Whether $1.20 a click is expensive depends on a number in your own P&L that you have probably never calculated.
Average cost per click on Amazon runs somewhere between $0.71 and $1.20.
That figure is the first thing every guide on this subject publishes, and it is close to useless on its own. A $1.20 click is cheap for a product carrying 45% margin and ruinous for one carrying 15%, and the number that separates those two cases is not on any benchmark page.
It is in your own numbers. Break-even advertising cost of sales is your pre-ad margin, and almost nothing about budgeting Amazon ads makes sense until you have it.
This guide covers what you can afford before what things cost, how to tell whether the spend created demand or captured it, and five ways to lower what you pay once both are settled.
At a Glance
Table of Contents
Five things to take away
- Break-even ACoS is your pre-ad margin. Nothing more complicated than that.
- A CPC benchmark cannot tell you whether a click is expensive for you.
- A profitable ACoS on branded search can still be worth close to nothing.
- TACoS measures against total sales, which is where the organic effect shows up.
- Cutting spend to lower TACoS kills the thing TACoS was measuring.
Quick Answer
Amazon advertising costs whatever your margin allows. Average cost per click sits between $0.71 and $1.20, but that range means nothing until you calculate break-even ACoS, which is simply your profit margin before advertising. A product with 40% margin can spend up to 40% of revenue on ads and break even. Below that number the sale is profitable, above it the sale loses money, and no bidding strategy changes the arithmetic. Once you have that ceiling, the second question is whether the spend created demand or captured demand you already had.
Definition
Three metrics get used interchangeably and they answer different questions.
ACoS is advertising spend divided by advertising sales. It measures efficiency inside the campaign and tells you nothing about the rest of the business.
Break-even ACoS is your pre-ad profit margin. It is a ceiling rather than a measurement, and it is fixed by your costs rather than by your bidding.
TACoS is advertising spend divided by total sales, paid and organic together. It is the only one of the three that sees the effect of advertising on the business rather than on the campaign.
A campaign can run a healthy ACoS while TACoS climbs, which means the advertising is buying sales the business was already going to make.
|
ONE CEILING, A HUNDRED PRODUCTS You have one target ACoS and it is applied across the whole catalog. Some of those products are paying for the ones that cannot carry it.
30 minutes. Bring your catalog export with cost of goods. |
1. What You Can Afford, Before What It Costs
Start here, because every other number depends on it.
Break-even ACoS is your pre-ad profit margin. If a product makes 40% before you spend anything on advertising, you can spend up to 40% of the sale price on ads and still break even on that unit.
The calculation
Take a product selling at $50.
Forty percent of $50 is $20, which is exactly the profit. Spend more than that and the unit loses money.
Run it on your own top ten products and the picture usually surprises people. Margin varies enough across a catalog that a single target ACoS applied to everything is quietly funding some items with the profit from others.
|
The ceiling is only as current as the fees underneath it. Referral rates change by category, FBA fees move with dimensions and weight, and a size tier reclassification after a packaging change can shift the number by several points without any announcement you would notice. A ceiling calculated in January is a different ceiling by June. Amazon account management keeps the inputs current, so the arithmetic stays true after the first month. |
Then work backward to a budget
Once you have a target below the ceiling, the budget follows from a revenue goal.
A $20,000 sales target at a 25% target ACoS gives you a $5,000 budget. The arithmetic is trivial. Everything hard about it happened in the previous step.
What to do
- Calculate break-even ACoS per product, not per account.
- Set the target below the ceiling, with the gap being your margin.
What to avoid
- Applying one target ACoS across a catalog with varied margin. Some items are subsidizing others.
- Setting a budget before the ceiling exists. The budget will be a guess wearing a decimal point.
2. The Benchmarks, in Context
Now the numbers are worth looking at, because you have something to read them against.
Office supplies sit near the bottom. Electronics and personal care sit above the top.
The spread is the point. A category average that ranges from $0.28 to above $1.41 is not describing your situation, it is describing a distribution you appear somewhere inside.
What the formats cost you differently
Sponsored Products appear in search results and on product pages, priced per click. They capture demand that already exists, which is why they anchor most accounts and why Amazon PPC management starts there.
Sponsored Brands sit at the top of search with a logo, a headline and three products. They carry a higher click cost and a broader job, which is category-level visibility rather than individual sales.
Sponsored Display follows shoppers off Amazon and retargets people who viewed and did not buy. Scaling past it means Amazon DSP management and a different measurement question entirely.
Cost per thousand impressions, rather than per click, applies to awareness placements where the goal is reach rather than response.
What to do
- Read the benchmark against your ceiling, not against your instinct.
- Check your own category rather than the overall average. The spread between them is wider than the average suggests.
What to avoid
- Judging a CPC as high or low without the margin number beside it.
- Treating a low CPC as a win. Cheap clicks that do not convert cost more than expensive ones that do.
3. Created or Captured
Here is the question a good ACoS cannot answer.
Some of your advertising is buying sales that were coming anyway.
Branded search is the clearest case. Somebody types your brand name, clicks your ad, and buys. The ACoS looks excellent. They were going to find you regardless, so the advertising captured a sale rather than creating one, and the money spent added nothing.
Where TACoS comes in
TACoS divides advertising spend by total sales, paid and organic. That makes it the only one of the three metrics that can see the difference.
Falling TACoS while revenue grows means the advertising is lifting organic performance. Rank improves, velocity builds, and the paid spend is carrying more than its own weight.
Rising TACoS while revenue is flat means the advertising is replacing organic sales rather than adding to them.
The trap is the obvious fix. Cutting spend lowers TACoS immediately, and it lowers it by removing the thing TACoS was measuring. The number improves and the business does not.
A note on ROAS
Return on advertising spend is ACoS inverted. Spend $100, earn $400, and it reads 4.0.
It answers the same question ACoS does, from the other side, and it inherits the same blind spot. A strong return on captured demand is still a weak use of money. That is why the measure worth watching is TACoS against revenue rather than either of the campaign-level numbers.
What to do
- Separate branded from non-branded before judging any ACoS. They are different businesses.
- Read TACoS against revenue direction, not as a number on its own.
What to avoid
- Cutting spend to improve TACoS. The metric improves and the growth stops.
- Scaling a campaign on ACoS alone. It cannot tell you whether the sale was incremental.
|
MARGIN THAT IS NOT IN THE AD ACCOUNT Cutting spend fixes the TACoS number and stops the growth it was measuring. There is margin sitting in your inventory adjustments that costs nothing to recover.
20 minutes. Bring your inventory adjustment report. |
4. Five Ways to Lower What You Pay
With the ceiling set and the incrementality question asked, these are the levers that move the number.
Harvest search terms from auto campaigns. Auto campaigns are for intelligence rather than scale. A term converting at 15% inside a campaign averaging 45% is being hidden by the terms around it, and pulling it into its own exact-match campaign lets you bid for it properly.
Add negatives aggressively. Selling premium stainless steel bottles and paying for clicks on “plastic water bottle” is budget leaving with nothing in exchange. Every dollar blocked is a dollar available for a term that converts.
Cut bids on high-click, zero-conversion terms. The search term report names them. Traffic without conversion is the most expensive thing in an account because it costs money and damages the performance signal at the same time.
Adjust bids by placement. Top of search usually converts better than product pages. Where top of search runs 20% and product pages run 50%, a placement modifier puts the money where it works.
Improve the creative on Sponsored Brands. A benefit-led headline outperforms a generic one, and a lifestyle image usually outperforms a product on white. Better click-through lowers the effective cost of the same placement.
None of these five works on a listing that does not convert. Advertising sends traffic and Amazon listing optimization decides what happens to it, which is why paying to send more people to a page that leaks is the most expensive mistake available.
What to do
- Work the list in order. Harvesting and negatives move the number fastest.
- Fix the listing first if conversion is the constraint. Bid work cannot compensate for it.
What to avoid
- Treating auto campaigns as a scaling channel. They are a data source.
- Changing five things at once. You will not know which one worked.
Cost grid
Your margin, what you can afford, and what it means in practice.
Not knowing the ceiling is where a surprising number of accounts sit, and it is the cheapest row to fix.
How Adverio Helps
Adverio is an agency that runs Amazon, Walmart and Target for consumer brands, measured on contribution margin against a baseline set before the work starts.
On advertising cost that means the ceiling gets calculated per product before any target is set, and branded gets separated from non-branded before any ACoS is judged.
Where a campaign clears its target on captured demand, the recommendation is to move the money rather than celebrate the number. Reading that across a catalog is what business intelligence is for, and it is the difference between a campaign that looks efficient and one that adds something.
Frequently asked questions
What is a good starting budget for Amazon ads?
Between $500 and $1,500 a month gets enough data to learn from without risking much, at roughly $10 to $30 per day per advertised product.
Consistency beats size. Thirty dollars a day for a month teaches the algorithm more than $900 spent in a week.
How long before Amazon ads show results?
Impressions and clicks appear within a day or two. Useful data takes two to four weeks, and a stable target ACoS usually takes 60 to 90 days.
Changing things early resets the learning. That is the most common reason an account takes longer than it should.
Should you lower ACoS or grow sales?
It depends which phase the product is in. A launch needs velocity and can carry a high ACoS to get it. An established product should be trimming toward a target below its ceiling.
Most accounts need both at once, run as separate campaigns with separate targets rather than one compromise number.
What is ACoS exactly?
Advertising spend divided by advertising sales, shown as a percentage. Amazon’s own definition of ACoS covers the mechanics.
The number that matters more is your break-even, because it tells you what ACoS you can afford rather than what you achieved.
Does lower TACoS always mean things are improving?
No, and this is the most common misreading. TACoS falls when advertising lifts organic sales, and it also falls when you simply cut spend.
Read it against revenue direction. Falling TACoS with growing revenue is health. Falling TACoS with flat revenue is a smaller business.
Closing
Calculate your break-even ACoS before you read another benchmark, because the benchmark cannot tell you what a click is worth to you.
|
You will calculate the ceiling this week and some products will not clear it. On a few of those the problem is the rating, and no bid change reaches it.
20 minutes. Bring the ASINs that fall below the ceiling. |
References
Cost data and Amazon’s fee structures change. Checked on 22 September 2026.
- Amazon advertising statistics, SequenceCommerce, 2025. https://sequencecommerce.com/amazon-advertising-statistics/
- Amazon Seller Central, referral and FBA fee schedules. https://sellercentral.amazon.com/