Should you run a Black Friday sale?

Black Friday is the one promotion most stores never actually decide to run. It just arrives, everyone else is doing something, and the question quietly changes from whether to how much off.

It can be a genuinely good week. It is also the single easiest promotion to misread, because the reporting is generous, the comparison is usually against nothing, and the costs land in a different month from the revenue.

So here is how to work out whether yours is worth running, before you commit to a depth in October.

The question the report does not answer

Your promotion report will tell you what the sale sold. That is not the question.

The question is how much of it would have happened anyway.

A store that sells 120 units in a normal week and 300 units in the sale week did not gain 300 units. It gained 180, at best, and probably fewer, because some of those 180 were people who were going to buy in early December and have now bought in November instead.

Measured against zero, every promotion works. Measured against baseline, a good number of them turn out to be a busy week that cost margin. If you cannot show the promotion drove sales that would not otherwise have happened, it was not a promotion. It was a margin reduction with a countdown timer on it.

Setting the baseline before you start

Do this in the last week of October, while it is still a decision rather than a post mortem.

Take the four weeks before the promotional window and work out average weekly units and average weekly profit. Four weeks, not one, because a single week is noisy. Adjust upward if November is normally busier than October for you, and be honest about the adjustment rather than generous.

Work out your break-even from the depth you are considering. Margin divided by margin minus discount, which is covered in full in how to calculate a discount break-even.

Write both numbers down where you will see them in December. Baseline units, and the units the promotion needs. That single act is what stops the same optimistic sale running again next year on the strength of a top-line number.

A worked example, including the part everyone leaves out

A product sells at £40 on a 50 percent margin. It costs £20, so it earns £20 a unit. Normal trade is 120 units a week.

You run 25 percent off. The price drops to £30, the cost stays at £20, and you now earn £10 a unit. Your profit per unit halved, which is what a 25 percent discount does to a 50 percent margin.

Break-even is 50 divided by 25, which is 2. You need to sell 240 units, double your normal week, just to stand still.

The sale week does 300 units. That is 300 times £10, so £3,000 of profit, against a normal week’s £2,400. You are £600 up, you cleared break-even by 60 units, and the report looks like a success.

Now look at the fortnight after. Sales settle at 70 units a week rather than 120, because the people who were going to buy in early December already bought in November. That is 50 units a week short, for two weeks, at £20 a unit of full margin.

Sale week gain        + £600
Two quiet weeks       - £2,000
Net                   - £1,400

The promotion beat its break-even and still lost £1,400, because the break-even was calculated on the sale week alone while the cost was spread across three.

This is the most common way a Black Friday result gets misread, and it is invisible unless you deliberately include the weeks either side. Add the fortnight before too, since customers who know the sale is coming stop buying in the run-up.

Four things to decide before the depth

Which products, and which are excluded. A blanket percentage across a catalogue with a twenty point margin spread means some products are being sold near cost and nobody chose that. Run the break-even on your thinnest margin line first, not your best one, and exclude anything that cannot carry the depth.

How deep, based on margin rather than on what everyone else is doing. A 30 percent cut on a 40 percent margin product needs 300 percent more units. That was never realistic at any level of enthusiasm, and no amount of traffic fixes arithmetic.

How long. A short window concentrates demand and makes the comparison easier to read. A month of rolling offers is indistinguishable from having lowered your prices, and your customers will treat it that way next year.

What you will do about the people who were going to buy anyway. Some proportion of your sale orders are your most loyal customers, the ones who read every email, buying something they would have bought at full price in a fortnight. Every one of them is pure margin handed over. You cannot eliminate this, but you should size it before you decide the depth, because on a well-engaged list it is a big share of the total.

What you are teaching, and to whom

The margin cost is one year’s. The habit is not.

Run a sale every November and your customers learn to wait for November. The dip in the weeks before the sale gets a little deeper each year, and the sale has to get a little bigger to compensate, which deepens the dip again. Stores that have been at this for five years often find the event no longer produces incremental revenue at all. It just relocates a chunk of Q4 into one week and charges them 40 percent of the margin on it.

There is a second cost. Customers acquired in a heavy discount window tend to be price motivated, and they churn hardest at the point the discount stops. Check yours rather than assuming: split the November intake from the rest of the year and compare repeat purchase rate six months on. If the Black Friday cohort never comes back, the event is an expensive way to buy one order. How discount codes train customers to wait covers the pattern, and what a customer is worth is what makes the cohort comparison meaningful.

If you want to take part without a sitewide sale

The pressure to do something in November is real, and ignoring the biggest shopping week of the year is a legitimate thing to feel uneasy about. There is a middle option, which is to have something to announce that is not a percentage off everything.

Early access rather than a lower price. A new product, or a limited run, released to your list before anyone else. Urgency without discount.

A bundle priced below its parts but above single item margin. The customer gets a real saving, you get a higher order value, and the individual product prices never move.

A bonus item with every order over a threshold. Costs you wholesale rather than a percentage of every sale, and there is no lower price for anyone to wait for next year.

Free shipping for the week. Customers weigh delivery cost far more heavily than its size deserves, and it costs you actual postage rather than a slice of every order.

Each of these gives you a November email that is not a discount, which means next November starts from where you are now rather than one notch deeper. The full list of alternatives has what each one costs.

The decision, in four lines

Run it if the incremental volume you need is a number your last comparable promotion actually hit, on products whose margin carries the depth, in a window short enough to read, having sized what you are handing to customers who would have paid full price.

Otherwise run something that adds value instead of cutting price, and keep the margin.

Either way, measure it against the four weeks before and the two weeks after, and write the result down. Next year’s version of this decision is much easier when you are looking at what actually happened rather than at what the promotion report said in the moment.

The wider case on discounting covers the other three situations where a discount is genuinely the right call.