS. It depends only on your own break-even line, which is your margin before advertising. [1]
- Orders driven by ads take days to be fully counted. Figures still inside that window are not a basis for changes. [3]
Three ratios, one question each
Marketplace sellers see many ad metrics every day; three of them decide whether advertising makes or loses money. All three have ad spend on top. The difference is what sits underneath. [1]
| Metric | Formula | The question it answers |
|---|---|---|
| ACoS | Ad spend ÷ sales attributed to ads | What it costs in ads to make 100 in ad sales |
| ROAS | Sales attributed to ads ÷ ad spend | What each 1 of ad spend brings back; the inverse of ACoS |
| TACoS | Ad spend ÷ total sales, organic included | What share of the whole business's revenue goes to advertising |
Helium 10 gives an example: $1,000 of spend against $5,000 of ad sales is a 20% ACoS; if total sales in the period were $20,000, TACoS is only 5%. [2] Both numbers describe the same spend and can point to opposite conclusions, which is why they belong next to each other.
ACoS is the Amazon term; the ad consoles of Shopee, Lazada and TikTok Shop more often show ROAS or ROI. The names differ and the arithmetic is the same, so everything below applies.
Work out your own break-even line first
Break-even ACoS equals your margin before advertising. If a product keeps 30% after cost, platform commission and fulfilment, then at a 30% ACoS an ad order neither makes nor loses money; the matching break-even ROAS is 100 ÷ 30, about 3.33. [1]
So the reference for judging an ad is this line, not a peer average. A supplement with a 60% margin and an electronics accessory with 25% can bear very different ACoS. Count returns, creator commission and voucher costs in the margin, and recalculate when costs or fees change.
ACoS holds steady while the business gets worse
ACoS counts only orders attributed to ads and cannot see organic orders. Seller Labs describes a typical case: ACoS steady at 25%, ROAS at 4.0, everything apparently fine, while TACoS climbs to 22%. Organic sales are shrinking and the store is leaning harder on paid traffic. [1]
TACoS is better read as a trend than as a single value: a rise from 8% to 11% is a signal to investigate, a fall from 14% to 11% means organic orders are taking over. [2] On what range is reasonable, the sources disagree. Helium 10 puts most established brands between 8% and 15%, and accepts 20% to 30% for 60 to 90 days during a launch [2]; Seller Labs suggests a target of 50% to 70% of break-even ACoS. [1] Both are starting points; the answer still comes back to your own margin.
Do not conclude inside the attribution window
Sales in an ad report are not live figures. On Amazon, Sponsored Products credits orders placed within 7 days of a click; Sponsored Brands and Sponsored Display use 14 days. [3][5] Attribution can take up to 12 hours to update, and cancelled orders and invalid clicks are removed within 72 hours. [3]
Yesterday's ACoS is therefore almost certain to look high and to settle over the following days. Helium 10's advice is to evaluate Sponsored Products campaigns only after at least 8 days and Sponsored Brands after at least 15. [3] Other platforms use windows of their own, and TikTok has a separate explanation of attribution for GMV Max [4]; confirm this in your platform's help pages before a review.
Set targets by purpose and cadence by level
Ads do different jobs and should not share one target. Ads on your own brand terms should run well below break-even; ads competing for a rival's traffic can accept a higher ACoS; during a launch, the thing to watch is whether TACoS falls month by month. [1]
The cadence is layered too. ACoS for a single campaign can be checked often, to adjust bids and negative keywords. TACoS reflects the whole business; weekly readings are too volatile, so review it monthly and adjust direction quarterly. [2]
Five questions to answer in every review
Brought into a routine review, all of this comes down to five questions.
- Is the period you are looking at already outside the attribution window?
- How does each main product's ACoS compare with its own break-even line?
- Is TACoS up or down against last month and last quarter? How has the organic share changed?
- For the campaigns that spend the most, is each one there for profit, defence or acquisition, and does its target match?
- Did you wait long enough after the last change, and is the result written down?
The answers come from reading the ad report against the store report. Aurovia Agent can read both for the stores you have authorised, set these ratios out product by product in a report, and note the source and time of every figure.
Sources
This article draws on the public material below. The ranges and rules of thumb come from tool vendors and are for reference only; for attribution rules, each platform's current help pages are the authority.
- Seller Labs: TACoS vs ACoS vs ROAS Explained Simply
- Helium 10: Why Does Your Amazon ACoS Look Great But You're Still Losing Money?
- Helium 10: What Are Amazon Attribution Windows, and Why Are They Important?
- TikTok for Business: About attribution for GMV Max
- Amazon Ads: How to get started with Sponsored Brands video ads
Sources checked on
Ongoing tasksSample data