Amazon TACOS is an output of four things: conversion rate, organic share, the cost of a click in your category, and how hard you are pushing for growth. Three of the four moved in 2026, so here is what healthy looks like now.
Strategy
Amazon TACOS is not a target, it is an output of four separate things: your conversion rate, your organic share, the cost of a click in your category, and how aggressively you are trying to grow. Change any one and the number moves without anybody doing anything wrong. Three of the four moved in 2026, so it is worth restating what healthy actually looks like now.
First, a definition, because the two words get used interchangeably and they are not two views of the same thing. ACOS measures your spend against the sales the advertising is credited with. TACOS measures the same spend against everything the brand sold that month, organic included. If that distinction is new, start with ACOS vs TACOS.
What is on this page
- The four things behind the number
- One: clicks kept getting more expensive
- Two: shoppers stopped typing the same way
- Three: Sponsored TV came within reach
- The one thing that did not move
- What healthy Amazon TACOS looks like now
- Read it per product, not per account
- How to read your own number
- What to actually do about it
The four things behind the number
This is why benchmark shopping goes wrong. Two brands with identical TACOS can be in completely different health, because one has high conversion and expensive clicks while the other has cheap clicks and a listing that barely converts. The number is the same and the businesses are not.
So the useful question is never whether your TACOS is good. It is which of the four inputs moved since last quarter, and whether that movement was a decision or something that happened to you.
One: clicks kept getting more expensive
Cost per click has risen every year since 2020 in almost every category, and the accounts we run are no exception. If your figures are a point or two worse than last year on identical work, that is not drift, it is the auction.
The consequence is uncomfortable and worth saying plainly: the account has to make up the difference in conversion rate or in organic share, because it will not make it up in bids. Bidding harder into a more expensive auction buys the same position at a worse price. Everything else in this article follows from that.
Two: shoppers stopped typing the same way
Amazon’s AI assistant and the longer, more conversational queries that come with it have widened the tail. In practice that means two things.
Your automatic campaigns are surfacing terms you would never have written, phrased the way people speak rather than the way keyword tools suggest. And exact match alone covers less of the demand than it used to, because the demand itself is spread across more phrasings.
Accounts that mine their automatic campaigns weekly are finding real volume in that tail. Accounts that set them up in 2023 and left them alone are not, and the gap between those two groups widened noticeably this year. The weekly routine for it is in the search term report guide.
A benchmark is only useful when you know which of the four things behind it is moving.
Want to know which of the four is moving in your account?
Send us the reports. We will tell you whether your number is a problem or just the market, free.
Three: Sponsored TV came within reach
Streaming inventory that used to require a DSP seat and a serious media commitment became buyable at far smaller budgets. That changes the calculation for brands that had written off video entirely.
Two cautions. It is an awareness channel, so judging it on immediate return will make it look like a failure in exactly the same way DSP does, and for the same reasons. And it belongs after search is fully worked rather than instead of it, because paying to create demand while cheaper demand sits uncaptured is the wrong order. The measurement approach that works is the one in new to brand, and the wider case for audience buying is in Amazon DSP explained.
What it means for TACOS is simple: a brand testing awareness channels will see the blended number rise for a period, and that is the plan rather than a fault. Report the two separately or you will end up cutting the thing that was working.
The one thing that did not move
Amid three changes, the fourth input stayed exactly where it was: conversion rate is still the lever with the most leverage in the whole account, and it is still the one most brands leave alone.
The reason is worth stating plainly. A bid change alters what you pay for one click on one keyword. A conversion improvement changes what every click in the account is worth, including the ones you have already bought and the ones arriving organically. When clicks get more expensive, as they did again this year, the account with the better listing absorbs it and the account with the weaker listing cannot.
That is why our first week on any new account is spent reading listings on a phone rather than reading campaign settings. The checks take an afternoon and they are in your listing is a mobile listing now.
What healthy Amazon TACOS looks like now
Those are bands rather than targets, and the width of each is deliberate. Our own best managed account runs at 8.68% total advertising cost of sale, while plenty of perfectly healthy accounts in our book sit at double that because they are buying category position on purpose.
What decides your number is contribution margin and stage, which is a calculation rather than a lookup. It takes about fifteen minutes and the method is in there is no correct TACOS.
Read it per product, not per account
Account level figures average away the information you needed. A brand at 12% across the catalogue can be carrying one product at 6% that funds three sitting at 25%, and until somebody looks per product nobody knows which is which.
Do it per product, or at least per range, monthly. Then group them into three piles: products where organic carries most of the sales, products where advertising carries most of them, and products where advertising carries most of them and always has. That third pile is the one to interrogate, because a product that has never developed an organic base after a year of support is usually telling you something about the listing, the price or the category rather than about the campaigns.
One caution on the split. A new product belongs in the second pile by design, and moving it to the third too early is how brands abandon products that were on track. Twelve months of support with no organic development is a signal. Four months is not.
How to read your own number
Read it monthly, per product or per range rather than per account, and always next to total sales. Four combinations and four meanings.
| What you see | What it means | What to do |
|---|---|---|
| TACOS down, sales up | Organic carrying more of the business | Keep doing exactly that |
| TACOS down, sales down | You shrank, the ratio flattered it | Put spend back on the terms you cut |
| TACOS up, sales up | Buying growth, usually correct | Check the contribution still works |
| TACOS up, sales flat | Costing more, producing no more | Investigate this week |
What to actually do about it
If clicks got dearer, the answer is conversion rate, not bids. A listing improvement lowers your cost per sale on every keyword at once, which no bid change can do. The checks are in your listing is a mobile listing now.
If the tail widened, mine your automatic campaigns weekly and promote what converts into exact match with a deliberate bid. Twenty minutes a week is enough.
If you are testing awareness channels, report them separately and give them eight to twelve weeks before judging, or you will kill them during the phase where they only cost money.
If nothing moved and the number still rose, look outside the account first: a competitor’s price, your review count, a coupon ending, or the featured offer slipping for part of the day. Those explain more sudden shifts than campaign settings do, and the diagnostic order is in Buy Box, stock and account health.
The honest summary for 2026 is that the same TACOS costs more work than it did two years ago. That is not a failure of management, it is the market, and the accounts holding their numbers are the ones treating conversion rate and organic share as the levers rather than bids.
Not sure whether your number is healthy or drifting?
Twenty minutes on a call and we will tell you which of the four inputs is moving against you.
Send us the reports and we will read the whole picture
Free, no card, written answer in three working days, and yours to keep either way.
Frequently asked questions
For an established brand not chasing rapid growth, high single figures to the mid teens is a common healthy range, and our best managed account sits at 8.68%. For a brand launching or taking a category position, twenty percent and above can be entirely correct. The range is wide because the number is an output of four other things rather than a target in itself.
Cost per click has risen every year since 2020 in almost every category. If your figures are a point or two worse on identical work, that is the auction rather than drift, and the account has to make up the difference in conversion rate or organic share, because it will not make it up in bids.
Conversational queries have widened the tail, so shoppers phrase things in longer and less predictable ways. In practice your automatic campaigns now surface terms you would never have written, and exact match alone covers less of the demand than it did. Accounts that mine their automatic campaigns weekly find real volume there.
It became reachable for far smaller budgets than the streaming inventory that used to require a DSP seat, which makes it worth testing once search is fully worked and your listings convert. Judge it on new to brand share and branded search movement rather than on immediate return.
Not necessarily, and forcing it down can shrink the business. What should fall over time is TACOS on a maturing product, because organic sales grow against steady spend. A brand adding new products will see the blended figure rise again, which is normal rather than a problem.
ACOS compares ad spend to the sales the ads were credited with. TACOS compares the same spend to everything the brand sold. ACOS judges campaign efficiency, TACOS judges whether the advertising is growing the business.
Higher than your mature products, deliberately, because early spend is buying the sales velocity that becomes organic rank. The number matters less than whether it is falling by month four or five, which is the sign the ranking work is actually happening.
Send the right reports. Get a straight answer back.
We read the account properly and write back with where the money is going, what it is buying, and the three things we would change first. You keep it either way.
No obligation · no card · a person replies, usually the same working day