Andy Warhol put it best. “You can be watching TV and see Coca-Cola, and you know that the President drinks Coke, Liz Taylor drinks Coke, and just think, you can drink Coke, too,” he declared in 1975. “A Coke is a Coke and no amount of money can get you a better Coke than the one the bum on the corner is drinking. All the Cokes are the same and all the Cokes are good.”
That is true not just for Coca-Cola or, for that matter, Campbell’s soup or Brillo pads. One could say much the same about a Hollywood movie, Gmail, Ikea bookshelves, Microsoft Office, the iPhone, the Uber app, a Big Mac, a Moderna vaccine shot and YouTube. If Warhol were still here to opine, he might object that you can splash out on a private movie screening or pay for a premium subscription to YouTube. True enough. But the broad principle applies: in the unlovely jargon of Silicon Valley, these products and services all “scale”.
But not everything does. Researchers and policy wonks have long studied pilot schemes such as criminal rehabilitation programmes, public health initiatives or innovative schools. They dread the familiar phenomenon of the pilot delivering sensational results, only to fade at a larger scale. This dismaying tendency was called “voltage drop” by the psychiatrist Amy Kilbourne and her colleagues in 2007.
The economist John List has been exploring the causes of this voltage drop, first in a 2019 paper with Omar Al-Ubaydli and Dana Suskind, then in a recent book, The Voltage Effect. List is well qualified for the task, with a mix of theoretical, policy and Silicon Valley experience. He is admired for conducting realistic experiments to shed light on economic theory and policy; he set up a pre-school in Chicago with the aim of “establishing a model that other school districts around the world might some day be able to implement”; and he is chief economist at Lyft, a position he previously held at Uber, two of the companies attempting to scale the process of getting driven around.
So why does the voltage drop for so many promising ideas? One common problem is that the original effect was illusory. Consider a famous experiment, conducted over 20 years ago by psychologists Sheena Iyengar and Mark Lepper, in which customers in a high-end supermarket were offered free samples of jam from a choice of either six or 24 flavours. The wider choice was dramatically demotivating. Ten times as many people bought jam after being shown the smaller range. It is one of the most famous results in psychology; it has also proved rather difficult to repeat in follow-up experiments. Perhaps the effect is completely non-existent, the result of a statistical fluke. Or perhaps the effect exists but with nothing like the force exhibited in the original experiment. Does anyone seriously believe your local supermarket would sell 10 times as much produce if only it simplified its product line?
Another source of voltage drop is when the original effect does not generalise beyond unusual circumstances. My favourite example is the Arch Deluxe, a gourmet hamburger launched by McDonald’s in 1996 with a marketing fanfare. The fast-food giant had every reason to expect success, because focus groups loved the Arch Deluxe’s bakery-style rolls, peppered bacon and stone-ground mustard-mayo dressing.
The problem, says List, is that the focus group enthusiasts were not a good guide to the attitude of the typical consumer: “A person who signs up to take part in a McDonald’s focus group is probably someone who is crazy about McDonald’s or loves all kinds of burgers, or both. But the average person, it turns out, goes to McDonald’s for the Big Mac, not a fancier version of one.”
Even if the idea is real, and generalises to a wide audience, it may be difficult to repeat the performance once it ventures beyond the control of the original creative team. A pilot school may work well, but it is easier to hire 20 good teachers than 20,000. A brilliant chef can work in only one kitchen at a time. I was once served a meal fresh from the research kitchens of a fast-food multinational. It was vastly better than I would expect on my local high street, which, alas, is a challenge to the whole idea of a research kitchen.
Yet pinning down a single explanation for voltage drops is impossible. Indeed, the wisest sentence in List’s book is the Tolstoy-inspired epigraph, “Scalable ideas are all alike; every unscalable idea is unscalable in its own way”. Quite so.
The world is a big, complex, bewilderingly diverse place. All the Cokes are the same and all the iPhone 13s are the same. But schools and restaurants and police departments and museums and comedy gigs and clinics are not much like Cokes or iPhones. Perhaps the mystery is not that ideas often fail to scale. The mystery is that we ever convinced ourselves that they should.
Written for and first published in the Financial Times on 25 February 2022.