"Why am I down in profit this month" has a five-step answer: what changed, over which periods, on which products, why, and what to do next. Skip a step and you get a fact without a cause, or a cause without a fix. Revenue flat and profit down $6,695 is a fact. Ad spend up $3,700, two SKUs moved into a higher fulfillment fee band, and returns spiked on one product is a cause. Cut two campaigns, repackage two products, and inspect one batch is a fix.
The chain works by hand in a spreadsheet, and it works faster when an AI runs it on reconciled data. The arithmetic below is built so you can reproduce it either way.
Illustrative numbers. Take an Amazon seller with 40 SKUs, comparing August to July. Referral fees run at 15 percent, Amazon's published rate for Home and Kitchen on sell.amazon.com. Fulfillment fee figures are illustrative; Amazon sets them by product size tier and weight, and the 2026 schedule is covered in Amazon FBA fees 2026.
July
August
Compute contribution for each month. Revenue minus every variable cost.
July: $142,000 minus $52,500 minus $21,300 minus $24,800 minus $11,200 minus $4,100 minus $1,900 equals $26,200.
August: $142,300 minus $52,600 minus $21,345 minus $26,400 minus $14,900 minus $5,600 minus $1,950 equals $19,505.
Profit fell $6,695, or 25.6 percent, on revenue that rose $300. That is the fact. Anyone looking at the Sales dashboard in Seller Central would see a flat month.
Month over month is the obvious comparison and the right one here, because the question was about "this month." Two checks before trusting it. First, is August seasonally weaker than July for this catalog? Look at August versus July last year. If the same drop showed up then, part of the answer is the calendar. Second, are both months on the same accounting basis? Settlement deposits do not line up with calendar months. The figures above are by order date, from settlements reconciled at the transaction level, so the comparison holds. A ledger that books each two-week settlement as a lump on its deposit date can put three settlements in one month and two in the next, and the "drop" is an artifact.
Take each line's change and sign it by its effect on profit.
Sum: $300 minus $100 minus $45 minus $1,600 minus $3,700 minus $1,500 minus $50 equals negative $6,695. The decomposition ties to the total, which is the check that you have not dropped a line.
Three lines carry the decline. Ad spend is $3,700 of $6,695, or 55.3 percent. Fulfillment fees are 23.9 percent. Refunds are 22.4 percent. The other four lines net to a small positive.
Cost lines are not causes. The cause lives at the SKU level, and this is where a blended ledger stops and reconciled data keeps going.
Fulfillment fees, up $1,600. Two SKUs changed fee bands. SKU-14, a bamboo cutting board, gained a retail box in August that pushed it over a weight threshold; 900 units at $0.95 more each is $855. SKU-22, a ceramic pour-over set, was repacked with a heavier insert; 700 units at $1.05 more each is $735. Together, $1,590 of the $1,600. The remaining $10 is mix.
Ad spend, up $3,700. SKU-03 accounts for $2,100 of it. A new Sponsored Products campaign added $3,900 of ad-attributed sales. At SKU-03's 33 percent margin before ads, those sales produced $1,287 of gross profit against $2,100 of spend. Net effect negative $813. SKU-11 added $1,100 of spend for $1,400 of attributed sales at a 28 percent margin, or $392 of gross profit, net negative $708. The remaining $500 is spread across a dozen campaigns with small increases.
Refunds, up $1,500. SKU-07, a glass storage jar set, went from a 4 percent to a 9 percent return rate; $1,300 of the increase is that one product. Return reasons in Seller Central cluster on "arrived damaged." The other $200 is noise.
Now the "why" reads as a sentence. Profit fell because two products were repackaged into a higher fulfillment fee band, one new campaign and one expanded campaign bought sales at a margin lower than their cost, and one product started breaking in transit.
Each cause has a distinct action.
For SKU-14 and SKU-22, run the fee arithmetic on the old packaging versus the new. If the retail box adds $0.95 per unit and no conversion lift, revert it. If it lifts conversion, price the lift against $855 a month.
For SKU-03 and SKU-11, the campaigns lose money at current margin. Either cut them, or raise the break-even by fixing something upstream (price, COGS, fee band). Amazon's own advertising guide makes the same point: to hold profit, ACoS has to sit below margin.
For SKU-07, pull a sample from the FBA inventory and inspect the packaging. A 9 percent return rate on glassware with "damaged" as the reason is a packaging problem until proven otherwise. The fix costs a few cents of foam and recovers $1,300 a month.
An AI on reconciled data runs the five steps in the order above, in seconds, across all 40 SKUs instead of the three you had time to check. Crunch, the analytics AI inside ConnectBooks, is built on exactly this chain: find what changed across sales, pricing, ads, fees, returns, storage, inventory, and margins; compare the right periods; find the products responsible; explain why; and recommend what to do next. The illustrative example on the Crunch page follows the same shape, with the decline traced to six SKUs, ad spend without enough incremental gross profit, and two products moving into a higher fulfillment fee band.
The prerequisite is the data. ConnectBooks reconciles each Amazon settlement into QuickBooks or Xero at the transaction level, so fulfillment fees, referral fees, refunds, and storage each post to their own account against the SKU that caused them, with FIFO COGS and landed cost per unit. The SKU-level profit reports are what Crunch reads. Without that layer, step four is impossible for any tool, because the SKU detail was never recorded.
It cannot tell you that SKU-07 is breaking because the foam insert was dropped from the spec. It sees a return rate and a reason code. The inspection is yours.
It cannot judge whether the retail box on SKU-14 is worth keeping for brand reasons. It can price the fee delta. The trade-off is a judgment call.
It cannot see a landed cost that was never entered. If SKU-22's last PO has no freight allocated, the COGS line is understated and the AI will say margin is better than it is. Ask any tool what it does with a missing cost; the right answer is a flag.
It takes no action. Crunch does not pause SKU-03's campaign or revert SKU-14's packaging. It tells you what it found and what it would do. The Seller Central login stays with you.
Run the chain by hand once on your own numbers. Then decide whether you want to run it by hand every month.
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
