Which SKUs Can Afford a Black Friday Discount?

A SKU can afford a Black Friday discount when its contribution after every holiday-rate fee, ad cost, and return allowance stays positive at the discounted price, or when it is aged stock whose already-paid cost should not drive the decision. Your own reconciled data answers the first test SKU by SKU in a few minutes, provided the data carries landed COGS and per-SKU fees. The second test needs a cash view, not an accounting one. A store-wide percentage fails both tests for some products, and the revenue report will not tell you which.

The five questions to ask before you mark anything down

These are the questions, in plain language, that decide the markdown list. Each one needs SKU-level revenue, fees, ad spend, refunds, and FIFO COGS for at least the last two quarters, plus last year's holiday weeks if you have them.

Which SKUs earned the most contribution per unit after fees and ads over the last 90 days? This ranks the products by how much room they have before a discount hits zero.
For each SKU that ran a discount last Black Friday Week, did contribution after ads beat the two weeks before? This tells you whether your past markdowns paid or just moved units.
Which products had stronger margins in earlier months than they have now, and why? A SKU whose fulfillment fee band changed or whose landed cost rose since summer has less room than its old numbers suggest.
Which SKUs have return rates high enough to eat the discount? Holiday returns arrive late. Amazon's 2026 extended holiday returns policy, announced on Seller Central in October 2026, lets items bought between November 1 and December 31, 2026 be returned until January 31, 2027.
Which aged SKUs should be held, discounted, liquidated, or removed before storage fees rise further? Here the paid COGS is sunk and the question is future cash, not margin.

Questions written this way get better answers from people and from software. The general method is in how to write questions for an AI financial assistant.

Worked example: four SKUs, one proposed 25 percent markdown

Illustrative numbers. Take a seller planning 25 percent off across Amazon for Black Friday Week, with deals carrying the 1.5 percent variable fee from Amazon's Holiday 2026 Seller Central announcement and the 15 percent Home and Kitchen referral fee from sell.amazon.com/pricing. Percentage costs total 16.5 percent, so each dollar of price keeps $0.835. "Dollar costs" below means landed COGS, the holiday-peak fulfillment fee, expected ad cost per unit, return allowance, and the upfront deal fee spread over expected units.

SKU A, $29.99, dollar costs $14.20. Breakeven floor: $14.20 divided by 0.835 equals $17.01. Maximum discount 43.3 percent. At 25 percent off ($22.49), contribution per unit is $22.49 times 0.835, or $18.78, minus $14.20: $4.58. Safe to discount.

SKU B, $44.99, dollar costs $31.10. A large-standard product with heavy ad spend. Floor $37.25, maximum discount 17.2 percent. At 25 percent off ($33.74), contribution is $28.17 minus $31.10: negative $2.93 per unit. Every deal unit loses money.

SKU C, $19.99, dollar costs $15.80. A low-priced item whose fulfillment fee is a large share of price and whose return rate runs high. Floor $18.92, maximum discount 5.3 percent. At 25 percent off ($14.99), contribution is $12.52 minus $15.80: negative $3.28. The worst candidate in the catalog, and probably one of the most popular on a revenue chart.

SKU D, $34.99, dollar costs $22.00, 215 days in the fulfillment network. Floor $26.35, maximum discount 24.7 percent. At 25 percent off ($26.24), accounting contribution is $21.91 minus $22.00: negative $0.09. On the P&L it is a marginal loss.

But $9.50 of SKU D's dollar costs is COGS paid in the spring. On a cash contribution basis, the costs still to be paid are $12.50, and each unit at $26.24 brings in $21.91 minus $12.50, or $9.41 of future cash. The cash floor is $12.50 divided by 0.835, $14.97, a 57.2 percent discount. Holding it at full price means more months of storage and aged-inventory surcharges on units that are not moving. SKU D should be discounted, possibly harder than 25 percent.

The store-wide plan discounted two SKUs that could afford it and two that could not. The right list: A and D on deal, B at a shallower 15 percent or not at all, C at full price.

The floor method for any SKU is set out step by step in setting a Black Friday discount floor for every SKU.

Asking the data instead of building the spreadsheet

The arithmetic above is simple for four SKUs. It is a lost weekend for four hundred, because each row needs the current FIFO cost layer, the fee charged on recent settlements, and ad spend assigned to the product.

Crunch, the AI CFO inside ConnectBooks, takes questions like the five above and answers them from the seller's reconciled ConnectBooks data. Its question areas include margins by SKU, pricing investigations, inventory and storage fees, and cash contribution. On every answer it works the same chain: find what changed, compare the right periods, find the products responsible, explain why, and recommend what to do next. For aged inventory, when asked, it treats paid COGS as sunk and weighs future storage and aged-inventory fees, selling fees, expected sell-through, and expected cash recovery, then ranks SKUs as hold, discount, liquidate, or remove.

A seller could ask it whether a product like SKU B or C had better margins in other months, and see which fee and ad lines moved since then. That is a better starting point than a revenue ranking. The underlying per-SKU figures are the same ones in ConnectBooks profit reporting.

What the AI cannot tell you

It cannot tell you how many units a discount will sell. Crunch analyzes what has happened; it does not forecast demand at a new price. The velocity at 25 percent off is your estimate, informed by last year's deals if you ran any. That estimate decides whether a discount above the floor makes more money than full price.

It does not know what your competitors will charge on Black Friday, whether Amazon will feature your deal, or whether your supplier can ship a reorder in January if the product sells out.

It cannot fix missing data. If SKU C has no landed cost entered, any margin figure for it is wrong by the size of that gap. A wrong number delivered with confidence is worse than no number, so check that every active SKU has a cost before you trust a ranking.

It takes no action. It will not change a price, create a deal, or edit a coupon. It recommends. The decision, and the January review of whether it was right, are yours.

Before Thanksgiving

Get the markdown list from the data, not from the revenue chart. Discount the SKUs with room, push the aged ones on a cash basis, and protect the ones whose fees already take most of the price. Then write down the units you expect from each deal, so the post-event review has something to compare against.

Take Control of Your E-Commerce Business with ConnectBooks

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