Like it or leave it

A History of Like

By Robert W. Gehl

The marketing field has long been obsessed with likability, but Facebook may be inadvertently revealing how shallow our liking goes
If you blog, run a university home page, do e-commerce, write news articles for a local paper, have a local government site, or do nearly anything with the Internet,  you’re pretty much required to have users “like” your pages. Otherwise, you’re going to be left out of the new economy of quantified affect.  We live in what Carolin Gerlitz and Anne Helmond call a Like Economy, a distributed centralized Web of binary switches allowing us to signal if we like something or not, all powered by the now ubiquitous Facebook “Like” button.
But why “Like”? Why not “Love,” or “I agree,” or “This is awesome”? At first it seems like one of those accidents of popular culture, where an arbitrary boardroom decision eventually dictates our everyday language. In fact, one history of Facebook’s Like button presents it in these very terms: Facebook engineer Adam Bosworth noted that the button began as an Awesome button but was later changed to Like because like is more universal. If it had stayed Awesome, perhaps we’d be talking about an economy of Awesomes binding together the social Web and we would sound more like Teenage Mutant Ninja Turtles than Valley Girls.
There’s a deeper history to “like,” though, that is far older than Facebook. The marketing subfield of Liking Studies, which began before Internet use became mainstream, is key to understanding how this somewhat bland, reductive signal of affect became central to the larger consumer economy we live in. It’s also explains why Facebook will never install a Dislike button.

***

What’s the best way to predict whether an advertisement increases sales or not? The marketing field has searched for the answer to this question for decades. (If I knew, I would be the head of some unctuous marketing firm instead of a state employee.) In the early 1990s, the Advertising Research Foundation’s (ARF) “Copy Research Validity Project” proposed a simple answer: the advertisement is “likable.” The massive study compared pairs of television ads in various settings and methods, against a wide range of accepted marketing measures: persuasive elements, how well the advertisement is recalled afterword, the clarity of the message, and (seemingly as an afterthought) how well the ad was “liked.”
Of all the measures, “likability” was the surprise winner. “The average impression of the commercial, derived from a five-point liking scale, picked sales winners directionally 87 percent of the time and had an index of 300 (i.e., picked winners and was significant 60 percent of the time).” In other words, you like, you buy.
As the authors of the report, Russel Haley and Allen Baldinger, explain, “It appears probably that ‘liking’ is what Gordon Brown has called a ‘creative magnifier’ for both persuasive messages and for messages that are recalled.” Likable ads, they conclude, are more persuasive and memorable than others. This triggered the development of Liking Studies, as academic and practicing marketers set out to further refine the contours of liking as an ad-copy-test measure.
Liking studies further decomposed likability into cognitive and affective elements. On the cognitive side, researchers theorized that viewers who like an ad pay attention and thus recall its message better. In terms of affect, a 1994 study by David Walker and Tony M. Dubinsky in the Journal of Advertising Research explores a “theory of ‘affect transfer,’” which “asserts that if viewers experience positive feelings towards the advertising, they will associate those feelings with the advertiser or the advertised brand.” A likable ad thus promises to encode brands into our bodies as pre-cognitive desire.
Of course, affect and cognition are complex phenomena. To be fair to academic marketers, there are repeated calls in the advertising research literature to resist reducing this complexity to liking and, at the very least, to continue to use other copy-testing measures in addition to likability to predict an ad’s success. However, the underlying complexity of cognition and affect actually reinforces the value of likability as a measurable aspect of ads. The value of like is that it abstracts and condenses the complex thoughts and emotions it contains and, like any good abstraction, provides a simple and commensurable quantification of complexity. If a test subject says, “I like this ad!” it seems to stand in for the less cut-and-dried aspects of recall and emotion.
For a largely empirical, positivist field such as marketing – which has pretensions of being a science, not an art –independent variables such as likability have value because of their perceived universal predictive power. With globalization, marketing is in greater need of just such a universal measure capable of predicting the success of global branding campaigns across cultural contexts. Cultural variations might change how marketers go about getting us to like brands, but the goal is always likability.
One vertigo-inducing marketing moment illustrates this well. Since 1989, USA Today has published metrics on the most well-liked Super Bowl ads. These metrics partly inspired the makers of FedEx’s 2005 “Perfect” Super Bowl ad, which combined the top 10 likeable elements of previous Super Bowl ads. (It featured Burt Reynolds getting kicked in the nuts by a dancing bear while a cute kid and sexy cheerleaders looked on. The bear talked to Reynolds about Smokey and the Bandit after Journey’s “Don’t Stop Believing” played. Oh, and somewhere in there is a message about how effective FedEx is at delivering things.) After this ad proved quite likable, the academic authors of a 2013 study of Super Bowl ad likability used many of its elements to measure the likability of other Super Bowl ads in the 2000s. Their discovery: animals, cute kids, and humor are all elements of a likable ad. Their study also used data from, where else, USA Today‘s Ad Meter likability measures. This self-referential loop will likely sentence us to ads featuring animals, cute kids, and bad humor for the rest of our Super Bowl watching lives.

***

Or maybe not. We are now allowed, even expected, to like other things than what’s sold in Super Bowl ads. If you read even only the first two sentences of any given marketing paper today, you invariably learn that we are living in a new age where the user is in control. Social media in general, and Facebook in particular, are supposedly driving a brand-new world where marketers, editors, and other gatekeepers are marginalized and mass culture is dissolving into niche cultures and individual expression.
But if we keep the history of marketing in mind, including the development of Liking Studies, we see how Facebook is caught up in longer histories, specifically the history of the desire to dissect, study, and recompose a particular subject, the sovereign consumer.
Facebook’s Like button has been lauded as a radically democratic tool allowing users to finally make their opinions heard, but the marketing field has always regarded the sovereign consumer’s opinions the most important element in the circuit of production. After all, sovereign consumers realize the value locked away in commodities: when they buy, the corporation gets paid. In order to have us be better value-realization machines, marketers know they need to know what we like. The Like button is a logical extension of the studies and practices developed by marketing since the 1990 ARF study and the USA Today Ad Meter system which, after all, have asked us to tell them what we like for decades.
Of course, the Facebook Like button does provide increased data about which ads are likable. With the Like Button, marketers can constantly experiment with variations on ads with test audiences, seeing what works and then scaling up small experiments quickly. With a universal measure provided by Like, marketers can test ads across different segments, greatly aiding the globalization of marketing campaigns.
But this only begins to describe the continual monitoring of users who have clicked Like in Facebook. The choice to like an ad or brand in Facebook is seen as an affirmative interaction with that brand – and an agreement to have one’s profile image associated with that brand and to have that approval follow you across the Web. If you like a brand, you must like to be a target for marketing messages, both from that brand and from others similar to it. You’re providing just a bit more information to help Facebook build a profile of your tastes and desires, all of which is for sale.
But again, these things are not really all that new. Since even before the advent of Liking Studies, marketers have experimented with advertising messages and tracked users to determine ad effectiveness. Whether the branding happens on TV, in a magazine, or online, we like, we buy. We like, they know. The science of marketing has always been the science of placing us in taxonomies based on what we like.
So what is new about Facebook and the Like button? Oddly enough, it reveals too much. The great sin of Facebook is that it made “like” far too important and too obvious. Marketing is in part the practice of eliding the underlying complexity, messiness, and wastefulness of capitalist production with neat abstractions. Every ad, every customer service interaction, every display, and every package contributes to the commodity fetish, covering up the conditions of production with desire and fantasy. As such, Facebook may reveal too much of the underlying architecture of emotional capitalism. The Like button tears aside this veil to reveal the cloying, pathetic, Willy Lomanesque need of marketers to have their brands be well-liked. Keep liking, keep buying. Like us! Like us! Like us!
Liking in marketing was always meant to be a metonym for many other complex processes — persuasion, affect, cognition, recall — but it wasn’t meant to be exposed to the public as such. In Facebook, however, the “Like” button further reduces this reduction and makes it visible, making the whole process somewhat cartoonish and tiresome. The consequences can be seen in “Like us on Facebook to enter to win!” promotions and the obsession with Like counts among businesses large and small (not to mention the would-be “personal branded“).
Because likability is now so visible, so prevalent as the preferred emotional response to brands and ideas, users have predictably called for the expansion of the emotional repertoire. They call for a Dislike button. At first glance, we might think this binary-emotional expansion would be welcome to marketers: it would add to their collected data on our desires. However, marketing’s sub-field of Liking Studies has already revealed that disliked ads poison everything they touch. Negative sentiment – disliking – is asymmetrical in its power to shape consumer’s opinions of a brand: for every 10 likes, 1 dislike could tear a brand apart. Such negative emotion requires much brand damage control. One thing Facebook will never do, then, is install a Dislike button.
This is not to say that Facebook won’t introduce other binary-emotional switches. Facebook’s flirtation with a Want button indicates their potential willingness to expand our binary-emotional repertoire. One could imagine users getting a Love button. But we are not allowed to dislike. And herein lies a way out of the Like Economy. Dissent, dissensus, refusal are not easily afforded in Facebook. Dissenters have to work for it: they have to write out comments, start up a blog, seek out other dislikers. They are not lulled into slackivism or “clickivism,” replacing the work of activism with clicking “like” on a cause as if the sheer aggregate of sentiment will make someone somewhere change something.
Instead, frustrated dislikers must think through their negative affect and find ways to articulate it into networks of dislike. If dislike scares off brands, so be it. Brands aren’t going to fix the world’s problems – but the dislikers might.