"My ACoS is 32%. Is that bad?" It is one of the first questions sellers and agencies ask us, and the honest answer is: it depends, and on its own it tells you very little. ACoS vs TACoS is not an either-or choice. The two metrics answer different questions. ACoS tells you how efficient your advertising is. TACoS tells you how dependent your whole business is on advertising. This guide walks through both formulas with numbers, shows how to work out your break-even ACoS, and explains how to read the two trends together without living in spreadsheets.
The problem: a "good" ACoS can hide a business that is getting worse
Looking at ACoS alone pushes sellers into two opposite mistakes. The first is cutting campaigns that are actually lifting organic sales. The second is celebrating a flat ACoS while total sales slide and advertising quietly becomes the only thing keeping revenue alive.
The reason is structural. ACoS only looks at sales attributed to ads. It knows nothing about organic sales, organic rank or repeat customers. That is what TACoS is for.
ACoS and TACoS formulas, with a worked example
ACoS (Advertising Cost of Sales)
ACoS = ad spend ÷ ad-attributed sales × 100
Say you spend $1,250 on Sponsored Products over 30 days and the campaigns drive $5,000 in attributed sales. ACoS = 1,250 ÷ 5,000 = 25%. ROAS is the inverse: 5,000 ÷ 1,250 = 4.0.
TACoS (Total Advertising Cost of Sales)
TACoS = ad spend ÷ total sales (ad + organic) × 100
Same month, same seller account: total sales on amazon.com were $12,500. TACoS = 1,250 ÷ 12,500 = 10%. For every $100 you sold, $10 went to advertising.
Watch the attribution window
Attributed sales are not final straight away. For sellers, Sponsored Products uses a 7-day attribution window; Sponsored Brands and Sponsored Display use 14 days. If you calculate yesterday's ACoS, you are overstating it. Always leave out the most recent two or three days.
What is a good ACoS? Start with break-even ACoS
There is no universally good ACoS. There is your break-even ACoS, which is simply your margin before advertising:
Break-even ACoS = (price − Amazon referral fee − FBA fee − landed product cost) ÷ price × 100
Example on amazon.com: a product sells for $29.99. Referral fee at 15% is about $4.50, the FBA fulfillment fee is $5.10 and the landed cost is $8.00. Margin before ads: 29.99 − 4.50 − 5.10 − 8.00 = $12.39. Break-even ACoS ≈ 12.39 ÷ 29.99 = 41%.
On amazon.co.uk, remember to strip VAT out of the price first, because your margin is calculated on the net price. Above break-even, every ad-driven sale on that product loses money. Below it, you are profitable on the ad sale. Running above break-even can make sense for a limited launch period; for a mature product it usually does not. These figures are illustrative only: rerun the maths with your real fees, which vary by category and size tier.
How to read ACoS and TACoS together
What matters is the direction of both numbers over the last 4 to 8 weeks:
- ACoS flat, TACoS falling: ads are pulling organic sales up with them. Healthy.
- ACoS flat, TACoS rising: organic sales are dropping and ads are filling the hole. Investigate now: Buy Box, stock, reviews, price.
- ACoS rising, TACoS flat: campaigns are decaying (bids too high, broad keywords, irrelevant search terms) but the business is holding up.
- Both rising: you are paying more to sell less. Top priority.
Two practical rules. Calculate TACoS per ASIN, not only in aggregate: a 10% account-level TACoS can hide one product at 35% and three at 3%. And compare like with like: amazon.com and amazon.co.uk have different fees, prices and competition.
How to check it manually
Without tooling you need three sources:
- the campaign or advertised product report from the Amazon Ads console (spend and attributed sales per ASIN);
- the "Detail Page Sales and Traffic by Child Item" report in Seller Central (total sales);
- a spreadsheet joining the two by ASIN over identical dates.
It works, but it takes 30 to 60 minutes every time and breaks easily: misaligned dates, parent and child ASINs mixed up, currencies across marketplaces.
With your AI connected: 4 prompts to copy
If your seller account and Amazon Ads are connected to ChatGPT or Claude through an MCP connector, you can run the same check in one conversation:
With Vendooly, the AI reads ad performance from the data-hub history and sales from your seller account, so the per-ASIN join is done for you. If you decide to act (lower a bid, pause a keyword, trim a budget), the AI prepares the change, shows you a before → after preview and executes it only after your explicit OK. Vendooly is not its own AI: it is the action layer that gives the AI you already use access to Amazon, through official Amazon APIs.
What to do this week
- Work out break-even ACoS for your top 5 products. That number is what makes an ACoS "high" or "low".
- Look at the TACoS trend over 8 weeks, not yesterday's value.
- For every ASIN with rising TACoS, look for the cause outside advertising first: stock, Buy Box, price.
- Only then work on campaigns, starting with keywords that spend without selling.
If you want the picture without opening reports, the Vendooly Monday recap shows ad spend and ACoS against the previous week, alongside the products that are slipping. For true margin per product, see the guide on calculating real profit by ASIN.
The 14-day free trial includes the full Seller + Ads plan, and it stays free under €2,000 of Amazon revenue per 30 days. Sign up here: vendooly.com/signup.html. All guides are in the wiki.