Almost every SKU has months with better margins, and the useful answer is why. Lay out contribution per unit by month, after referral fees, fulfillment fees, ad spend, refunds, and the FIFO cost layer that sold. Then split the gap between the best month and the current one into those five lines. The month with the best margin percentage is often not the month that made the most money, and the reason a margin slipped is usually two or three small changes landing together rather than one big one. You cannot see any of that on a revenue chart.
A monthly margin figure is only as good as the costs dated into that month. Four of them cause most of the trouble.
COGS from the layer that sold. Under FIFO, units sold in May may come from a purchase order landed in January at one cost while June's units come from a March PO with higher freight. A monthly margin built on a single standard cost hides the step change. The reasons one product carries several costs at once are laid out in why one SKU has four different costs.
Fees at the rate in force that month. Amazon's fee schedule moves within a year. Seller Central's 2026 FBA fee page states that a 3.5 percent fuel and logistics-related surcharge applies to US FBA fulfillment fees starting April 17, 2026, and that holiday peak fulfillment fees apply from October 15, 2026 to January 14, 2027. A margin history that runs across either date shows a step that has nothing to do with your pricing.
Refunds matched to the sale. If refunds land in the month the refund posted rather than the month of the original order, a busy month looks clean and the following quiet month looks terrible. Pick one convention and hold it; the tradeoffs are covered in returns by order date.
Ad spend attributed to the product. A single monthly ad invoice spread evenly across SKUs makes every product's margin history wrong in the same direction, which is worse than useless because it looks consistent.
If any of those four is missing or averaged, the monthly margin line is a smoothed guess. That is where reconciled books matter: ConnectBooks reconciles each Amazon settlement, including its fee, refund, and reimbursement lines, and reports SKU-level profit on FIFO COGS.
Illustrative numbers. Take a seller with an insulated tumbler in Home and Kitchen, where the referral fee is 15 percent per Amazon's pricing page on sell.amazon.com. The fulfillment fee is an illustrative $5.12 before the April surcharge and $5.30 after it, with April blended at $5.21. Returned units in this example are not restocked. On a refunded unit, Amazon credits back the referral fee minus its refund administration fee, which Seller Central's fee page sets at the lesser of $5 or 20 percent of the referral fee. On a $27.99 sale that fee is $0.84, so the seller gets $3.36 of the $4.20 referral back.
The seller asks: were there months when this product had better margins?
Here is March in full, so you can check the method. Gross sales are 700 times $27.99, or $19,593.00. Refunds are 28 times $27.99, or $783.72, leaving net revenue of $18,809.28. Referral fees net of the credit on refunded units: $4.20 times 700, minus $3.36 times 28, equals $2,845.92. Fulfillment is $5.12 times 700, $3,584.00. Ads are $1,610.00. COGS is $4,760.00. Contribution: $18,809.28 minus $2,845.92 minus $3,584.00 minus $1,610.00 minus $4,760.00 equals $6,009.36, which is 31.9 percent of net revenue.
February has the best margin percentage. April has the most contribution dollars, even at a $3 discount, because it sold 50 percent more units than March. Which one counts as "better" depends on the decision. If you are setting a price floor, February tells you what the product earns at full price with modest ads. If you are deciding whether to run the April promotion again, April's $6,553.02 against March's $6,009.36 says the discount paid for itself in dollars, though it cost about $2.34 of contribution per unit.
Same price, similar volume, and $2.41 less contribution per unit. Break it into lines:
Those four add to $2.42, which matches the $2.41 gap after rounding. Nearly half the decline is ad spend. A quarter is the new cost layer. The fee change, which sellers tend to blame first, is the smallest piece.
That ranking changes what you do next. You cannot negotiate the surcharge. You can look at whether the extra $1.10 a unit of ad spend bought incremental sales or bought the same buyers at a higher price, and whether the June refund spike traces to a batch.
Compare unit contribution, not just percentage. A price cut lowers margin percentage even when it raises total profit. Read both lines side by side.
Mark the dates where costs stepped. New PO landed, fee schedule changed, ad strategy changed, price changed. Most margin histories are a series of flat stretches broken at those dates, and a chart with them labeled answers half the question.
Watch the denominator in small months. A SKU selling 40 units can swing 8 points of margin on two refunds. Below a few hundred units a month, look at a rolling quarter.
Check the period boundaries. Amazon settles roughly every two weeks, per its sell.amazon.com guide to seller payments, so a calendar month is assembled from parts of settlements. If the books post settlements as single lump entries dated to the settlement close, a month can carry another month's fees.
Building the six-month breakdown by hand means pulling settlement lines, PO costs, and ad reports for each month and lining them up per SKU. Crunch, the AI CFO inside ConnectBooks, takes the question directly ("did this tumbler have better margins in other months, and why") against the seller's reconciled ConnectBooks data. It finds what changed, compares the periods, isolates the products responsible, explains the drivers, and recommends a next step, such as reviewing June's ad spend or the refund reasons on that batch.
Check any figure it cites against the profit reports for the same month before you act on it.
It cannot correct a missing cost layer. If the March PO was never entered with its freight and duty, every month after it shows a margin that is too high, and the AI will report the inflated figure with confidence. Check that each PO carries landed cost before trusting a margin history.
It does not know why the refund rate moved. It can show that 43 June refunds clustered in one week. It cannot know that the factory changed lid suppliers. You make that call, or the phone call.
It does not decide what "better" means. February and April are both answers. Which one matters depends on whether you are protecting a price, sizing a promotion, or planning a reorder, and that judgment stays with you.
When the best month and the current month differ by a few dollars a unit, the split usually points to one controllable line and one or two you live with. Fix the controllable one first. If it is ad spend, test a lower bid on that SKU for two weeks and watch total units, not attributed units. If it is the cost layer, the next PO negotiation has a number attached to it. If it is refunds, read the reasons before you touch price. Then rerun the same monthly view in thirty days and see whether the line you changed moved.
Running an e-commerce business comes with plenty of challenges, but ConnectBooks is here to make your life easier. With real-time insights, seamless integrations, and detailed tracking of your profitability and inventory, you can stay ahead of the game. Whether you’re selling on Amazon, Shopify, Walmart, TikTok or eBay, ConnectBooks helps you manage your finances with 100% accuracy and confidence, so you can focus on growing your business.
Ready to level up? Start making smarter, data-driven decisions every step of the way. Try ConnectBooks Free Today or Schedule a Demo