Editor’s note: Brave Bison, which acquired Mark Ritson’s MiniMBA in 2025, owns about 28% of System1 and sought to acquire the rest this year.
Joy is stuck at the airport on Christmas Eve, the snow coming down hard, and every flight cancelled. So she picks up her bag of presents and sets off for home the long way: a bus, a lift from some donkey farmers, a night in a boat cabin on a heaving sea, a taxi driven by a giant. At every stop, she thanks a stranger with one of the Amazon parcels meant for her family.
This is “Joy Delivered”, Amazon’s new holiday ad, which launched earlier this month and runs like a proper movie. A new version of the Beatles’ “Golden Slumbers” does a lot of the emotional heavy lifting and the hug at the end will get you.
There are two problems with the ad, however. One is peripheral. One central.
The peripheral one first. Amazon filmed its white Christmas in New Zealand, in and around Dunedin, to be specific. But it’s summer down there, and it will be even warmer by Christmas, when most of the country will be at the beach in a pair of shorts.
The central problem is bigger. Amazon already has a great Christmas ad: “Joy Ride.”
Three old ladies sitting on a snowy hill watch the local kids go sledding. One of them orders some seat cushions from Amazon. The box turns up and the next thing you know they are flying down the hill and, for a few seconds, they are girls again. There are no words. Just an instrumental of “In My Life.”
“Joy Ride” first ran in 2023 and got the maximum 5.8 stars from System1. Basically perfection.
Then Amazon did something very unusual for a big marketer. Last year, instead of making a new ad, it ran the old one again. It scored 5.7, again putting it in the rare company of less than 1% of all ads made all year.
None of this should surprise anyone. Coca-Cola first ran “Holidays Are Coming” in 1995. The trucks have rolled out every Christmas since, and System1 says the ad always scores five stars in its testing. When System1 looked at more than 50,000 ads, it found that scores barely move as ads get older, and that some of the very best actually improve as people get to know them.
Ads don’t wear out, at least not for consumers.
Amazon’s problem is that its advertising decisions are made by marketers — people who like to make new, shiny stuff. Jo Shoesmith, Amazon’s global chief brand officer, told ADWEEK that people want nostalgia but also “something they haven’t seen before.”
I call that a complete contradiction.
The average brand building campaign reaches its maximum impact about two to three years into execution. Seasonal stuff can often, as Coke has proved, live on for decades. But the average marketer is tired of their ad, and wants to make something new.
It’s the same story with slogans, with logos, with distinctive brand assets. Marketers forget they are not the consumer. They look at their ads for at least 9,000 hours before they release it into the wild, and are over it before most of the target market has noticed it.
There is an intermediate position. Amazon could pull a Coca-Cola: bring out their old ad every year but also make a crappy, inferior new execution alongside it. But why create a bat to beat yourself with? Better to remove the outstanding, already made, already free, already proven, already much-loved Christmas ad completely, and then run something less good, with lots more risk and cost attached. Makes perfect sense.
We read all the time that marketers need more support from their CEO, more budget, more seats in the boardroom.
But what they mostly need is patience. It takes years, not months, for a good campaign to work. Amazon knew that last Christmas, but it has already forgotten it.

