Most Amazon accounts are reported on ACOS. Most Amazon businesses should be judged on TACOS. The gap between those two sentences is where a surprising amount of money disappears.
The two formulas
ACOS, advertising cost of sale:
ad spend ÷ advertising-attributed salesTACOS, total advertising cost of sale:
ad spend ÷ total sales (advertising + organic)Same numerator. Different denominator. That single change moves the number from measuring a campaign to measuring a business.
What each one is actually for
ACOS is a campaign instrument. It answers: did this keyword, this ad group, this placement return more than it cost? It is the right number for setting a bid, because bids are set at that level.
TACOS is a business instrument. It answers: how much of this brand’s revenue is being bought? It is the right number for judging a quarter, because it is the only one that notices when paid traffic is quietly replacing traffic you used to get for free.
The failure ACOS cannot see
Here is the scenario that makes TACOS non-optional.
A product sells 1,000 units a month. 300 come from advertising, 700 organically. ACOS on the paid portion is a healthy 20%.
Six months later the same product still sells 1,000 units a month. Now 700 come from advertising and 300 organically. ACOS is still 20%.
Every campaign report looks identical. Every campaign target is being met. And the business is significantly worse: advertising spend has more than doubled to hold flat revenue, because organic rank decayed and paid traffic backfilled the gap. TACOS caught it. ACOS could not, because ACOS never looks at the organic side of the ledger.
Reading the direction, not the level
The absolute TACOS figure matters less than which way it is moving against revenue.
| TACOS | Revenue | What it means |
|---|---|---|
| Falling | Growing | Organic rank is compounding. This is the goal. |
| Falling | Flat | Advertising is being cut faster than organic can replace it. Watch rank. |
| Rising | Growing | Buying growth. Fine during a launch, a problem if it never ends. |
| Rising | Flat | Paying for sales you used to get free. Diagnose immediately. |
The bottom row is the one that hides behind good ACOS reporting.
Launches invert the rule, temporarily
A new product has no organic rank, so every sale is bought. TACOS on a launch can sit at 50% or higher and be entirely correct. You are purchasing the sales velocity that produces organic rank, and the rank is the asset you are actually buying.
What matters is the curve. TACOS on a launched product should fall steadily over the following months as organic takes over. If it plateaus high, the product is not building rank and the launch has become a subsidy.
This is also why launch TACOS should be reported separately from the mature catalog. Blending them hides both.
Where branded search fits
Branded terms distort both numbers, and they distort them in opposite directions.
Branded campaigns convert extremely well, so folding them into your ACOS makes overall advertising efficiency look better than the acquisition work actually was. That is why I separate branded defence into its own campaigns and report it on its own line. The campaign structure piece covers how.
For TACOS, branded advertising is more subtle. Some of it is genuinely defensive: a competitor conquesting your brand name will take sales if you do not bid. Some of it is buying customers who typed your name and would have found you anyway. The honest test is to pull branded spend on a subset of ASINs for a period and see whether total sales actually move.
What to report
A monthly report that shows only ACOS is incomplete. The minimum useful set:
- ACOS, by campaign type, with branded separated from non-branded
- TACOS, for the account and split between launched and mature products
- Total sales, so TACOS has context
- Organic share of sales, which is the thing TACOS is really tracking
Four numbers. Together they make it very difficult to report a good month that was not one.