What is a good ACOS on Amazon?
Most sellers ask this hoping for a single number. There is one — the industry average sits around 30% — but using it as your target is how profitable accounts end up losing money.
First, the formula
ACOS — advertising cost of sale — is what you spent on ads divided by the sales those ads produced:
ACOS = Ad spend ÷ Ad sales × 100
Spend $200 and make $1,000 in ad sales, and your ACOS is 20%. ROAS is the same relationship upside down: $1,000 ÷ $200 = 5x. A 20% ACOS and a 5x ROAS are the same thing said two ways, so there is no point arguing about which to use.
The number everyone quotes
Several independent 2026 datasets land in roughly the same place for Sponsored Products across the whole marketplace:
| Dataset | Median / average ACoS | Average CPC |
|---|---|---|
| Ad Badger (Jan–Aug 2026 app data) | 29.6% | $1.22 |
| Autron (2026 medians) | 34% | $1.13 |
| SellerPlex (38 accounts, Aug 2026) | 38% median | $1.07 |
Call it 30–35%. It is a real number and it is useless as a target, because it says nothing about whether those sellers are making money.
The number that actually matters: your break-even ACOS
Your break-even ACOS is simply your profit margin before advertising. Below it you make money on an ad sale; above it you lose money on it.
Break-even ACOS = (Profit per unit before ad spend ÷ Selling price) × 100
Work it through on a $30 product:
| Line | Amount |
|---|---|
| Selling price | $30.00 |
| Cost of goods | -$8.00 |
| Amazon referral fee (15%) | -$4.50 |
| FBA fulfilment fee | -$5.77 |
| Storage, returns, overhead (est.) | -$1.50 |
| Profit before ads | $10.23 |
Break-even ACOS is $10.23 ÷ $30 = 34%. At 34% ACOS this product makes nothing. At 20% it makes $4.23 a unit. At 45% every ad sale costs you money.
This is why the industry average is a trap. A seller with 50% margins can run at 40% ACOS and print money. A seller with 20% margins running at the "average" 30% is quietly going broke.
Benchmarks by category, if you want them
ACOS varies more by category than by competence. From the same 2026 datasets:
| Category | ACoS range across datasets | Typical CPC |
|---|---|---|
| Books | 18–26% | $0.38–$0.89 |
| Food & Grocery | 21–35% | $0.56–$1.98 |
| Toys & Games | 26–28% | $0.78–$1.41 |
| Home & Garden | 28–32% | $0.74–$1.52 |
| Electronics | 25–30% | $1.11–$1.89 |
| Beauty & Personal Care | 24–36% | $1.18–$2.47 |
| Health & Household | 27–41% | $1.05–$2.89 |
| Clothing & Apparel | 30–57% | $0.69–$2.15 |
Note how wide the apparel range is — 30% to 57% depending on whose data you read. Where datasets disagree that violently, treat the benchmark as a rough sanity check and nothing more.
Why a low ACOS is not automatically good
This is the mistake we see most often in audits. A seller proudly reports 12% ACOS, and when you look at the account, spend has been cut to almost nothing. They have optimised themselves into a corner: profitable, but invisible, losing rank to competitors who are willing to buy traffic.
If your ACOS is well below break-even and your sales are flat, you are not efficient — you are underinvesting. The right move is usually to raise bids on the terms that convert and accept a higher ACOS in exchange for volume.
TACOS: the number that catches what ACOS misses
TACOS is ad spend as a share of total sales, not just ad sales:
TACOS = Ad spend ÷ Total sales × 100
ACOS only sees the sales ads produced. TACOS sees the whole business. If ACOS stays flat but TACOS falls, your ads are driving organic rank — new customers are finding you without clicking an ad. That is the signal a campaign is actually working.
A healthy, established product often sits at 5–10% TACOS. During a launch, 25–40% is normal and correct, because you are buying rank, not profit.
So what should you target?
Three different jobs need three different numbers:
- Launching. Target above break-even on purpose. You are paying for reviews, rank and the sales velocity that drives organic placement. Expect 40–70% ACOS and treat it as a marketing cost with an end date.
- Defending an established product. Target 50–70% of break-even. On the $30 example above, that is 17–24%. Profitable, and still buying enough traffic to hold position.
- Harvesting a product you will discontinue. Target well below break-even and let rank slide. You are extracting margin, not building.
Most accounts need all three at once, because most catalogues contain products at all three stages. A single account-wide ACOS target is the surest way to overspend on your winners and starve your launches.
How to improve your ACOS without losing sales
- Pull your search term report for the last 60 days. Sort by spend. The terms at the top with no conversions are where your money is going.
- Add negatives. Exact-match negative anything with meaningful spend and zero orders. This alone often cuts 10–20% of wasted spend in a neglected account.
- Separate what converts. Move proven converting terms into their own exact-match campaign where you can bid on them properly, instead of letting them compete with everything else in a broad campaign.
- Fix the listing before raising bids. If your conversion rate is below 10%, the problem is the page, not the bid. More traffic to a page that does not convert just costs more.
- Check placement data. Top-of-search usually converts best and costs most. Adjust placement modifiers rather than blanket bids.
- Change one thing at a time and wait 14 days. Amazon’s attribution window is long enough that faster judgement is guesswork.
The short version
A good ACOS is any number below your break-even that still buys you the volume you need. For most sellers that lands between 15% and 30%, but the only way to know yours is to calculate your break-even first. Anyone quoting you a target ACOS before asking your margins is guessing.
We do this every day, in real accounts.
Our amazon advertising management team will audit yours free and show you exactly where the opportunity is. No obligation, and you keep the findings either way.