ACoS equals ad spend divided by ad-attributed revenue. That is the complete formula.
It says nothing about Amazon referral fees, FBA fulfillment costs, return processing charges, storage, inbound placement fees, or what you paid for the product. A 28% ACoS can look healthy while every unit you ship loses money.
I ran PPC on my own seven-figure Amazon brand for two years before this clicked. We hit our ACoS targets every month. Revenue climbed. Margin stayed thin and we couldn’t understand why. The answer was sitting in fee columns we weren’t reading.
ACoS is a ratio. Contribution margin per SKU is the actual dollar amount. Managing PPC by ACoS without knowing your contribution margin is like managing a restaurant by table turns without knowing what each dish costs.
Here is the formula:
Selling price, minus COGS (product cost plus inbound freight plus prep), minus Amazon referral fee, minus FBA fulfillment fee, minus storage allocation, minus inbound placement fee, minus return allowance (return rate times return processing fee per unit), minus PPC ad spend per unit (total monthly ad spend divided by total units sold) equals contribution margin per unit.
Run the real numbers on a specific SKU. A large standard-size home and kitchen product, selling at $45:
A 39% break-even ACoS gives real room to advertise. Running 28% ACoS means $12.60 in ad spend per unit. Net contribution margin after ads: $4.94, roughly 11%. Thin, but workable.
Now run that same SKU through Q4. Peak FBA surcharges add $0.40. Storage climbs from $0.78 to $2.57 per cubic foot. CPCs spike 30% during Prime Day, pushing ad spend per unit from $12.60 to $16.38. Contribution margin: -$1.04. Negative. On a campaign with 28% ACoS that looked fine in July.
This is what happens to sellers who run seasonal campaigns without a margin stress test. The ACoS number doesn’t change. The math underneath it does.
Amazon’s individual fee increases look modest. The 2026 FBA hike averaged $0.08 per unit. Fine. Rounding error.
The problem is what stacks on top of it. The 2026 fee layers on a large standard SKU priced $10-$50:
| FBA fulfillment: | $5.73 (up $0.31 from 2025). |
| Return processing (apparel, on every return) | $1.65-$4.01 per return, no threshold exemption. |
| Inbound placement fee | $0.20-$0.70 per unit depending on shipment routing. |
| Low-inventory fee | triggered when stock drops below 28 days of supply. |
| Peak FBA surcharge (Oct 15 – Jan 14) | $0.20-$0.40 per unit extra. |
| Q4 storage | $2.40-$2.57 per cubic foot versus $0.78 off-peak. |
Every one of these was introduced or increased between 2024 and 2026. Together they can swing margin by 15 percentage points on the same SKU, year over year, with no change to your bid structure.
Apparel is the most exposed category. Online return rates run around 24%. Amazon charges return processing fees on every apparel return with no exemption threshold. A $45 clothing item at a 24% return rate generates roughly $0.48-$0.96 in return processing costs per unit sold, on top of referral and fulfillment fees that are never refunded regardless of why the customer returned it. That original $5.73 FBA fee and $6.75 referral fee: gone on roughly 1 in 4 units.
46% of Amazon sellers operate at 11-25% net margins. Given this fee structure, many don’t realize that fees alone can consume that entire range on certain SKUs.
Most sellers reverse the order of operations here.
The typical approach: find an industry benchmark for TACoS (10-15% for mature sellers, 20-25% for growth phase), then manage bids toward it.
The correct approach: calculate contribution margin per SKU first, then derive the maximum TACoS that margin can support.
If TACoS falls below your organic contribution margin, total ad economics are profitable. If TACoS exceeds your contribution margin, ads are subsidizing negative returns across both paid and organic units. Your TACoS ceiling is a mathematical output of your unit economics, not a benchmark from a blog post.
A seller with 35% contribution margin can absorb 22% TACoS and keep 13 points. A seller with 18% margin running the same TACoS target is destroying the business. The same number that looks conservative in one account is ruinous in another.
Set SKU-level TACoS ceilings from your margin calculation, then segment your catalog into three tiers.
This is how the team at ALFI structures bid decisions across every account. Contribution margin per SKU drives tier assignment. ACoS stays visible as a secondary data point.
If you operate a Shopify store and also sell on Amazon, the fee comparison is direct.
Shopify’s total cost on a $50 sale: roughly $2-$5. Amazon’s total cost on the same $50 sale: $7.50 referral plus $5.73 FBA plus ad spend, so $15-$20 before margin. That’s $29.77 net on Amazon versus $42.50 net on Shopify for the identical product.
The mistake is treating Amazon as a margin channel. For most Shopify-first brands, it isn’t.
Frame it this way: Amazon’s 20-35% total cost structure is your blended CAC for accessing a marketplace with 200 million Prime subscribers. If a customer acquired on Amazon then makes a second purchase on your Shopify store at 5% total transaction cost instead of 35%, the lifetime value math flips entirely.
Use this filter to allocate SKUs across channels:
Amazon and Shopify serve different functions in the same customer journey. The contribution margin calculation tells you exactly where each SKU belongs.
Run this before every Prime Day and before Black Friday. It takes 30 minutes.
One thing most sellers miss: return processing fees from Q4 hits accounts in January and February, two to three months after the return. You discover the problem when the charge appears, not when the advertising decision was made.
Does this framework work if I don’t have exact COGS by SKU?
Start with estimates. Pull your last purchase order, carrier invoice, and any prep costs. An approximation close to accurate will still identify your high-margin and low-margin SKUs. You don’t need accounting-software precision to start making better bid decisions.
What should I do if contribution margin per unit is under $3 on my best-selling SKU?
Don’t run growth PPC on it. Run exact match on brand and competitor terms only. Redirect that budget to a higher-margin SKU. Also audit the FBA tier: is the SKU in a larger fulfillment tier than necessary? A packaging change that drops one size tier can add $1.50-$2 back to margin without touching price.
How does this apply if I sell on both Amazon and Shopify?
Run the contribution margin calculation separately for each channel on the same product. Most brands find their Shopify margin is 15-20 points higher than Amazon on identical SKUs. That gap is your Amazon CAC. The question is whether lifetime value justifies that acquisition cost. For repeat-purchase products, it usually does.
My ACoS has been on target for months. Do I actually need to change anything?
If you don’t know your contribution margin per SKU, you don’t know whether those ACoS targets are producing profit or subsidizing losses. Run the formula on your top five SKUs. The results will answer the question faster than any benchmark can.
What’s a realistic contribution margin target before factoring in ad spend?
For a SKU you plan to scale aggressively: aim for $8 or more per unit before ad costs. That gives enough room to absorb meaningful ad spend while keeping net margin positive. Below $4 per unit pre-ad, meaningful growth spend rarely fits the math without pricing or cost changes first.
Naeela Shah is the Founder of ALFI, an Amazon growth agency capped at 18 clients. She previously built and exited a seven-figure Amazon brand.
