The like stuff
In a world of likes, shares, thumb sups and star ratings, our personal taste is subject to constant assault by the tastes of others.
In a world of likes, shares and star ratings, our personal taste is subject to constant assault by the tastes of others.
Tom Vanderbilt was walking his five-year-old daughter to school one morning when she asked him what his favourite colour was. Without really thinking, he gave a quick response. Blue. “It was the answer that popped into my head,” he explains. “It had always been there as a narrative in my life, this blue preference. Then I started to wonder whether it really was a preference or if I’d talked myself into it, if saying blue was my favourite made things like shopping for clothes easier, whether I’d come up with a shortcut that I didn’t necessarily believe in. And if blue was my favourite, where would a preference like that come from? These were the thought processes that unlocked when I gave my daughter this answer I thought was a little bit too quick.”
This simple, everyday exchange sent Vanderbilt, a New York-based journalist and author, on an exploration of the science of preference that took him inside the worlds of military food engineering, pet-show judging, art galleries and online entertainment companies like Netflix. In each place he was trying to find out why we like what we like, and what it means for us, and he documents his exploration in a new book, You May Also Like: Taste in an Age of Endless Choice.
Our taste, so personal to us, so essential to our identity, is something we have very little control over.
THE AGE OF EVALUATION
Never has our personal taste been so public. We live in a world of likes, thumbs-ups and star ratings. Even before he had that conversation with his daughter, Vanderbilt had been feeling as if every day he was weighing up some new product or service that had come into his life.
“And not only that, but I was being exposed to what other people felt,” he says. “[There was] this torrent of opinion, what they felt about
restaurants, hotels, a product on Amazon, films, books.”
It is the sheer size of what he calls “this general tidal swell” of opinion that makes his central discovery so interesting: our taste, so personal to us, so essential to our identity, is something we have very little control over. When it comes to what we like, we are easily swayed, prey to unconscious bias and social influences, full of contradictions. We rarely understand why we like what we like. And we may not like it tomorrow, says Vanderbilt. “So many of the things out there that we think we choose and enjoy are imbued with hidden or not-so-hidden contextual factors that may really be influencing our liking without us thinking about it.”
A key part of marketing any product or service is manipulating our preferences. But people’s likes can be notoriously tricky to predict or influence. Vanderbilt examines the case of Crystal Pepsi, a clear soft drink released in the early 1990s.
“There was a trend at the time for clear products and Pepsi decided to capitalise on it,” he explains, “but it didn’t realise that it was going to be violating people’s expectations in all sorts of ways. Consumers liked it when they couldn’t see it [in a blind tasting], but once they realised it
Humans are omnivores, born with a wariness of foods that could potentially be toxic, and so we’re primed to seek familiarity.
had no colour, that violated their expectation of what it should taste like. The very fact it was called Pepsi led them to believe it would taste more Pepsi-like but it was slightly different.”
Very quickly the fizz went out of Crystal Pepsi and it was taken off the market. The reason it failed, says Vanderbilt, can in part be traced back to evolutionary theory. Humans are omnivores, born with a wariness of foods that could potentially be toxic, and so we’re primed to seek familiarity. As Vanderbilt puts it: “What didn’t kill you yesterday was what you liked today.”
The single biggest predictor of whether we like a food is whether we’ve had it before. Our liking is tied up with memory: when we order a favourite dish at a restaurant, we are looking back to the last time we enjoyed it and expecting to like it again. Crystal Pepsi failed because it played around with those expectations of what is familiar.
And yet, we humans can be perverse creatures. At some point, familiarity becomes boring and we start craving variety. Evolutionary theory has an explanation for that, too.
“If a food source was depleted, then suddenly you couldn’t rely on the familiar,” says Vanderbilt. “So the argument is that the brain must also be primed to seek novelty and be excited by it in some way.”
TALKING OURSELVES INTO IT
Psychologists have shown how our liking mechanisms can be directed. For one study, published in the British Journal of Nutrition four years ago, researchers were able to change how much people liked a microwaveable chicken meal after they had eaten it simply by having the subjects rehearse “the enjoyable aspects” of the meal. As a result of talking about what was good about the food, not only did it seem better, but people wanted to eat more of it.
To understand the challenges of making food more likeable, Vanderbilt went to the US Army’s Soldier Systems Centre in Natick, Massachusetts, where they produce readyto-eat meals with a minimum shelf life of three years that can be airdropped into the field and that soldiers will find acceptable to eat.
There is a greater chance we will like something when we expect to like it, and few people expect to enjoy military rations. A Natick study had soldiers eat in the dark and found they liked things more when they were told what they were eating. In another study, people were given a salmonflavoured ice cream that was labelled either as “ice cream” or “frozen savoury mousse”. The mousse eater liked it far more than the ice cream eaters, many of whom described it as disgusting.
“Experiments have shown that the same bottle of wine is ranked lower when it’s labelled as coming from New Jersey than when the label reads Bordeaux,” says Vanderbilt. “We’re often construing these mental models of things before we’ve had a chance to consume them, based on what we expect from previous experience. It’s the role of branding in a way to build a whole picture of expectation so you don’t have to.”
In fact, flavour has far less influence over whether we like a food than you might imagine. How it looks and even where we eat it has an effect. Psychologist Debra Zellner presented the same meal plated up
Our tastes define us. They are intrinsically tied up with who we think we are and who we want to be.
conventionally and also with more flair to participants in a study. The people whose food got the special treatment reported liking it more. In Natick, they found food is rated as better when served in a restaurant than in an institutional cafeteria. Even the time of day has an influence on what we like to eat. Dutch researcher EP Köster, who has
examined the psychology of food choice, has observed that the most adventurous of foodies will eat the same thing for breakfast every day but by dinnertime they’re seeking something more exciting.
DECISIONS, DECISIONS
Our tastes may have started out as a basic survival tool, but in modern life they have evolved into something far more complicated. Cornell University researcher Brian Wansink has estimated we make 200 decisions a day just about food. At the same time, we’re choosing what we wear, listen to, watch and do: we are engaged in a constant process of being influenced and forming preferences.
“It’s pretty hard to have an expectationfree experience,” Vanderbilt says. “In part thanks to social media, everything comes wrapped in recommendation.”
To find out how electronic word-ofmouth has changed the landscape of what we like, he visited Netflix’s headquarters in Los Gatos, California. There Vanderbilt discovered the difference between what we say we like and what we actually watch. As Todd Yellin, the company’s vice-president of product innovation, explained, people might give a five-star rating to a quality drama like Hotel Rwanda and two stars to a super-hero flick like Captain America, but they are far more likely to actually watch the latter.
Given that the Netflix data is private, the only people we’re lying to about our taste here is ourselves. That is just one reason companies like Netflix are de-emphasising the five-star system. It is unreliable because it is filled with bias. For instance, people avoid the ends of the scales – so you get more two- or four-star reviews than oneor five-star ones. Plus we have a positivity bias – on the book-recommendation site Goodreads, the average rating is 3.8 stars and on TripAdvisor it’s 3.7.
We also have a tendency to imitate, so we’re more likely to rank things highly if other people before us have (although paradoxically, when a novel wins a big literary prize, its ratings on Goodreads often get worse). Sociologist Amanda Sharkey and professor of organisational behaviour Balázs Kovács analysed this and concluded that the prize raised expectations. It went from being a book you might like to being one you should like.
YOU ARE WHAT YOU LIKE
Our tastes define us. They are intrinsically tied up with who we think we are and who we want to be, where we belong and who we want to belong to.
“Psychologists have identified a sense of optimal distinctiveness,” says Vanderbilt. “We all want to claim membership to a certain human group, but we also want to feel distinct. In societies where you have these large bands of people who have very similar amounts of capital and access to products, how do you create this sense of distinctiveness?”
In our quest for individuality, few things have become more complicated than coffee. “Once it was a banal drink you just ordered,” says Vanderbilt. “Now I go to the store down the street and have this sense that every few months there is some new brewing technique being rolled out. Connoisseurship has drifted down into more humdrum everyday products. There’s a toothpick company here that makes artisanal toothpicks: you read the language and it almost sounds like a parody, but it’s not.
“You can view this as exciting, in the sense that we are high enough on the hierarchy-of-needs scale here that we don’t have to worry about eating and we can move on to think about our toothpicks with this new discernment. But when does the search for distinction end? And does it create happiness – or more insecurity?”
In an experiment conducted at Stanford University by marketing guru Jonah Berger and professor of organisational behaviour Chip Heath, yellow rubber Lance Armstrong Foundation Livestrong wristbands were sold in a target dorm at a time when they were becoming popular. The next week, they were sold in a dorm known as being full of geeks. A week later, the number of target dorm wearers dropped by 32%. The people from the target dorm said it was not that they disliked the geeks, only that they thought they were not like them. Ownership of that simple band of rubber was a signal of taste, and the only
We have a positivity bias – on the bookrecommendation site Goodreads, the average rating is 3.8 stars and on TripAdvisor it’s 3.7