Key takeaways
- ACoS = ad spend ÷ ad-attributed sales. TACoS = ad spend ÷ total sales, organic included. The link between them: TACoS = ACoS × the share of sales that came from ads.
- ACoS rising while TACoS falls means total sales are outgrowing ad spend, usually because ads are feeding organic sales. Confirm that organic sales grew before crediting the ads.
- Both rising means ad spend is outgrowing sales, usually because ads are buying sales you would have got anyway or the listing is weak. Your branded share of spend and your conversion rate tell you which.
- A good TACoS is one your margin can pay for while total sales grow: break-even ACoS minus the profit margin you want to keep.
- Sponsored Products credits orders for 7 days after a click (14 for vendors); Sponsored Brands and Sponsored Display for 14. Recent ACoS looks worse than it will end up.
TACoS vs ACoS comes down to what you divide ad spend by: ACoS (advertising cost of sale) divides it by the sales Amazon credits to your ads, and TACoS (total advertising cost of sale) divides it by all your sales, organic included. Use ACoS to run campaigns and keywords, and TACoS to judge whether advertising is growing the business.
What is ACoS?
ACoS is the number for decisions inside a campaign: which keywords to bid up and which to cut.
ACoS = ad spend ÷ ad-attributed sales × 100
Worked example. You spend ₹20,000 on ads in a month and Amazon attributes ₹80,000 of sales to them, so ACoS is 25%. ROAS (return on ad spend) is the same result the other way up: ad sales ÷ ad spend = 4.
What ACoS misses: it only counts orders Amazon attributes to an ad. It can't see organic sales the ads helped create, or tell whether an attributed order would have happened anyway, as when a shopper searched your brand name and would have found you without the ad. If your ACoS is above break-even, start with why ACoS gets too high and how to fix it.
What is TACoS on Amazon?
TACoS shows whether ads are growing the business or only paying for sales you already had.
TACoS = total ad spend ÷ total sales × 100
Total ad spend is Sponsored Products, Sponsored Brands and Sponsored Display, plus DSP if you run it. Total sales is ordered product sales from Seller Central's Business Reports for the same dates and marketplace (vendors: ordered revenue in Vendor Central's sales report).
Continuing the worked example: total sales for the month are ₹2,50,000, so TACoS is ₹20,000 ÷ ₹2,50,000 = 8%. Ads drove 32% of sales (₹80,000 of ₹2,50,000), and one line links the two metrics:
TACoS = ACoS × (ad sales ÷ total sales)
Here, 25% × 32% = 8%. What TACoS misses: it can't point to the campaign that is wasting money, and it moves with things ads don't control, such as price changes, deals, stock-outs and festival demand.
TACoS vs ACoS: how to read them together
Read the two as a pair, week by week, on settled dates. Their directions tell you where to look first.
| ACoS | TACoS | What it usually means | Check first |
|---|---|---|---|
| Falling | Falling | Ads are getting cheaper per sale and total sales are keeping up | That total sales aren't shrinking: a spend cut can lower TACoS while sales fall |
| Rising | Falling | Total sales are growing faster than ad spend, usually because ads are feeding organic sales | That organic sales really grew, and why |
| Falling | Rising | Ads carry a growing share of sales while organic sales lag | Organic rank, reviews, stock, Buy Box and price |
| Rising | Rising | Ad spend is growing faster than both ad sales and total sales | Branded search spend, conversion rate and cost per click |
ACoS rising, TACoS falling: check that organic sales really grew
This usually means ads are feeding organic sales: Amazon lists sales history among the factors that decide where a product appears in search, and ad-driven orders add to that history. But total sales can also grow for reasons unrelated to ads, so check:
- Work out organic sales each week as total sales minus ad-attributed sales. Treat it as an estimate: Amazon's ad sales include orders placed days after a click and some orders of products you didn't advertise.
- Line the rise up with what changed. Higher bids or budgets on non-branded terms point to the ads; a price cut, coupon, deal, new reviews or a competitor's stock-out point elsewhere.
- If your brand is registered, check your brand's purchase share in the Search Query Performance report (Brand Analytics) for the terms you advertise. A rising share on exactly those terms supports the case; our guide to Search Query Performance and bids shows how to read it.
ACoS and TACoS both rising: find out what the ads are buying
There are three causes to rule out, each with its own check:
- Ads buying sales you would have got anyway. Split search term spend into terms with and without your brand name. If branded terms take a growing share, you are paying for shoppers who were already looking for you.
- A weak listing. Compare conversion rate on ads (orders ÷ clicks) with unit session percentage (units ordered ÷ sessions) in Business Reports. If both fell, the cause is price, reviews, images, the Buy Box or out-of-stock variants, and no bid change will fix it.
- Auction pressure. If CPC (cost per click) rose while conversion held, competitors are bidding more. Decide term by term whether the new price still fits your break-even ACoS.
To test the first cause, run a holdout. Worked example. For two weeks, pause the branded exact keywords where you already rank first organically, then compare those terms' purchases in Search Query Performance with the two weeks before. If purchases hold, those ads were mostly buying demand you already had. The risk: a competitor may take the ad slot on your brand search meanwhile, so watch who appears there.
What is a good TACoS?
A good TACoS is one your margin can pay for while total sales keep growing. No single figure fits every seller; it depends on your margin and on how much of each product already sells organically.
Break-even ACoS = margin before ad spend ÷ selling price × 100
TACoS ceiling = break-even ACoS − the net profit margin you want to keep
Worked example. A product sells for ₹999. After Amazon fees, GST and product cost you keep ₹300 before advertising, about 30%, so break-even ACoS is 30%. To keep 12% of every sale as profit, advertising can cost at most 18% of total sales. Single campaigns can run above 18% ACoS, as long as organic sales keep the blended TACoS under that ceiling. Fees change, so take them from Seller Central's current rate card, or use our Amazon profit calculator.
Two things move the right level:
- Product stage. A launch has few organic sales, so its TACoS sits close to its ACoS. Set a spend limit and a date by which TACoS should fall; see ads across the product lifecycle.
- Direction. A flat or falling TACoS while total sales grow is healthy. A falling TACoS with falling total sales can mean ads were cut too far.
For reference, portfolio TACoS on the accounts we manage is 13.5%. It is a blend across many products, so treat it as a sense check, not a target.
New-to-brand: are your ads finding new customers?
New-to-brand (NTB) shows whether ad spend is buying new customers or re-buying existing ones, which neither ACoS nor TACoS can tell you. An order is new-to-brand when the shopper hasn't bought from your brand on Amazon in the past 12 months. Sponsored Brands and Sponsored Display reports include it for brand owners.
NTB order share = new-to-brand orders ÷ attributed orders × 100
On the accounts we manage, new-to-brand order share is 69.1% (₹28.6 Cr of new-to-brand sales). Two ways to use it:
- Expect a lower share on branded campaigns, whose shoppers already know your name. A low share on a campaign meant to find new customers means it mostly reaches existing ones.
- A high-NTB campaign can justify a higher ACoS if customers reorder, as with consumables. Count the margin on expected repeat orders, not only the first one.
What NTB can't do: it only sees Amazon purchases, so a shopper who bought from your own website still counts as new, and a lapsed customer counts as new again after 12 months.
How the Amazon attribution window changes recent ACoS
The latest days' ACoS looks worse than it will end up. Spend is counted on the day of the click, but Amazon keeps crediting orders back to that day for the length of the attribution window:
- Sponsored Products: 7 days for sellers, 14 days for vendors.
- Sponsored Brands and Sponsored Display: 14 days.
Worked example. You spend ₹10,000 on Sponsored Products on a Monday. On Tuesday the console credits ₹25,000 of sales to Monday's clicks, an ACoS of 40%. Orders keep arriving all week, and by the following Tuesday Monday's sales have grown to ₹40,000, a settled ACoS of 25%. Cut bids on the first reading and you cut keywords that were on target.
Three rules follow:
- Judge bids and campaigns only on dates older than the window.
- Compare periods with the same time to settle: last week against the week before is unfair to last week.
- Reports pulled through the Amazon Ads API can use other windows for Sponsored Products, so a tool and the console can disagree. Check which window a report uses.
What about Google and Meta ads?
Google and Meta can't see Amazon orders on their own, and Amazon's ad reports don't credit them. Judge that spend with a blended TACoS: all ad spend, on and off Amazon, ÷ total Amazon sales. Where Amazon offers it, Amazon Attribution tags off-Amazon links for brand owners and reports the Amazon orders that followed.
The Amazon ads metrics an owner should check every week
Seven numbers cover a weekly review if each is tied to an action. Read them on settled dates and compare with the previous month, not yesterday.
| Metric | Where to find it | Act when |
|---|---|---|
| Total sales, split into ad and organic | Business Reports and the ads console | Organic sales keep falling: check rank, stock, reviews, price and the Buy Box before touching ads |
| TACoS | Total ad spend ÷ total sales | It goes above your TACoS ceiling: find the products pushing it up |
| ACoS by product, against its break-even ACoS | Advertised product report | A product stays above break-even on settled dates: cut bids on its worst search terms, not the whole campaign |
| Branded share of ad spend | Search term report | It rises while total sales stay flat: test lower bids on branded terms |
| New-to-brand order share | Sponsored Brands and Sponsored Display reports | It falls on campaigns meant to win new customers: move budget to category and competitor targeting |
| Conversion rate: orders ÷ clicks on ads, unit session percentage on the listing | Ads console and Business Reports | It drops on a product: check price, reviews, images and variant stock before changing bids |
| Days of stock cover: units available ÷ average daily units sold | Inventory reports | Cover falls below your restock lead time: lower bids on that product so stock lasts until the restock arrives |
During sale events, compare with the same event last year rather than the week before. For the campaign-level fixes these numbers point to, see Amazon PPC mistakes and how to fix them.
ScaleSKUs reads ad sales from the Amazon Ads API and total sales from the Selling Partner API, so ACoS, TACoS and organic sales sit on one daily trend for each account. See how it connects to Amazon.
Frequently asked questions
Is TACoS more important than ACoS?
Neither, because they answer different questions. ACoS shows which campaigns and keywords earn their spend; TACoS shows whether advertising is growing total sales. Manage campaigns on ACoS against break-even, and judge the account on TACoS and total sales.
Why does my ACoS for the last few days look so high?
Because not all the sales have arrived yet. Amazon credits orders back to the day of the click for 7 days on seller Sponsored Products and 14 days on Sponsored Brands, Sponsored Display and vendor Sponsored Products. Recent ACoS falls as those orders arrive, so judge only on older dates.
Should TACoS include Sponsored Brands, Sponsored Display and DSP spend?
Yes. Total sales include orders that every ad type helped produce, so count every ad type's spend too. Leave some out and TACoS looks better than it is.