Fallacy number 5001: Peak became a plateau. Shoppers didn't get the memo.

In the latest episode of us lot in the industry swimming in circles like goldfish: fallacy number 5001, peak being peak.

In the mid-to-late 2010s, peak was Black Friday, and over the years it stretched to Cyber Monday. Four days, a tidy long weekend to make bank.

Then the industry got K-holed on discounting and FOMO, and stretch-armstronged the whole thing at both ends, starting earlier and running past Cyber Monday. That's what retailers did. What consumers did is a different question. More on that in a bit.

Once one brand goes early, the rest follow, because nobody wants to be the one still waiting while everyone else discounts. Watch this year: I'd put money on Black Friday comms landing straight after Scream 11 (is that where we're at now? Wild guess) finishes on ITV2 on Halloween night.

Peak has grown new ridges too. There's Singles Day, a clever Chinese export that turned a joke about being single into the biggest shopping day on earth. There's Click Frenzy, from our friends Down Under.

So. What "peak" did consumers actually have?

Not what the retailers did. The story everyone repeats is that peak spread out, so shoppers must have started buying earlier too. It lines up neatly with the discounting calendar. In the UK, at least, it's not true.

Last year, 58 UK retailers had Black Friday campaigns live by 1 November, against 12 in 2021, according to IMRG. DataWeave tracked 71,642 UK products and found retailers had front-loaded their discounts well before Black Friday week. IMRG's online revenue, year on year, barely moved: down 1.6%, 0.2% and 1.9% across the first three weeks of November, then down 1.2% in Black Friday week itself. Weeks of deals, and not one week got a lift from them. That's what retailers built: a plateau.

Shoppers didn't buy it, literally. The BRC recorded October's weakest growth since May, with shoppers holding off on toys, electronics and clothing. Barclays reported November card spending down 1.1%, the steepest fall since February 2021, and put it partly down to shoppers "awaiting seasonal discounts". The discounts were already live. They waited anyway.

Then Black Friday arrived and behaved exactly like Black Friday. Adobe measured it as the biggest online day of the UK year, at £1.16 billion. Barclays called it the busiest day of the year so far, with transaction volumes up 62.5%. One graph is flat. The other has a spike standing straight up out of it on 28 November, and it's the shoppers' graph.

It didn't stop there. On Super Saturday, the day the industry expects to be the biggest of the year on the high street, in-store visits fell 6.9%. Boxing Day footfall rose 4.4%, the strongest growth in a decade. January retail grew 2.7% because, in the BRC's words, plenty of shoppers "had held off Christmas spending and waited for the January sales".

It's the same in the US and Europe. Everyone is at it.

US retailers started in October, with Amazon's Prime Big Deal Days on the 7th and 8th. DataWeave checked nearly 80,000 US products and found the same shape as the UK: electronics were already 14.6% off before Black Friday week, and the week itself added just 2.6 points. Meanwhile, in the first week of October, 63% of American shoppers told the National Retail Federation they'd wait until Thanksgiving weekend to do most of their holiday shopping, up from 59% the year before. The deals came early, and a bigger share of shoppers planned to wait anyway. Cyber Monday then became the biggest online day of the American year: by my maths on Adobe's numbers, about 3.4 times an average day of the season.

America is where the "buying earlier" crowd, my former self included, have a case. Adobe had US online spend in October up 8.2%, faster than Cyber Week's 7.7%, and Bank of America reckons earlier promotions "may be encouraging consumers to spread their shopping over a longer, and earlier, period". Fair enough. Some of the growth did turn up early, but the biggest days kept their spot. At the other end, Salesforce had US online sales in the last two weeks of December up 9%, more than double the growth for the season as a whole.

Parts of Europe are spikier still. Dansk Erhverv, the Danish trade body, found a quarter of online shops doing Black Friday had offers out from 1 November. Danish Black Friday still did 6.1 times the transactions of an average Friday in the five weeks before it, on Adyen's payments data. The US managed 2.8. In Sweden, clothing sales grew just 3% in October and 7.5% in November, and Svensk Handel, the Swedish retail trade body, put the slow October partly down to customers waiting for the Black Week offers to start. Worth asking yourself whether customers trained to wait for the offer are the ones you want.

In Spain, the consumer group OCU tracked 2,475 electronics and appliance products and found that only 9% were genuine bargains on Black Friday itself, while 37% cost more than their lowest price in September and October. Spain still did 4.5 times a normal Friday on Adyen's measure.

The Dutch show you whose timetable it really is. In 2024, the busiest day of the Black Friday period on iDEAL, the Dutch online payment system, wasn't Black Friday. It was Monday 25 November, right around payday.

So… the plateau is real. It's just sitting under the retailers, not the shoppers. Every ridge, every blackened ad campaign, every discount live from 1 November, and the money still arrived in spikes, on its own timetable, mostly at the end.

This year there's a sinkhole too, and it lands right where your P&L doesn't want it. Grand Theft Auto VI comes out on Thursday 19 November, eight days before Black Friday, with the first full weekend of play on the 21st and 22nd, precisely when you were planning to get your pixels firing en masse. A meaningful chunk of the male population will be unreachable, unwashed, and busy maintaining a relationship with a fictional woman called Lucia.

I wouldn't build a media plan on this, and I'm not going to pretend the evidence exists. I couldn't find anyone who has measured a game launch against retail conversion. What has been measured is adjacent and mildly encouraging. At the 2023 Super Bowl, Cloudflare found food delivery traffic 19% below baseline at kickoff, and reckoned the orders had been placed earlier rather than lost. Barclays found UK pub and bar transactions up 195.6% on the night of the Euro 2024 final, in a month when retail fell. Big attention events don't destroy the money. They move it about. So the play isn't to write anyone off. It's to be in front of them the week before, while they can still use a card, and again in the first week of December, when a certain kind of man emerges blinking from a fortnight in Vice City and realises he has bought nobody a present.

Here's a playbook, of sorts, for peak:

Get your plumbing ready now. Data connections, pixels, audiences, feeds, clean email lists, product data. The standard advice is fine; it just never comes with a deadline, so it drifts into November and loses to everything else. Whatever isn't firing cleanly by the end of October won't get fixed in November.

Have your winning creative ready. Nobody thinks it strange to spend a million pounds on a TV ad and run it four hundred times: reach, reception, recall. Put the same logic to a social team and it flips, because the people telling you content "wears out" are the people paid to make more of it. Bank your proven ads in October and cut new versions from them, instead of starting from scratch in November.

Spend into the tail end. Your November non-converters aren't lost buyers. Most weren't ready yet, including everyone in the GTA gutter. Have urgent messaging ready for December: it's a less contested window, and usually cheaper in the ad auctions.

Go and ask your customers. I got the "shoppers are buying earlier" line wrong. I said it with total confidence, for years, and only found out when I checked. Most peak advice, mine included, runs on last year's anecdote instead of this year's customer. Before you brief the agency, stand in a shop or get on the phone, and ask ten actual customers what would make them buy from you in November. Otherwise you're going into Blind Friday.

One more thing, and it's the one I want brands to take from this. Every peak calendar defaults to the same lever: money off. Discount earlier, discount deeper, discount longer. It's a bloody disease at this stage.

There is a cure, though. Look at Record Store Day. It doesn't discount anything. It works because things are rare and specific: a vinyl pressing you can't get anywhere else, on one day, in select shops. People queue overnight for that. Nobody queues overnight for 15% off.

Peak doesn't have to mean cheaper. It can mean early access, a limited run, something built for the people who already buy from you rather than a wider net at a lower price. Value doesn't have to be money off. It just has to be something your competitors aren't also offering on the same Friday.

Be interesting for your customers this year.