Week-over-week analysis on one Amazon product answers a narrow question: of the last several weeks, which one made the most contribution, and what was different about it. Done properly, the answer is often not the week with the most units. A deal week sells three times the volume at a discount, with bid-up ad spend, and the returns from it land two weeks later in a week that did nothing wrong. The most profitable week is frequently the quiet one after, when the rank lift from the deal is still paying and the price is back to full. Five alignment problems stand between a seller and that finding, and each one is a way to pick the wrong week.
Weekday alignment. Amazon demand has a weekly shape. Compare Monday to Sunday against Monday to Sunday, never a seven-day window that starts on a Thursday against one that starts on a Saturday. If your reporting tool defaults to calendar weeks starting Sunday, fine, as long as every week uses the same rule.
Deal and holiday weeks. A Prime Day, a Lightning Deal, or a holiday shifts both price and volume for a few days and pulls demand forward from the following week. A deal week is not comparable to a normal week and should be labeled as such before any week-over-week figure is read. The comparison that matters is deal week against the last deal week, and post-deal week against the last post-deal week.
Ad spend timing. Ad spend is recorded by the day it is incurred in the ad console, and billed on invoices that do not line up with weeks. Use daily spend by campaign, attributed to the SKU, on the day it ran. A week that carries an invoice for the prior week's spend will look worse than it was.
Settlement lag versus order date. Amazon settles every two weeks and deposits after that. If you book revenue when the deposit lands, a week's sales show up in the ledger one to three weeks later, and the week-over-week series is a series of deposit timings, not of sales. Use order date. The settlement report guide covers how the two-week statement maps back to orders.
Returns lag. Amazon's return window runs weeks, so refunds from a high-volume week land in the following two or three weeks. On an order-date basis, the refund is posted when it happens, which charges the deal week's returns to the weeks after it. For a fair comparison, match each refund to its original order and attribute it back.
Illustrative numbers. Take a seller with an insulated lunch bag at $32.99 in Home and Kitchen, where sell.amazon.com lists the referral fee at 15 percent. Fulfillment fee is $4.60 per unit, Amazon's 2026 non-peak large standard rate for 12 to 16 ounces at $10 to $50, fuel surcharge aside. FIFO landed cost is $9.20 per unit. Weeks run Monday to Sunday. Refunds are shown net of the referral fee Amazon credits back and the refund administration fee it keeps.
Week 1, September 7 to 13. Normal week.
Week 2, September 14 to 20. Seven-day deal at 20 percent off, $26.39.
Week 3, September 21 to 27. Post-deal hangover.
Week 4, September 28 to October 4. Rank lift, full price.
On raw contribution, week 4 wins at $2,570.37, week 2 is second at $2,158.38, then week 1, then week 3. Week 2 sold more than twice the units of week 4 and earned less.
Now fix the returns lag. Of week 3's $1,010 in net refunds, $830 traces to deal-week orders. Move it back. Week 2 falls to $1,328.38 and week 3 rises to $1,488.64. The deal week drops from second to last, behind the ordinary week 1, and the week that looked like a disaster turns out to have been a normal week absorbing someone else's returns.
Why week 4 won, in three parts. Price: every unit at $32.99 rather than $26.39, worth $6.60 of revenue per unit, of which $5.61 survives the referral fee. Ads: $2.76 of spend per unit against $4.13 in the deal week, because the deal bought rank that the following weeks used for free. Returns: normal, because the units were bought by shoppers who wanted a lunch bag rather than shoppers who wanted a deal.
The action is not "never run deals." The deal produced the rank that made week 4 possible. The action is to evaluate the deal on its full cost, including the returns it generated and the week 4 lift it earned, rather than on the revenue tile for the week it ran. That is a three-week window, and a seller who reads only week 2 will either overrate the deal or, after week 3's refunds, underrate it.
The five alignment rules above are the reason an AI on reconciled data gets this right where a spreadsheet built from Business Reports does not. Order-date revenue, per-order fees, daily ad spend by SKU, and refunds matched to original orders all have to exist before the comparison is meaningful. ConnectBooks reconciles each settlement into those pieces, and the SKU-level profit reports hold the result by product and by day. The comparison report is the period-over-period view built on it.
Crunch, the analytics AI inside ConnectBooks, runs the analysis on request: ask it to analyze this lunch bag week over week for September and find the most profitable week and why, and it compares the periods, finds what changed across sales, pricing, ads, fees, and returns, and explains the difference in the terms above. It will name week 4, and it will say the deal week's returns landed in week 3. It will also flag that the deal week is a deal week, because the price series shows it.
It cannot know a deal is coming. It sees a price drop on September 14 and a price recovery on September 21. Whether that was a Lightning Deal, a coupon, or a repricer misfire is in your Seller Central, and you tell it.
It cannot attribute a return to a cause. Thirty-eight units came back in week 3. Whether the deal attracted the wrong buyers, or a batch had a zipper defect, is in the return reason codes and the units themselves.
It cannot decide whether rank lift is worth the deal's cost. It can price the lift at the week 4 figures. Whether to run the deal again is a call about your category and competition.
And it will not schedule the next deal. It tells you what the last one did.
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
