Judge Black Friday on contribution, not revenue. Know how many extra orders each discount depth needs to stand still, raise your break-even MER, the revenue each ad euro must bring, for the sale weeks, scope codes away from orders you would get anyway, order stock against lead times, and plan for December's refunds and the VAT that follows.
Black Friday is judged on revenue and paid for out of contribution, and that mismatch is why so many record weekends end in a quiet January. A discount comes off the price, while the goods, the parcel and most of the payment fee cost exactly what they did the week before, so the whole discount comes out of the slice each order was actually keeping. In 2026 Black Friday falls on Friday 27 November and Cyber Monday on 30 November, which leaves about eight weeks to set the sale up on numbers instead of habit. This checklist is the ten checks worth running before anything goes live, in the order they matter, with the arithmetic done on an example store's own margins so every figure can be redone with yours.
In short
- A discount is a small share of the price and a large share of what an order keeps: at a 48,9% contribution margin, a 20% code needs about 69,3% more orders just to stand still.
- The ad floor rises during a sale: on the same margins, break-even MER moves from 2,05 at full price to 2,77 at 20% off.
- Scope beats depth, because every order that would have come anyway takes the discount and gives nothing back.
- The sale's refunds land in December and its VAT leaves in December or January, so the cash plan belongs before the sale, not after it.
01 Write down what an order keeps before any discount
Everything below rests on one number: the share of net revenue an order keeps after its goods, its parcel and its payment fee, which is the margin after goods, parcel and payment fee, CM2 on a contribution ladder. Revenue is the wrong starting point, because a discount does not take a share of revenue away from a cost; it takes it away from what is left after the costs. On an example store this site's studies run, priced by the same engine as the product, that share came to 48,9% over fourteen months: of every €100 of net revenue, €48,88 survived the goods, the parcels and the fees.
Measure yours on a normal month rather than last November, because a promotion month already has the discount inside it. If you only know your product margin, take the parcel and the payment fee off as well before going on. A sale that looks comfortable at a 60% product margin can be a loss at the contribution that actually has to pay for the advertising, and the advertising is the part of the sale that grows fastest.
02 Price every discount depth in orders
A code takes its percentage off the price, so on an order that kept €48,88 per €100 of net revenue, a 10% code leaves €38,88. To make the same total contribution, the sale has to sell enough extra orders to replace what each order gave away, and the requirement grows much faster than the discount does, because the discount comes out of the thin slice rather than out of the whole price.
| Discount | Contribution left | Extra orders needed to stand still |
|---|---|---|
| None | €48,88 | none |
| 10% | €38,88 | about 25,7% |
| 15% | €33,88 | about 44,3% |
| 20% | €28,88 | about 69,3% |
| 30% | €18,88 | about 158,9% |
Read the table as a question to put to last year's sale: did the 20% weekend really sell about seventeen orders for every ten it would have sold anyway? Only the percentage part of the payment fee shrinks with the price, so the real requirement is a little lower than the table says, and the shape does not change. The discount impact calculator runs the same arithmetic on your own margins in a minute, and its answer is the number to write at the top of the sale plan, where everybody who argues about depth can see it.
03 Scope the code before you deepen it
Every customer who would have bought anyway takes the discount and gives nothing back, and on Black Friday that share is large: loyal customers wait for the date, and people who were going to buy in December buy in November instead. A sitewide 20% code hands the discount to all of them. Scoping keeps most of the effect and removes most of the giveaway: a code for first orders, for one category, for slow stock or for baskets above a threshold, and a deeper offer only where the extra volume is genuinely new.
Two configuration habits matter more than the headline depth. Put an end date on every code, so it does not circulate on coupon sites into December, and decide whether codes stack with sale prices, because a 15% code on top of a 20% markdown is a 32% discount on exactly the products that can least afford it. The full arithmetic of a single promotion is in what each code costs before it sells anything, and a compare-at price left in place after the sale keeps reporting a discount for months.
Be careful with last year's evidence as well. On an example store's own order book, orders with a shallow discount averaged €102 of net revenue against €93 for orders with none, which looks like proof that codes raise baskets and is mostly selection: threshold offers and cart-level codes pick the larger baskets in the first place. The full caution is in discounted orders are bigger orders, and it applies to every report that compares discounted orders with the rest.
04 Raise the ad floor for the discounted weeks
Break-even MER is net revenue divided by the contribution an order keeps before advertising: the revenue every ad euro must bring before the ads pay for themselves. A discount lowers the contribution faster than the revenue, so the floor rises exactly when budgets usually rise too. On the same store's margins it moves like this:
| Discount | Break-even MER |
|---|---|
| None | 2,05 |
| 10% | 2,31 |
| 15% | 2,51 |
| 20% | 2,77 |
| 30% | 3,71 |
A campaign that ran comfortably at a blended MER of 2,5 in October loses money at the same 2,5 during a 20% sale, and no platform dashboard will say so, because each one reports the revenue it claims rather than the margin behind it. Set the sale's floor before the budget, judge the weekend on blended numbers from your own revenue and every platform's billing page, and use a MER calculator to redo the floor on your own figures. Ad platforms judge the same floor as ROAS, return on ad spend, against the purchase value they report, with VAT and shipping inside it, which is why the break-even ROAS you set in an ad account sits above these figures, and a break-even ROAS calculator gives you that figure for your own order.
In nouz the floor is kept for you. Insights, Marketing shows the store's break-even MER beside POAS, profit on ad spend, recomputed every night from the margins the statement already holds, and the Products tab carries a breakeven ROAS for every product, which is the figure to put into a product-level campaign rule before the sale rather than after it.
05 Check the shipping offer against the bigger basket
Free shipping thresholds are set for ordinary months, and a sale month changes their arithmetic twice. The discount shrinks what each extra euro of basket keeps, so a threshold that paid for itself at full price may not at 20% off, and the carriers add their own surcharges in the fourth quarter. Rerun the free shipping threshold calculator with the sale's discount taken off the margin before deciding whether to lower the line for the weekend.
Read your carrier contract for the season too. Fuel and peak surcharges, oversize rules and the cut-off dates for Christmas delivery all change what the same parcel costs in December, and the time for negotiating a carrier rate card is months before the season, not during it. A parcel price that rises by fifty cents in the busiest weeks takes it from every order the sale brings, discounted or not.
06 Order the stock against your lead times
A sale that sells out on Saturday morning has spent its advertising on a stockout. The point to order at is daily sales times the supplier's lead time, plus a buffer, and the trap is using an ordinary month's daily sales for a week that sells several times as fast. Work out the sale week's expected units per product, then count back from Black Friday by each product's lead time: a supplier who needs six weeks has to have the order by mid-October.
The reorder point calculator does the division for one product, and a safety stock sized for the sale protects the products whose sell-out would hurt most. Leftover stock after the sale is the opposite mistake and a quieter one, because it is cash parked on a shelf until spring, and the products most likely to be left over are the ones bought deepest for a promotion that did not happen as planned.
nouz's Forecast page does this for every product at once, in a purchase plan that takes each variant's forecast units, which include last year's Black Friday season and any sale you add as a planned event, counts the stock already on order as arriving at the end of the product's lead time, and names the products to order today and how many, priced at your own unit costs.
07 Put the sale into the forecast before it happens
A forecast built from your own last year already contains last November. nouz's Forecast page looks for last year's Black Friday season in your own daily orders, the run of unusually busy days around the day itself measured against the weeks before it, and carries it into this November, where you can switch it off if this year will be different. A sale you are planning goes in as one of the forecast's planned events, with its dates, its discount, the extra orders you expect and its ad budget, and the page shows what it moves on net revenue and EBITDA before anything is committed.
Run it twice: once with the discount and the extra orders you hope for, once with the extra orders you can defend from last year. If only the hopeful version clears your EBITDA goal for the month, the sale is a bet rather than a plan, and it is far better to know that in October. The method behind the forecast itself is set out in forecasting sales from your own orders.
08 Expect the refunds in December
Returns lag sales by weeks, and a sale's returns lag furthest, because impulse purchases cool off. On an example store's own order book the median refund arrived 18 days after its order, only 21,7% within the first week and 37,7% in the third and fourth weeks, so a Black Friday order that comes back most likely comes back around 15 December, in the middle of the gift season's busiest days.
Two consequences follow. Read November's margin only once its returns have arrived, at the same age you read other months, or the sale will look better than it was; the shape of the lag is in when refunds actually arrive. And hold a refund reserve sized on the sale's own revenue rather than an average month's, because the reserve exists for exactly the month after a spike.
09 Plan the cash: stock before, VAT after
The sale moves cash on three calendars at once. The stock for it is paid before it sells, on your supplier's terms. The VAT collected during the sale leaves later in one payment: in Germany a monthly filer pays November's VAT on 10 December, or on 10 January with the one-month extension, and in Austria on 15 January. And the refunds drain December. A store that reads its bank balance in the first week of December can feel rich and be short by the middle of January.
Put all three on one calendar before the sale, with the cash you have today at the top. The full method, with the VAT rules of Germany, Austria and Switzerland, is in the VAT side of a cash flow forecast; in nouz, the cash plan on the Forecast page's Stock & cash tab carries the same plan month by month, with the stock payments from the purchase plan and the VAT return on your own filing cadence.
10 Decide in advance how the sale will be judged
Write the test down before the sale starts, because afterwards every number will look like success. Compare the sale period's contribution, not its revenue, against a comparable period without a sale, once its returns have arrived. Count the share of discounted orders that came from customers who were already buying. And judge November as a month, not the weekend: a sale that lifted revenue and lowered contribution is a loss with good publicity.
If you keep margin targets on the P&L, expect the sale days to show orange or red against them; that is the discount doing what it was meant to do. The question is whether November as a whole still clears its own targets once the returns are in, which is the same question this checklist started with: what did the sale leave behind?
The checklist in one place
- Measure what an order keeps after goods, parcel and payment fee, in a normal month.
- Price each discount depth in extra orders, and write the requirement at the top of the plan.
- Scope the codes, give them end dates, and decide whether they stack with sale prices.
- Raise the break-even MER for the sale weeks before setting the budget.
- Rerun the free shipping threshold with the sale's margin, and check the carrier's peak surcharges.
- Order stock against the lead time of each product, counting back from 27 November.
- Put the sale into the forecast as a planned event, with a cautious version beside the hopeful one.
- Expect the refunds in December, and hold a reserve sized on the sale.
- Put stock payments, the VAT return and the refunds on one cash calendar.
- Write down how the sale will be judged before it starts, and judge it on contribution after returns.
None of this argues against Black Friday. A well-scoped sale on known margins, with stock that arrives and a cash plan that expects December, is one of the best weeks of the year. The difference between that and a busy weekend that quietly cost money is decided almost entirely in October, on a page of arithmetic that takes an afternoon.