Amazon often takes 25% to 45% of the sale price once you add up the main seller fees. If you are trying to figure out how much does Amazon take from sellers, the real question is whether your margin can survive Amazon’s fee structure at all.
A lot of products look profitable on paper and fall apart after fees. The problem is usually not one giant charge. It is a stack of smaller charges that hit at different moments, which makes the business look healthier than it really is.
The fix is simple, but not easy. You need to treat Amazon like a unit economics exercise, not just a sales channel.
So how much will you really pay to sell on Amazon?
A founder launches a product at $30, sees steady sales, and assumes the business is working. Then the payout hits, and the gap between revenue and profit becomes obvious.
Amazon rarely takes one clean percentage. Your real cost comes from a stack of charges tied to how you sell, what you sell, and how the order gets fulfilled. For many products, the all-in cost ends up in the 25% to 45% range of the sale price once those pieces are added together.
That is why pricing from topline revenue leads to bad decisions.
Practical rule: Never price an Amazon product from gross revenue. Price from expected payout.
The choice simplifies to this. Start with the fixed cost of your selling plan. Add the referral fee Amazon takes on each order. Then add the operating costs created by fulfillment, storage, and the product itself.
That last layer decides whether a product is attractive or dangerous. A small, durable item with low returns can survive Amazon’s fee structure. A bulky item, a low-ticket item, or a product with frequent returns can lose money even when sales volume looks healthy.
I usually tell founders to build the model in the same order the cash disappears:
- Fixed cost first: your selling plan sets your baseline overhead.
- Commission next: Amazon takes a category-based percentage from each sale.
- Operations last: fulfillment, storage, prep, and returns determine what is left.
The business question is straightforward. After Amazon gets paid, is there still enough contribution margin to cover your product cost, ad spend, and overhead?
If you already sell through your own site, the logic is familiar. You still need to account for platform costs, payment processing, shipping, and returns. Amazon just compresses those tradeoffs into one channel, which makes it easier to launch and easier to misprice. Founders comparing channels should also weigh how marketplace economics differ from owning the stack on a platform such as WooCommerce vs BigCommerce.
A product that sells well on Amazon is not automatically a good Amazon business. The numbers have to work after fees, not before.
Your first choice, selling plans and monthly fees
Before your first sale, Amazon asks you to choose a selling plan. That decision is small on paper and important in practice.
Individual vs Professional
Amazon offers two standard plans. The Individual plan costs $0.99 per item sold, and the Professional plan costs $39.99 per month. The breakeven point is 40 sales per month.
That is clean math.
If you expect low volume and you are testing a product, the Individual plan keeps your fixed cost low. If you are trying to build a real operating channel, the Professional plan usually makes more sense because the per-sale charge disappears and you get the tools serious sellers need.
What founders usually miss
This choice is not just about cost. It is also about capability.
The Professional plan matters when you want bulk tools, advertising access, and better workflow control. If you are trying to scale beyond a casual test, those features matter more than saving a few dollars in month one.
For founders comparing Amazon to running their own store, this is the same kind of early platform decision you make when reviewing WooCommerce vs BigCommerce. The sticker price matters, but the operational limits matter more.
A simple way to think about it:
- Use Individual if you are validating demand and do not know whether the product should exist on Amazon yet.
- Use Professional if you already have traction, need ad access, or expect regular volume.
- Switch quickly once the numbers point there. Hanging onto the cheaper-looking plan too long is usually false savings.
If you are close to the breakeven line, choose based on workflow, not just arithmetic.
I have seen founders overfocus on the monthly fee and ignore the bigger issue. Amazon rarely gets expensive because of the plan. It gets expensive because the plan is only the first line item.
The main cut, Amazon’s referral fees
If you remember one Amazon fee, remember this one. The referral fee is Amazon’s core commission on each sale.
It is not flat across the marketplace. That is why two products with the same sale price can produce very different payouts.
The referral fee range
Amazon’s referral fees typically range from 8% to 15%, but some categories run much higher. Clothing and Accessories are 17%, some Pet Supplies reach 25%, and Amazon Device Accessories can go up to 45%.
The fee is calculated on the total sales price, including shipping. That is a detail people miss when they do quick napkin math.
Example referral fees by category
| Product Category | Referral Fee Percentage |
|---|---|
| Clothing and Accessories | 17% |
| Footwear | 15% |
| Home and Kitchen | 15% |
| Automotive | 12% |
| Pet Supplies | 25% |
| Electronics accessories | 15% up to $100, then 8% thereafter |
| Watches | 16% up to $1,500, then 3% above |
| Amazon Device Accessories | 45% |
A founder usually makes one of two mistakes here.
The first is assuming the average applies to their category. The second is ignoring tiered rules inside a category and pricing as if everything gets one simple percentage.
Why this changes product strategy
A small fee difference can change the whole business. A founder selling kitchen products and a founder selling pet products may face a very different margin picture even if both products cost the same to make and ship.
Here is the practical takeaway:
- Check category before sourcing. Do not treat category as an admin detail.
- Model the fee on your actual sale price. Referral fees hit revenue, not hope.
- Watch for category edge cases. Amazon has categories where the economics are much harsher than founders expect.
A product does not become good because customers like it. It becomes good when the margin survives the channel.
This is one reason many founders end up with better economics on their own ecommerce site for some product lines and use Amazon selectively for others. If you are weighing that tradeoff, it helps to think through your broader ecommerce technology partner needs before you commit to one channel too heavily.
Fulfillment costs, FBA vs FBM
A lot of founders get the pricing right on paper, then lose the margin on fulfillment.
That usually happens after they choose FBA or FBM too quickly. Amazon can ship for you, or you can ship orders yourself. The right answer depends less on preference and more on your product’s size, velocity, and operational setup.
What FBA costs in real terms
With Fulfillment by Amazon, you send units into Amazon’s network and pay Amazon to store, pick, pack, and ship them. Those fees rise with size, weight, and the space your package takes up.
Packaging matters more than many founders expect. A product can be light and still get expensive if the box is bulky. That changes your fee tier, increases storage pressure, and can turn a product that looked healthy at launch into a weak SKU a few months later.
The practical question is simple. Does Amazon’s fulfillment cost buy you enough extra conversion and operational relief to justify the margin hit?
For many sellers, FBA earns its keep when the item is compact, turns quickly, and benefits from Prime delivery. It also helps teams that do not want to build warehouse processes too early.
When FBM can be the better margin decision
With Fulfillment by Merchant, you store and ship orders yourself or through a 3PL. That usually gives you more control over packaging, carrier choice, and storage costs.
FBM often works better for products that are oversized, fragile, slow-moving, bundled in ways Amazon handles poorly, or expensive to keep in FBA for long periods. It can also make sense if your team already ships efficiently from your own warehouse and can meet customer expectations without paying Amazon’s fulfillment premium.
This choice should be modeled, not guessed. Founders who use predictive retail analytics to forecast demand and inventory flow usually make better fulfillment decisions because they know which SKUs turn fast enough for FBA and which ones sit too long.
A simple rule of thumb helps:
- Choose FBA for smaller, faster-moving products where Prime speed and outsourced operations are likely to increase sales enough to cover the added fees.
- Choose FBM for bulky, awkward, seasonal, or lower-velocity items where storage and fulfillment costs can eat too much margin.
- Review the decision by SKU, not for the whole catalog. One product line can work under FBA while another performs better under FBM.
A bulky product can look profitable at the factory and unprofitable once fulfillment is added.
Putting it all together, two real-world examples
A founder prices a product at $25, sees steady sales, and still wonders why the bank balance is not improving. That usually means the fee model was too shallow.
Example one, a $25 kitchen gadget
Start with a $25 selling price for a standard Home & Kitchen item using FBA.
A 15% referral fee takes $3.75. Add an FBA fulfillment fee of about $4.50, plus roughly $0.99 if you are on the Individual plan, and Amazon’s share is about $9.24 before product cost, inbound freight, packaging, and ads. On the Professional plan, that same sale would carry about $8.25 in Amazon fees before those other costs.
Now look at the business decision. If your landed product cost is $6 and you spend $3 to acquire the order through ads, the margin is already tight. A small price drop, a coupon, or a higher than expected return rate can turn a decent-looking SKU into a weak one fast.
Example two, a $60 yoga mat shipped FBM
Now switch to a different type of product. A bulky yoga mat often changes the math because size affects fulfillment costs more than ticket price does.
At $60, a 15% referral fee is $9. If you fulfill it yourself for $8 in pick, pack, and shipping, Amazon plus fulfillment takes $17 before product cost. If your landed cost is $18, you have $25 left to cover returns, ad spend, labor, and profit.
That can still work. But only if the product has enough room in the price to absorb bulky shipping. If the same mat has to be discounted to $49.99 to stay competitive, the margin shrinks quickly.
What these examples tell you
The point is not to memorize fee tables. The point is to model each SKU before you launch it or restock it.
Use a simple worksheet with five inputs:
- Sale price
- Referral fee by category
- Fulfillment method and cost
- Landed product cost
- Average ad spend or promo cost per order
Then stress-test the result. Run the math at your target price, at a discounted price, and with a higher fulfillment cost than you expect.
Founders do not need a more complicated spreadsheet. They need one that shows whether a SKU still makes money after the costs that actually hit the P&L.
That is how you decide whether Amazon is a good channel for a product, not just whether the product can get sales.
The hidden costs that erode your profit
A founder launches a SKU that looks profitable in the spreadsheet. Referral fee, fulfillment fee, product cost. All covered. Sixty days later, the payout says something else.
The gap usually comes from costs that were treated as side issues instead of part of the model. On Amazon, those side issues decide whether a SKU scales or stalls.
Returns, storage, and ads are not side costs
Returns are one of the fastest ways to wreck contribution margin. A product with a decent gross profit can still underperform if too many orders come back, especially in categories where fit, quality perception, or damage in transit create friction. The sale shows up first. The return-related pain shows up later.
Storage is another quiet drag. Slow inventory ties up cash, raises carrying cost, and can push a product from healthy to mediocre without any change in sale price. This is common with seasonal items, oversized products, and SKUs that looked promising in a launch month but never built steady velocity.
Advertising pressure matters too. Many founders model ads as temporary. In practice, some products need ongoing spend to hold rank, defend branded search, or win placement against entrenched competitors. If your SKU only works before ad spend, it does not really work on Amazon.
A simple way to model the real margin
Use a second margin view after your basic fee math. Start with contribution profit per order, then subtract the costs that tend to appear after launch:
- average ad spend per order
- average return cost per order
- expected storage cost based on sell-through speed
- coupon or promo cost
- prep, labeling, or packaging adjustments
- overhead tied to the channel, such as customer service or reimbursement follow-up
This is the number founders should use when deciding how aggressively to reorder, whether to run a discount, or whether the product belongs on Amazon at all.
Where sellers get caught
The pattern is usually operational, not theoretical.
Some send in too much inventory because unit economics look good on a best-case sales forecast. Some underprice because they benchmark against a competitor without knowing that competitor may have lower product cost, better packaging, or lower ad dependence. Some treat Amazon as the default channel for every SKU, when a product may perform better on a DTC site or another marketplace. If you are weighing channel economics more broadly, this comparison of the best ecommerce platforms for small business is a useful reality check.
Profit erosion on Amazon usually comes from small costs that were never built into the first version of the model.
That is the hidden cost problem. Amazon does not just charge fees. It forces tighter operating discipline. If a product cannot absorb returns, storage drag, and ad pressure, the issue is not the calculator. The issue is the SKU.
How to keep more of your money
You cannot remove Amazon’s fees, but you can make better decisions around them. That is where margin is won.
A good operating habit is to model each product before you list it. Do not guess the payout. Use Amazon’s own calculator, estimate the category fee, and check whether the packaging puts you into a more expensive fulfillment profile.
A short operating checklist
- Tighten packaging early: Smaller and lighter packaging can reduce fulfillment pain, especially when dimensional weight is in play.
- Watch inventory age: Products that move slowly can become expensive even if they eventually sell.
- Choose the channel on purpose: Some products belong on Amazon. Others work better on your own store or in a mixed-channel setup.
- Compare channel economics: If you are still choosing where to focus, this guide to the best ecommerce platforms for small business helps frame Amazon alongside owned-channel options.
- Build the right store foundation: If Amazon is only one part of your sales mix, strong ecommerce development work on your owned channel can protect margin and reduce platform dependence.
- Choose your platform carefully: For brands moving toward owned commerce, the right Shopify development setup can make pricing, fulfillment, and reporting easier to control.
The strongest sellers I know do not treat fees as an annoyance. They treat them as design constraints.
That mindset changes everything. You stop asking whether Amazon is expensive and start asking whether this product, with this packaging, in this category, through this fulfillment model, can support a healthy business.
If you are sorting through channel strategy, pricing, or a rebuild of your ecommerce operation, contact Refact. We have helped 100+ founders build products, our average client relationship lasts 2+ years, and our strategy phase comes with a money-back guarantee. That is how we work: Clarity before code.
