Neither metric, when considered independently, should serve as the sole basis for decision-making.

That answer irritates about half the people who ask because the ACoS vs TACoS argument is rarely about measurement. It is a proxy. One side wants to cut ad spend, the other does not, and each picks the metric that supports them.

Here is the position this ACoS vs TACoS piece defends: ACoS is a control. It tells you what to do with a bid. TACoS is a diagnostic. It shows whether your business is becoming less dependent on advertising. Neither authorizes a decision on its own, because the key metric is the contribution margin per unit, considered against the product’s life cycle stage.

Below is the evidence, including the arithmetic that shows exactly where each metric lies for you.

What ACoS measures, and what it cannot see

ACoS is ad spend divided by ad-attributed sales. Spend $200, generate $1,000 in sales that Amazon credits to those ads, and your ACoS is 20%.

That is the whole metric. The interesting word is “attributed.”

Amazon credits a sale to an ad inside a fixed post-click window. Anything outside it does not count, and neither does anything the ad accomplished indirectly. The shopper who clicked, left, and came back three weeks later through organic search. The halo sale is on a different SKU. The rank improvement the click bought you. ACoS sees none of it.

That is not a defect. It is the question ACoS was built to answer. It measures campaign efficiency in a vacuum, and a vacuum is the correct environment for a bid decision. When you are working out whether a search term deserves $1.40 or $0.90, the halo effect is noise. You want the clean number.

The trouble starts when someone reads a vacuum measurement as a business metric and then acts on it. 

What TACoS measures, and why Amazon does not report it

TACoS is total ad spend divided by total revenue, including both organic and paid channels.

Run the same numbers. Spend $200, take $1,000 in ad-attributed sales, and add $1,000 in organic sales on top. ACoS is still 20%. TACoS is 10%. Nothing changed about the advertising. What changed is what you measured it against.

Total revenue sits in Business Reports. Ad spend sits in the advertising console. You pull both and divide. That is the real reason most sellers have never looked at TACoS: it is a number you have to want.

What makes it worth wanting is a quirk of Amazon that does not exist on other advertising platforms. Run Google Ads at a product and make sales, and your Google organic ranking does not move an inch. Run Sponsored Products on that same product and convert on a keyword, and Amazon hands you organic rank on that keyword. Ad sales buy organic sales.

So a falling TACoS at a flat ACoS means the flywheel is turning. The same advertising now sits atop a larger organic base. A rising TACoS at a flat ACoS means the reverse, and ACoS will never once warn you about it.

ACoS vs TACoS: The Debate Is Real, and Both Camps Are Half Right

This is a genuinely contested question, and anyone who tells you it was settled years ago is selling something. Three positions are worth knowing.

The tool and education sides of the industry have argued for years that TACoS is the metric that matters, and that argument is sound. Helium 10 frames it as a budgeting question: if you treat PPC as your marketing budget rather than a transaction cost, the only sensible denominator is all of your sales, because PPC is what got the product indexed and ranked in the first place.

The second camp concedes what the first usually skips. Jungle Scout’s own coverage of TACoS grants that monitoring and optimizing ACoS at the campaign level are equally important, which they have to be, because there is no such thing as a TACoS at the keyword level. You cannot manage a bid using a metric that does not exist at the account level.

The third camp is smaller, works mostly out of PPC software companies, and thinks the first two are arguing about ratios when they should be arguing about dollars. Their position: ACoS and TACoS are both denominated in revenue, which means both are structurally blind to what a unit actually earns. The real number is the per-unit contribution margin after ad spend, and the correct KPI varies by lifecycle stage rather than by preference.

We think the third camp has it, and that the first two are each arguing that their half of the answer is the whole answer.

Amazon declines to take a side, which is instructive. Amazon’s own guidance on ACoS says to work out the break-even ACoS against your profit margin first, then decide whether the campaign goal is sales or awareness, and it states plainly that there is no definitive “good” ACoS because it depends on industry, company size, and campaign frequency. On TACoS, Amazon says nothing at all, because Amazon does not publish it.

ACoS vs TACoS: The Decision Table

Seven criteria. Read the fourth column, because that is the one every other version of this table refuses to fill in.

Two rows there do most of the work, and neither shows up in the arguments people actually have about this.

Row six is why a metric can be perfectly honest and still be useless: ACoS answers to whoever is reporting it. Row seven is why most TACoS analyses are quietly wrong before anyone reads them. The rest is bookkeeping.

Where each metric lies to you, in arithmetic

Take a product doing $10,000 a month in ad-attributed sales on $2,500 of spend. ACoS is 25%. Organic contributes another $10,000, bringing total revenue to $20,000 and TACoS to 12.5%. Contribution margin before advertising runs at 30%, so the product earns $6,000. After ad spend, it clears $3,500.

Now somebody decides the ACoS is too high and pulls the bids back.

Spend falls to $1,200. Ad sales fall to $5,000. ACoS is now 24%. It improved. Over the following eight weeks, sales velocity on the ranking keywords drops, the product slides down the page, and organic settles at $5,000. Total revenue is $10,000. TACoS is now 12%. It also improved.

Both metrics moved in the right direction. Revenue halved. Contribution before advertising fell to $3,000, and after ad spend, the product now clears $1,800, down from $3,500.

Nobody did anything wrong on the dashboard. Both numbers improved, and the business was roughly half as good. That is the case against optimizing either metric in isolation, and no amount of tracking would have caught it because both lay in the same direction at the same time.

TACoS has a second failure, and it is quieter. Picture an account at $500,000 per month with $50,000 in spend. The account TACoS reads 10%. Healthy. Underneath it, four hero ASINs do $400,000 on $20,000 of spend, a 5% TACoS. The other eleven products do $100,000 on $30,000, a 30% TACoS.

The heroes are carrying eleven products that would not survive on their own. The account-level number is not wrong, it is just averaging away the only thing you needed to know. The day a competitor comes for one of those heroes, the headline number moves fast, and you will not have seen it building.

Why your TACoS will not fall if your listing does not convert

Every article on this topic tells you advertising lifts organic rank. Almost none of them tell you what Amazon is measuring when it decides whether to give you any.

It is not spend. It is not bids. It is sales per click on the keyword. Amazon watches what shoppers do after they click, and it promotes the products that convert the traffic it sent. Which means the flywheel has two inputs, and the bid strategy only owns one of them.

The other input is conversion rate, which is set by your main image, six competitors in the search results, your infographics, your A+ modules, your title, your review count, and your price. Not one of those is a PPC lever.

So here is a failure mode we see constantly. A brand hires a good PPC manager. The manager restructures the account, reduces waste, and lowers ACoS from 38% to 26%. Real work, real improvement, nothing to criticize. Two quarters later, TACoS has not moved a single point because every click they bought landed on a listing that converted at 6% in a category that converts at twice that rate. The manager bought the traffic. The listing lost it. Amazon watched a product fail to convert and gave the rank to somebody else.

You cannot bid your way out of that. Doubling spend on a listing that loses the click just buys Amazon more evidence that the product does not deserve to rank.

This is why we do not sell bid management as a standalone product. Advertising and listing creative are two halves of one whole, and a partner who owns only one of them can move only half of it.

The product-level TACoS flaw nobody warns you about

Pull an advertising report, and you will find two columns most sellers never read: Advertised SKU Sales and Other SKU Sales. The first is revenue from the product you actually advertised. The second is revenue from every other product in your catalog that the shopper bought after clicking that ad.

Both roll into the ad sales figure on your dashboard.

Which means that when you calculate TACoS for a single ASIN, you are dividing that ASIN’s ad spend by revenue that partly belongs to other products. The number is not slightly imprecise. It is measuring something other than what you think it is measuring.

This matters because product-level TACoS is exactly what sellers use to decide which ASINs to cut. A product that looks like it is burning spend at 40% TACoS may be the click magnet feeding sales to three other SKUs. Cut it, and you find out what it was doing about six weeks later.

The fix is a reporting-level fix, not a metric fix. Account-level TACoS hides subsidies. ASIN-level TACoS is corrupted by other SKU sales. The level at which the number stays honest is the parent-child product family, as shown in the Purchased Product Report, which shows what shoppers actually bought after clicking your ad rather than what you hoped they would.

Break-even ACoS is the number both metrics answer to

Break-even ACoS is the point where an ad-attributed sale earns you nothing. Below it, you make money on the sale. Above it, you are buying something other than profit and should be able to say what that something is.

Worked example. A $40 product with $12 in COGS, $8 in FBA and referral fees, and $2 in other variable costs leaves an $18 contribution per unit before advertising. Spend $18 to sell one, and you break even. That is a 45% break-even ACoS.

Put a second product beside it. Same $40 price, but $28 in combined COGS and fees. The contribution is $12. Break-even ACoS is 30%.

Run both against a single 25% account-wide ACoS target, and you underspend badly on the first and squeeze the second. One target across two margin structures guarantees you are wrong on at least one, probably both.

This is why “What is a good ACoS?” has no answer, and “What is a good TACoS?” has no answer either. Both are ratios of ad spend to revenue. Neither contains any information about what the unit earns.

How each metric gets gamed

Anyone can cut your ACoS by 30% inside two weeks. Here is the method.

Pause the broad and auto campaigns. Kill the discovery targets. Move the budget onto branded search, where shoppers are already hunting for you by name and convert at rates no other traffic comes close to. Your ACoS falls off a cliff. Your dashboard has never looked cleaner.

What you bought is a report. Branded search was going to convert with or without the ad, so a large share of that spend is now paying for sales you already had. Meanwhile, new-to-brand traffic is gone, nobody new is finding you, and in ninety days your organic rank starts sliding on the keywords the discovery campaigns were feeding.

TACoS is harder to fake, and not because it is a better metric. It is because it is a worse one to lie with. Cut spend and revenue follows it down, so the ratio does not obligingly improve the way ACoS does. At product level, though, it lies without any help at all. See the previous section.

Benchmarks by lifecycle stage, and where they actually came from

These are the ranges you will find quoted everywhere.

Now the part that never gets printed underneath the table.

These ranges are not Amazon guidance. Amazon does not publish a TACoS target because it does not publish TACoS. They are operating ranges that have been restated across the industry often enough to read like a standard, and if you trace them back far enough, you find blogs citing blogs. We use them. We are printing them. They are a reasonable place to start an argument about your account.

They are not a place to finish one. A 12% TACoS on the $18-contribution product above is a completely different business from a 12% TACoS on the $12-contribution product. Same number. Opposite decisions.

What to hold a PPC partner accountable to

If you are reading this because you are about to hire someone, these are the questions that set the field apart.

Which number goes in the engagement and at what reporting level? “We will lower your ACoS” is a promise about a dashboard. “We will hold TACoS at family level while protecting new-to-brand” is a promise about a business. Only one is hard to keep.

What happens to new-to-brand traffic if ACoS drops? Anyone who cannot answer clearly plans to cut discovery spend and hopes you do not check.

What is my break-even ACoS per ASIN tier? If a candidate quotes a target before asking for your COGS and fee structure, they plan to optimize toward a number unrelated to your economics.

And the one almost nobody asks: What happens if the conversion rate is the problem? A partner who owns bids and nothing else has one honest answer: “Not my department.” Ask anyway. The answer shows whether they manage your account or their report.

Book a PPC strategy call

If your ACoS looks fine and your TACoS has not moved in two quarters, the answer is sitting in the gap between your bids and your listing” → “If you’re still weighing ACoS vs TACoS for your own account, and your ACoS looks fine while TACoS hasn’t moved in two quarters, the answer is sitting in the gap between your bids and your listing. 

Book a PPC strategy call and we will pull your family-level numbers, work out your break-even ACoS by tier, and tell you which of the two is actually costing you money.