Amazon's Real Take Rate: 45% of a $30 Sale, Fee by Fee

Most founders I talk to can quote the Amazon referral fee off the top of their head. They say 15%, they price around it, and they think they know what the channel costs. Then they open their Seller Central payout report and the deposit is roughly half of what the sales dashboard said they sold. The gap is not a mistake. It is six overlapping fees that never appear together on any single page, plus the advertising spend that no one writes down as a fee but everyone pays. This post totals all of it, line by line, on a plain $30 apparel product using 2025 US rates.

The short version: direct Amazon fees run about 34% of the sale price before you spend a dollar on ads, and the all-in take rate for a typical private-label brand lands at 45-55% of gross revenue. That is not an Amazon-is-evil argument. It is a pricing and channel-mix argument, because every COGS target and price you set without the full stack in view is leaving real margin on the floor.

The six Amazon fee layers most brands never total up

Amazon's fee structure is not one number. It is a stack. Here is every layer on a 1 lb apparel item in a 10x7x4 inch box selling at $30.

The referral fee is Amazon's cut of the sale price. For clothing, Amazon uses a bracket system: a $30 item falls into the above-$20 bracket, which charges 17% on the entire $30. That is $5.10, not the 15% most founders quote and not a marginal blend. The FBA fulfillment fee (the pick, pack, and ship charge) is about $4.26 for this item, which sits in the large-standard size tier because the shortest side is 4 inches and small-standard caps at 3. Storage runs $0.87 per cubic foot from January to September and jumps to $2.40 in the October-December peak; for this box that is roughly $0.14 per unit per month off-peak.

Then come the three most brands miss. The inbound placement fee, introduced in March 2024, charges $0.27 to $0.41 per unit if you ship to a minimal number of warehouses, or $0 if you use Amazon-optimized splits and accept the logistics complexity. Returns processing for apparel is $1.78 per returned unit with no exemption threshold, so a 20% return rate adds about $0.36 to every unit sold. And the low-inventory surcharge can add up to $0.91 per unit when your days of supply drop below 14, though it disappears entirely if you keep stock above the 28-day line.

When I talk to founders running a brand at $5M to $20M on Amazon, the moment that lands hardest is when they see returns and placement as separate line items. One operator I worked with had built his entire price model around "referral plus FBA" and had never once accounted for the returns fee on a category running 25% returns. That single omission was quietly eating three points of margin.

Fee	2025 Rate	$ on $30 Sale	% of Sale
Referral fee (17% bracket, full price)	17% on $30	$5.10	17.0%
FBA fulfillment (apparel 1 lb, large std)	~$4.26/unit	$4.26	14.2%
Storage (off-peak avg, 0.162 cu ft)	$0.87/cu ft/mo	$0.14	0.5%
Inbound placement (minimal splits)	$0.27-$0.41/unit	$0.30	1.0%
Returns processing (20% return rate)	$1.78/returned unit	$0.36	1.2%
Low-inventory surcharge (if triggered)	up to $0.91/unit	$0.00	0.0%
Subtotal direct Amazon fees		$10.16	33.9%
Advertising (12% TACoS estimate)	12-18% TACoS	$3.60	12.0%
Total all-in Amazon take		$13.76	45.9%

Referral fees: the bracket method that changes the math

Here is the single most common pricing error I see. A founder reads "15% referral fee," applies 15% flat, and moves on. For apparel, the real fee is not 15% and it is not a blended rate that comes in lower than expected. It is higher.

Amazon's apparel schedule lists rates of 5%, 10%, and 17% at different price thresholds, but this is a bracket system, not a marginal one. A $30 item clears the $20 threshold, so the entire $30 is charged at 17%. That is $5.10 (our average Amazon referral fee by category breakdown has the full schedule). The mental model of a blend down to some sub-15% effective rate is wrong. The actual referral fee on most apparel items priced above $20 is more expensive than the 15% most founders use in their models.

The bracket structure does create a real pricing decision: an item sold at $19.99 falls into the 10% bracket and incurs $2.00 in referral fees. The same item at $21 triggers the 17% bracket, raising the fee to $3.57. That $1.57 jump in fees on a $1.01 price increase is the threshold every apparel brand should know. When we've helped brands rebuild an Amazon P&L, this bracket boundary is almost always where the "we thought we were making money" story starts to unravel.

The practical move: know which bracket your price point falls into, not what the blended rate looks like. Price below $20 to stay in the 10% bracket if your product can support it, or price well above $20 so the fixed 17% is spread over a larger base. The bracket, not the blend, is the lever.

The advertising spend no one calls a fee

This is the layer that turns a 34% direct-fee load into a 45-55% take rate. Amazon search is increasingly pay-to-play. If you are not a dominant, well-reviewed listing, holding your Best Seller Rank means running Sponsored Products, and that spend behaves exactly like a platform fee: mandatory, scaling with revenue, and impossible to switch off without watching your rank slide.

TACoS (total advertising cost of sale, meaning ad spend divided by total sales, not just ad-attributed sales) for apparel runs 10-20% in 2025, with platform average ACoS around 25-35% on Sponsored Products. The uncomfortable finding is that small brands spend roughly three times more of their revenue on ads than large brands do. A dominant listing coasts at 5-8% TACoS on the strength of organic rank and reviews; a launch-stage private label burns 20-30% to buy the same visibility. The fee schedule is nearly identical for both. Advertising is the entire difference in effective take rate.

When I talk to founders about their TACoS, the framing I use is blunt: what is your TACoS, and what margin do you have left after fees to fund it? If a brand is running 25% TACoS on a product that only clears 20% contribution after direct fees, that is not a growth investment, that is a subsidy to Amazon with your name on the invoice. The number has to be sized to the margin that survives the fee stack, not to a rank ambition.

Amazon's own disclosures show where this is heading. In FY2024 the company reported $156.1B in third-party seller services and $56.2B in advertising services, a combined $212.3B, up 13.6% year over year. Advertising is the smaller half in absolute dollars but the faster-growing one, which means the "invisible" component of the seller take rate is the part that keeps expanding.

Fiscal Year	3P Seller Services ($B)	Advertising ($B)	Combined ($B)	YoY Growth
FY2021	103.4	31.2	134.6	-
FY2022	117.7	37.7	155.4	+15.5%
FY2023	140.1	46.9	187.0	+20.3%
FY2024	156.1	56.2	212.3	+13.6%

Why the take rate is higher for small brands than for enterprise brands

The take rate is not one number, and the spread between seller types is almost entirely an advertising story. Direct fees are roughly flat across the board because everyone pays the same referral and FBA schedule. What moves is how much ad spend you need to buy visibility you have not yet earned organically.

A large brand with review moats and organic rank pays maybe 20% in direct fees, 5% in ads, and 3% in ancillary costs, for a roughly 28% all-in take. A typical FBA brand pays 30% direct, 12% ads, 3% other, landing near 45%. A launch-stage private label pays similar direct fees but 18% in ads, pushing past 50%. And a brand sitting on slow movers, triggering aged-inventory and low-inventory surcharges while running high TACoS, can see 65%.

This is worth sitting with, because it inverts the usual instinct. Scale does not get you a volume discount on the fee schedule. It gets you off the advertising treadmill. The brands paying the lowest effective take rate are not the ones negotiating rates; they are the ones that no longer have to pay for rank.

The referral fee is the number Amazon shows you. The take rate is the number you actually live with. For most private-label brands the difference between them is advertising, and advertising is the one lever that scales with how much demand you have not yet earned. Price to the take rate, not the referral fee.

When Amazon clears at a 50% take rate, and when it doesn't

A 45-55% take rate is not automatically a bad deal. Amazon is renting you demand you would otherwise have to buy through Meta, Google, and a CAC line that can run 20-30% on its own. The question is never "is the take rate high," it is "does my product clear at this take rate." (We run the head-to-head math in our Amazon vs Shopify margin comparison.)

The products that clear are consumables with genuine repurchase urgency and high-margin items with low advertising dependence. When someone runs out of a consumable, they do not want to wait five days for a DTC shipment, so Amazon's Prime logistics is a real feature they will pay for, and the repeat purchase amortizes the acquisition cost across many orders. When I talk to founders with a highly consumable product, my default is often just be on Amazon, because the total addressable demand sitting on the platform outweighs the take rate. I have told operators who were fretting about channel diversification to slow down, precisely because their TACoS was strong and there was clearly more untapped demand on Amazon than they were capturing.

The products that struggle are one-time purchases with high return rates and thin margins, the exact profile of a lot of mid-priced apparel. And there is a quieter dynamic worth naming: even DTC-first brands often find their repeat customers migrating to Amazon on their own, which means the channel captures your best buyers whether you court it or not. That is an argument for pricing your Amazon listing to clear a full take rate, because it may end up carrying more of your revenue than you planned.

The practical moves to lower your effective take: keep inventory above 28 days to kill the low-inventory surcharge, use Amazon-optimized splits to zero out the placement fee if you can handle the logistics, price with the referral tier in mind, and above all size your TACoS to the margin that survives the fee stack. None of that is a spreadsheet you build once. It is the ongoing work of knowing your real contribution margin per SKU, which is exactly the work most brands skip until the payout report forces the question.

Related reading. For the rest of the Amazon fee stack, see the all-in take-rate breakdown and FBA fees as a share of revenue. For how we help brands model margin and cash, see our fractional CFO work.

Sources and methodology

Amazon Seller Central fee schedules (2025 US marketplace). All six fee layers are drawn from Amazon's published rate cards: referral fees, FBA fulfillment fees, and storage, inbound placement, returns processing, and low-inventory surcharge schedules. All figures are 2025 US rates in USD, modeled on a 1 lb apparel item in a 10x7x4 inch box at $30. EU and Canada rates differ.

Amazon.com Inc Form 10-K, FY2024 (filed February 2025). Segment revenue line items for third-party seller services ($156.1B) and advertising services ($56.2B) are taken from Amazon's annual report on SEC EDGAR. Amazon does not disclose third-party GMV in its filings, so any implied take rate against GMV relies on third-party estimates.

Marketplace Pulse seller P&L analysis. The canonical decomposition of Amazon's effective take rate on seller revenue, including the widely cited 50%+ figure for private-label sellers, comes from Marketplace Pulse. Our model uses a more conservative 12% advertising midpoint, which places the typical all-in range at 45-55% rather than their 50-60%; both are defensible depending on the advertising assumption.

Advertising and TACoS benchmarks. Apparel TACoS ranges and the finding that small brands spend roughly 3x more of revenue on ads than large brands are drawn from published agency benchmark guides and dated trade press, including Forbes coverage of Amazon ad spend by brand size.

Seller profitability survey data. The margin distribution (57% of sellers above 10% net margin, 28% above 20%) comes from the Jungle Scout State of the Amazon Seller 2025 report, a survey of roughly 1,500 sellers.

Operator context. The founder-facing framing throughout reflects patterns from our own advisory work with consumer brands selling on Amazon at the $5M to $50M range, anonymized. No client figures or identities are disclosed.

Frequently asked questions

what percentage does amazon actually take from each sale?

For a typical private-label FBA brand, the all-in take runs 45-55% of gross revenue once you total referral fees, FBA fulfillment, storage, placement, returns, and the advertising you need to hold rank. Direct Amazon fees alone are about 34% on a $30 apparel item. The referral fee everyone quotes is only one layer of six.

what is the referral fee for clothing on amazon in 2025?

Clothing and accessories priced above $20 are charged 17% on the entire sale price, not a marginal blend. On a $30 item that is $5.10, which is higher than the 15% most sellers assume. Items priced below $20 use lower brackets (5% on the first $15, 10% from $15.01 to $20), but once the price crosses $20 the 17% bracket applies to the whole transaction.

does amazon advertising count as part of amazon's fees?

Not officially, but for most brands it functions as one. If you need Sponsored Products spend to hold your Best Seller Rank and stay on page one, that spend is a cost of selling on the platform, not optional marketing. That is why we model it inside the take rate rather than beside it.

what is a good tacos for amazon apparel?

For a growing apparel brand, 10-15% TACoS (total ad spend divided by total sales) is a sustainable target. New launches often run 20-30% to buy rank, and dominant products can operate at 5-8%. The right number is whatever your post-fee margin can actually absorb without going negative.

how much does the fba fulfillment fee cost for apparel?

A 1 lb apparel item in a 10x7x4 inch box lands in the large-standard tier at about $4.26 per unit in 2025. That box does not qualify for small-standard because the shortest side (4 inches) exceeds the 3-inch cap. Apparel adds roughly $0.40 over the non-apparel rate for the same box.

what is the amazon inbound placement fee and how do i avoid it?

Introduced in March 2024, it charges $0.27 to $0.41 per standard-size unit when you ship to a minimal number of Amazon locations. You can drop it to $0 by using Amazon-optimized splits, which means shipping inventory to more warehouses. You are trading logistics complexity for the fee either way.

is selling on amazon still profitable in 2025?

It can be, but the margin is thinner than the headline referral fee suggests. In a 2025 seller survey, 57% reported net margins above 10% and 28% reported 20% or more, which means roughly 43% are at 10% or below. Consumables and high-margin products with low ad dependence clear best.

how do amazon fees compare to selling on my own shopify store?

On your own store you swap the 45-55% Amazon take for payment processing (about 3%), your own fulfillment (often 10-15%), and your own customer acquisition cost, which for many brands runs 20-30%. Amazon buys you demand you do not have to acquire. The question is whether that demand is worth the take rate on your specific product.