A product with a healthy gross margin can still lose money, and it usually loses it in five places: returns that never get netted back against the sale, storage fees that compound on slow movers, advertising allocated at the account level instead of the product level, landed cost that omits freight and duty, and fee tier changes triggered by packaging. Each one sits outside the gross margin calculation, which is exactly why the product still looks fine on the report you are reading.
Work through a single SKU and the gap becomes obvious.
The example
A kitchen product sells for $34.99. Unit cost from the factory is $9.40. The seller’s spreadsheet shows a gross margin of roughly 73 percent before marketplace fees, and around 40 percent after the referral fee and fulfillment. On four hundred units a month, that reads as a solid contributor.
Now apply the five.
1. Returns cost more than the refund
A returned unit costs you the refunded revenue, the original fulfillment fee in most cases, a return processing charge where applicable, and the unit itself if it comes back unsellable. One return does not cancel one sale. It consumes the margin from several.
At a 9 percent return rate on 400 units, that is 36 returns. If two thirds are resellable and a third are written off, the seller has destroyed roughly 12 units of inventory at $9.40 plus the fulfillment already spent on all 36. The refunded revenue is the visible part. The rest is the part that shows up as an unexplained inventory variance at year end.
Sellers routinely model returns as a revenue reduction. It is a revenue reduction plus a cost event.
2. Storage compounds on exactly the products you should have cut
Storage is charged on volume and duration, which means it is heaviest on the slowest-moving, bulkiest inventory in the catalog. That is precisely the product a gross margin report flatters, because a slow mover’s per-unit margin looks identical to a fast mover’s.
Two SKUs at the same unit economics are not the same business. One turning six times a year and one turning once are separated entirely by carrying cost, and fourth-quarter storage surcharges widen that gap sharply. Check the current fee schedule in your Seller Central help documentation rather than working from last year’s numbers, because these change.
The discipline is to look at margin per unit of storage volume per month, not margin per unit.
3. Advertising is spent by SKU and booked by account
This is the largest of the five for most sellers. Advertising posts to the ledger as one expense. It is spent on individual products, at wildly different efficiencies.
Return our kitchen product to the example. Suppose it carries $2,100 of monthly sponsored spend against 400 units. That is $5.25 per unit, against a post-fee contribution of roughly $14. More than a third of the product’s contribution goes to advertising, and none of that appears anywhere near the product in a standard profit and loss statement. Meanwhile a second SKU with half the gross margin and almost no ad spend may be the more profitable line.
Until ad spend is pushed back to the SKU that generated it, you are ranking products by a number that ignores their largest variable cost. The difference between the margin you are looking at and the margin that pays you is set out at https://www.connectbooks.com/blog-posts/glossary-gross-margin-vs-net-margin, and the distinction is the whole of this problem.
4. Landed cost is not unit cost
The factory invoice is the number most sellers use. Landed cost includes ocean or air freight, customs duty, tariffs, brokerage, drayage, prep, inbound shipping to the fulfillment center, and inspection.
On low-value, high-volume goods, landed cost can run well above the invoice price, and it moves. Freight rates are volatile, duty depends on classification, and a single reclassification can change the economics of a product line. Duty and classification questions belong with a licensed customs broker, and the official tariff schedule is published by the US International Trade Commission.
The practical failure is subtler than omission. Many sellers do include freight, but apply an average across a mixed container, which quietly subsidizes the bulky items at the expense of the small ones. Allocate by volume or weight, not by unit count.
5. The packaging changed and the fee tier moved with it
Fulfillment fees are set by size and weight bands. A product sitting near the top of a band is one packaging revision, one thicker insert, or one humid warehouse away from the next band up, and the step between bands is not proportional to the size change.
This is the reason a product’s economics can deteriorate with no change to price, cost, or volume. It is also the most fixable of the five: measuring and weighing finished units as packed, then checking the band, sometimes recovers real margin by shaving a fraction of an inch.
Re-measure after any supplier change. Factories adjust packaging without telling anyone.
Putting it together
Run the kitchen SKU through all five and the 40 percent post-fee margin lands closer to the mid teens, with returns and advertising doing most of the damage. That is still a profitable product. It is not the product the seller thought they had, and it does not deserve the reorder priority a gross margin ranking would give it.
Two habits fix most of this. Build one report that carries returns, storage, allocated advertising, and full landed cost down to the SKU, and refuse to make reorder decisions from anything else. Then re-run it quarterly, because every input moves.
The products that quietly lose money are rarely the ones with obvious problems. They are the ones with good gross margins and a cost structure nobody has allocated properly. General guidance on tracking business costs is available through the SBA, but the allocation work is specific to your catalog and nobody else will do it for you.
About Post Author
