Real profitability on Amazon: which products actually earn

By the Finnex Agency team

The essentials, before we start

Your Amazon dashboard isn't lying to you. It's simply showing you part of the story.

Amazon knows how much you sold and can calculate many of the fees it charges on each transaction, but there's one thing it can't know: what it actually cost you to sell that product. It doesn't know what you paid the manufacturer, what you spent getting the product to its fulfillment center, or what you invested in content, software, outside services, or management.

That's why a product can show solid sales and still be losing money.

In September 2025 Amazon introduced Profit Analytics, a tool that lets you add extra costs and get a fuller picture of profitability. But there's an important difference: those costs that sit outside Amazon have to be supplied and kept up to date by the seller.

And that's the real challenge. Knowing how much you sold is easy. Knowing how much you actually make on each product is another story.

In this article we'll look at how to run that calculation, which costs tend to show up late, and how to use that information to decide which products to scale, which to fix, and which to let go.

Net proceeds aren't profit, and the gap is your whole business

There are three numbers worth separating from the start.

Net proceeds include certain referral fees, fulfillment fees, advertising, refunds, and promotions. Profit also includes product cost, inbound logistics, duties, seller-funded promotions, content, tools, and other operating costs.

The difference sounds small when you put it in one sentence. In practice, it can be your entire margin.

Plenty of accounts don't necessarily have a sales problem: they have a visibility problem with their own numbers. If you don't know what you actually make on each product, it's very hard to know what you should be scaling. It's the same logic we use to approach account economics inside a professional audit, where per-product margin is usually the first thing we straighten out.

What Amazon calculates on its own, and what it can't

This distinction matters because a lot of decisions get made by looking only at what shows up inside Seller Central.

Amazon can calculate automatically:

• The referral fees and charges for its services.
• Fulfillment by Amazon fees.
• Refunds and certain adjustments.
• Amazon Ads spend.
• Some charges tied to promotions and returns.
• The revenue and costs available in its economics and profitability reports.

But Amazon doesn't automatically know:

• Your product's manufacturing cost.
• Freight from the supplier to Amazon.
• Duties and import taxes.
• Prep, labeling, or inspection done by third parties.
• Photography, video, and content production.
• Tool and software subscriptions.
• Fees for agencies, freelancers, or consultants.
• The real cost of a returned unit that can no longer be resold.

And here's one of the quietest problems. Amazon can record that a unit was returned and whether it became sellable again, but it doesn't know what that unit cost you or how much value you lost if it ended up damaged or never made it back into sellable inventory.

Where Profit Analytics fits in

Amazon currently offers tools that let you analyze your products' economics and add certain costs that originate outside the platform. The upside is that you can get much closer to a real view of profitability. The downside is that the quality of the output depends entirely on the quality of the data you load.

If your product cost is out of date, if you leave out certain logistics costs, or if you never refresh your expenses, the report can be technically correct and still fail to represent your reality.

For pre-launch estimates, Amazon also offers the FBA Revenue Calculator, which lets you plug in product cost and estimate a unit's economics.

The fees worth knowing by heart

Amazon's fees change often and also depend on the marketplace, the category, and the product's size, weight, and characteristics. So rather than memorizing every number, it's better to understand which charges can hit your margin and where to verify them.

Referral fee

The referral fee is one of the main variable costs of selling on Amazon.

In Spain, for example, the vitamins, minerals, and supplements category currently has its own structure: for products up to €10, the fee can be lower than in categories like beauty, health, and personal care. That matters most on low-priced products, where a few percentage points can make a meaningful difference to the margin.

So before you calculate a product's profitability, verify:

• The category it's classified under.
• The selling price.
• The applicable referral fee.
• The per-unit minimum, where it applies.

Common mistake: assuming two products at similar prices necessarily have the same fee structure.

Subscription

Amazon charges a monthly fee for the professional selling plan. The amount depends on the marketplace and the account structure, so it's always worth checking each country's current fee schedule. Even though it's a relatively small cost next to other expenses, it's part of the fixed cost of operating on Amazon.

Fulfillment by Amazon

Fulfillment is another major component of per-unit economics. The cost depends, among other variables, on weight, dimensions, product type, marketplace, season, and which fulfillment services you use.

On a supplement, for example, a small difference in weight or size can change the per-unit cost — and as volume grows, that difference can turn into thousands of euros a year.

Amazon can also apply specific surcharges on certain fees, such as adjustments tied to fuel and logistics. So to work out a product's real margin, an average fee isn't enough: you need the fee that actually applies to that product.

Peak season

During certain periods of the year, Amazon applies storage and fulfillment fees that differ from off-peak rates, and this matters most in the fourth quarter. A product that looks profitable in the first months of the year can end up with a significantly thinner margin if it goes into peak season carrying too much inventory.

The five costs that show up late

Some costs are easy to spot because they land immediately. Others arrive later, and those are the dangerous ones — because you may have made pricing, advertising, or replenishment decisions based on a margin that didn't yet include every cost.

1. Storage

Fulfillment storage isn't static across the year: rates can rise during peak season and also vary by product type. That has an important consequence — inventory you bought months ago can still be generating costs today.

If you send too much stock for the fourth quarter and turnover doesn't keep up, you don't just have capital tied up. You also keep paying to occupy that space.

2. Aged inventory surcharges

Amazon applies additional charges when certain units sit in storage too long. The problem is that this cost doesn't appear when you decided to buy that inventory: it appears afterward. A poor demand forecast can end up hitting profitability several months later.

The product didn't necessarily become less profitable. The inventory you bought too long ago started costing you more.

3. Storage utilization surcharges

Amazon can also apply additional charges when your inventory level exceeds certain thresholds relative to historical sales. The logic is simple: if you're holding a lot of inventory with little turnover, Amazon is giving up storage capacity it could use for products that sell faster.

For the seller, that's one more reason to look at inventory not just as available stock, but as capital that carries a cost.

4. Returns processing fee

Returns can hit your margin in a way that isn't always immediate. Amazon can apply return-related fees when a product exceeds certain thresholds set for its category and analysis period.

That means the cost showing up today may be tied to return behavior from weeks or months earlier. Looking only at the current month's sales and refunds can give you an incomplete picture.

5. Low-inventory-level fee

The opposite extreme can be expensive too. When a product holds inventory levels that are too low relative to its historical sales, Amazon can apply a low-inventory-level fee under certain circumstances.

That adds another variable to planning: it isn't simply about having little or a lot of stock, but about having the right stock for the level of demand you expect. And it matters most on products that are growing fast.

The discounts you fund without noticing

Not every cost arrives as an invoice. Some just show up as less money per sale.

Subscribe & Save

Subscribe & Save can be an excellent tool for increasing repeat purchases and customer lifetime value, but any seller-funded discount reduces the effective revenue from that sale.

So when you calculate a product's profitability, you shouldn't work from list price alone. You have to look at what the customer actually ends up paying and how much of that discount you're absorbing.

The exact treatment of certain discounts can vary by program and applicable terms, so it's worth verifying in your account's transaction detail.

Coupons

Coupons are a powerful lever for click-through rate and conversion, but they have a cost, and that cost belongs in the profitability calculation.

In the United States, for example, Amazon changed the coupon fee structure in 2025. The important lesson isn't to memorize a figure: if your profitability model still runs on coupon fees from one or two years ago, you're probably overestimating your margin.

Vine

Amazon Vine also carries costs based on how many units you enroll. But there's something more important than the program fee: those units still have a product cost and can generate fulfillment costs.

So if you use Vine to launch a product, those units belong in the launch budget. They aren't lost sales: they're an investment in initial review generation and product validation.

In Spain the calculation has one more step

When you analyze profitability in Spain, you also have to consider the tax treatment of your transactions and of Amazon's fees.

Amazon's published fees may appear excluding VAT, and the treatment that follows depends on factors such as:

• Where your company is established.
• Where it's registered for VAT.
• Which Amazon entity is providing the service.
• What type of transaction you're carrying out.
• The tax treatment that applies to your product.

So VAT shouldn't be dropped into a profitability model automatically at a single rate for every seller. If your company can recover input VAT, for instance, that amount may mainly represent a cash-flow impact rather than a permanent cost to the business.

Our recommendation is to always separate margin from tax, and to validate the specific treatment with a tax advisor. This matters especially for supplements and food-related products, where the applicable VAT rate can depend on the product's classification and characteristics.

The three-line method

Now for the most important part, because knowing all these costs doesn't help much if you don't know what to do with them.

At Finnex we recommend thinking about profitability per unit and per product, not just at the account level. You can start with three simple lines.

Line 1. What the sale actually leaves you. Start from the revenue a single unit generates and subtract the charges Amazon records for that sale. That includes applicable fees, fulfillment, refunds, and any other relevant variable costs.

Line 2. What it costs you to have that unit. This is product cost, freight to Amazon, duties, prep, labeling, and other variable costs tied directly to the product. If you bought a lot of 10,000 units, spread those costs across the corresponding units.

Line 3. What it costs you to sell that unit. This is mainly Amazon Ads, seller-funded discounts, promotions, and other variable costs tied to acquiring the sale.

The result is much closer to a per-unit contribution margin:

Per-unit contribution margin = revenue per unit − variable costs per unit

After that, if you want to get to the business's operating profit, you still have to account for fixed costs: software, team, agency, admin, and overhead.

Two rules that make the difference

Spread costs that don't belong to a single sale. Photos, content, tools, and certain services don't necessarily belong to one month or one unit. Distribute those costs consistently so one month doesn't look terrible and the next artificially good.

And analyze advertising per product. Account-level TACoS can be useful for judging the overall health of the business, but it doesn't tell you how much each product is contributing. A profitable product can be subsidizing one that loses money, and the account average can hide it.

How to decide what to do with each product

Once you have real per-unit margin and cross it against sales volume, the catalog starts to sort itself. You can think of your products in four groups.

Earns and sells

This is your best case: good economics and real volume. Here it makes sense to look at more stock, more advertising investment, keyword expansion, new formats, and expansion into other marketplaces. That's the territory of scaling on Amazon Ads with real growth.

Secure the economics first. Then scale.

Earns but doesn't sell

Here the problem probably isn't the product's profitability but demand, visibility, or conversion. It could be keywords, advertising, the listing, price, creative, reviews, or ranking.

It's one of the most interesting groups to work on, because the product already has economics that work. If it's also been flat for months, the diagnosis is how to revive a stalled listing.

Sells but doesn't earn

This is the dangerous product, because from the outside it looks successful: it has sales, movement, and possibly strong revenue. But if each sale leaves little margin or outright loses money, more volume doesn't necessarily improve the business. It can make it worse.

Scaling a product that loses money is just accelerating the loss.

Before you raise the advertising budget, you need to understand what's destroying the margin. Price? Product cost? Fulfillment fees? Advertising? Discounts? Returns? The answer determines the strategy.

Neither earns nor sells

Here you probably have the clearest candidate to leave the catalog. That doesn't necessarily mean liquidating it tomorrow: it means stopping treating it as a product with growth potential when the numbers don't back that up.

You can stop replenishing, liquidate the inventory, keep it from becoming aged stock, recover capital, free up space, and concentrate resources on products with better prospects.

The order matters

A common mistake is to start by scaling the products that already sell. We prefer a different approach: first eliminate or contain the losses, then improve the economics of the products that sell, and only then scale the winners.

Because scaling before you clean up the catalog can mean spending more money only to discover later that the problem was in the product's economics.

What you shouldn't mistake for an unprofitable product

Not every negative margin means the product is bad. Some situations distort the numbers temporarily.

A launch in its first months. A new product usually carries high launch costs: advertising, Vine, promotions, and initial review generation. A negative margin early on can be normal. What matters is the trajectory. Is conversion improving? Is cost per click coming down? Is ranking improving? Is organic conversion rising? Does margin improve as volume grows? The important figure isn't always today's margin. It's also the direction it's moving in.

A month with extraordinary inventory costs. Storage or aged-inventory costs can reflect decisions made months earlier. A single month may not represent the product's structural economics.

A product with delayed return costs. Returns and their associated costs can land in periods after the original sale. If you look only at the month's result, you can charge September with a cost that was actually driven by earlier months' sales.

A product that works as an entry point. Some products aren't meant to maximize the margin on the first purchase. They can generate customers who later buy other products from the brand, subscribe, or carry high lifetime value. In those cases individual margin still matters, but it doesn't necessarily tell the whole story. That's where customer lifetime value comes in.

Wrapping up

Amazon gives you precise numbers, but precision doesn't necessarily mean profitability. You can know exactly how much you sold, how much you paid in fees, and how much you invested in advertising, and still not know how much money you're actually making per product.

The reason is simple: Amazon doesn't know all your costs. Product cost, inbound logistics, duties, certain outside services, and other expenses have to be built into your own model.

And when you do that, a completely different picture of the business appears. A product that looked like a winner may be funding another one. One that looked unimportant may have excellent economics. And one that generates a lot of revenue may be consuming capital without producing healthy returns.

So before you raise advertising, buy more inventory, or launch another product, there's one question you should be able to answer: how much money does each unit I sell actually leave me?

If you can't answer that with confidence, that's the first problem to solve. Before optimizing the listing, before raising the budget, and before scaling.

Understand the economics first. Then decide what to do with them.

This article is informational and does not constitute tax or accounting advice. Amazon's fees change frequently and vary by marketplace and category, and VAT treatment depends on each company's specific situation. Always check the current fee schedule and validate tax treatment with qualified professionals.

---

If you don't know what each product in your catalog actually leaves you, you can request your free audit. We review per-unit economics, identify which products are funding others, and define what's worth scaling, what to fix, and what to let go.