Pick Your Black Friday

Bettina Dargie ยท 2026-09-21

Pick Your Black Friday

Black Friday used to mean a sleeping bag, a 5am opening and a queue that formed on Thursday for a doorbuster television. In 2024, for the first time, more people shopped on the weekend online, 87.3 million of them, than went into a store, 81.7 million.

The queue turned up somewhere else. On Tuesday, September 15th, people slept outside Starbucks for the Snoopy cold cup: $39.95, two per customer, sold out the same day, trading around $200 on StockX by the afternoon. They used to queue for a markdown; these days, they queue because there aren't many.

The weekend itself is in fine health. Shopify merchants alone did a record $14.6 billion over it last year, up 27%. What's changed is what gets people to show up, and most brands are picking between three plays for November right now. About fifty years of research has something to say about each one.

The straight discount

Every shopper carries a rough sense of what you charge, built from every price they've seen from you. Discount often enough and that number settles at your sale price, and your regular price starts to look like a markup.

There's a sharper problem underneath. In 1978, Dodson, Tybout and Sternthal went through purchase panel data on two supermarket products. Customers brought in by a cents-off deal were less loyal after it ended than if there'd been no deal at all. Their reading was that someone who buys on a deal decides the deal is why they bought.

A sitewide markdown aimed at strangers is the only one of the three plays with half a century of evidence that it costs you something after the weekend. It still moves volume. Price that in.

VIP access for people who already buy from you

The same 1978 study found the opposite effect from a different mechanism. Coupons printed on the packet, which you could only use by buying the thing first, built loyalty that held when the offer was withdrawn.

Early access for your list runs on the same instinct. The reward goes to someone who has already chosen you, so there's nothing for them to pin the purchase on except the choice they'd made anyway. It also leaves your public price alone, which the straight discount does not.

The limited drop

In 1975, Worchel, Lee and Adewole did an experiment by sitting people down with a jar of cookies. Some jars held ten and some held two, and the near-empty jar rated better.

There are two details from this specific study that could direct how a Black Friday drop gets run. The study says a jar that started full and emptied beat one that had always been nearly empty, so scarcity you leave switched on all year stops doing much. It also says the reason for the emptying changed the rating: told everyone else had wanted them, people rated the cookies highly; and when told a researcher had carted most off by mistake, they rated them lower. 

A run that sells out because people wanted it beats a run that was simply small. Say why there are only so many.

Parke does this on a schedule rather than once a year. Bootstrapped, no ad spend, $16 million a year, sold out in under five minutes on the first of every month. The jar refills on a date everyone knows, then empties in public.

And whichever play you pick, skip the countdown timer

Hmurovic, Lamberton and Goldsmith went looking for time pressure online in 2023 and, by Lamberton's account, never got a ticking clock to beat the same offer without one. Shoppers read it as pressure and shrug it off.

Which one fits what you sell

A 2022 meta-analysis in the Journal of Retailing pooled 416 results from 131 studies. On things bought for pleasure, limiting the quantity was about three times more effective than simply telling them it was popular. Nearly the same gap was the result for anything used where other people can see it. For the things people buy because they ran out, limiting the quantity did nothing measurable, and popularity was the only factor that moved them.

Services land within the same logic. Limiting supply actually worked on experiences better than on physical goods. For example, a photographer with only four December slots and a restaurant with a set menu for one weekend would perform similarly as a physical good purchased for pleasure. Both of which can skip discounting as the impression of scarcity will do the work.

So, two questions before you build anything.

One. Is what you sell bought for pleasure?

Two. Is it used where other people can see it?

If you said yes to either one

Run the drop, at full price, and open it to your list first. You get the mechanism the Snoopy cup used and the loyalty effect the 1978 packet coupon found, without touching your everyday pricing. Those two questions came out of the research separately, each holding up on its own, so one yes is reason enough to try it. Starbucks has form: its Bearista cup last year drove an estimated $64 million in extra sales, with daily transactions up 16.3% the week after.

If you said no to both

A limited run of something people buy because they ran out will do nothing, and the research is fairly blunt about it. Popularity is what moves that shelf, so lead with reviews, reorder rates and how many people already buy it. Ritual took a multivitamin past $250 million in gross revenue in 2024 by publishing where every ingredient comes from, which is the same job done with evidence instead of shortage.

If you discount anyway, and plenty of brands on this shelf will, point it at people who've already bought from you rather than at the whole internet. Starbucks got a queue at 3am without discounting anything. Worth putting that option on the list before November.