Cue deep, booming movie-trailer voice: In a world where some people go to a movie because of great reviews and some because they’ve got nothing better to do, how can marketers tell if a particular audience member is actually enjoying a particular flick?

Truth is, they cannot. Neuroscientists have yet to identify the patterns of neural activity that signal whether we are watching something interesting. There is no way to tell by looking at a moviegoer’s brain whether, for example, that person finds Hugh Jackman’s Wolverine boring or bewitching.

But monitor the brains of the entire audience, and clear patterns emerge. Suddenly, you can tell whether people are spellbound by the X-Men’s antics, or would just as soon be having a snooze.

In other words, by comparing activity in the brains of multiple moviegoers, “you can actually predict the success of a movie,” says Kellogg’s Moran Cerf. “It is the ultimate focus group.”

Why does this happen? And how might marketers take advantage of it?

That is what Cerf, a neuroscientist and associate professor of marketing at Kellogg, explored in new research with PhD student Samuel B. Barnett. Their study focused on how engaged moviegoers are with trailers. 

“A lot of money and effort go into trailers. ... There has been a lot of intuition, but no data. We finally have it.”

And the stakes are high. Film studios spend more than $3 billion annually on movie marketing. Small wonder that studios and theater executives negotiate fiercely over which trailers accompany movies and in what order.

“A lot of money and effort go into trailers,” Cerf points out. “They are tricky things to make. You want to show some explosions, some kisses, but not tell too much of the story, and they have to be short and exciting. For the last 50 years, studios had very little data on what works. There has been a lot of intuition, but no data. We finally have it.”

Neuromarketing Research in the Movie Theater

An encounter in 2013 with the CMO of AMC Theatres at the Kellogg School’s Marketing Leadership Summit paved the way for Cerf and Barnett to conduct their study in an ideal real-world setting: a cinema.

“Most scientists, they bring their subjects to a lab where everything is controlled—the light is perfect, the distance between the chair and the screen is fixed,” Cerf says. “You can always argue that a study made in a lab only works in a lab. Whereas a movie theater is as natural as possible.”

AMC agreed to let Cerf and Barnett partner with a movie theatre in Northbrook, IL. For one month in 2014, people who came to the theatre were offered free admission to any of the movies playing there in exchange for taking part in the study.

Of the 122 people who participated in the study and watched their preferred movie, 58 people also got there early enough to watch all of the trailers that preceded it while wearing an EEG cap. With a chinstrap and many protruding black cables, each cap looked a bit like something Medusa would wear to swim. The caps allowed the researchers to detect electrical activity in the regions of the brain associated with attention to visual stimuli.

The researchers also wanted to learn whether there were simpler tests they could do to get information about which trailers people find interesting.

So participants also had their eyes tracked using a special high-resolution camera that can record in the dark. Additionally, they had their motions, head movements, and gestures tracked; their heart rate, respiration, hormones, and skin conductance measured; and their facial expressions analyzed. A subset of participants also had their genetics tested via a 23andMe kit. (The team is analyzing all this data in a separate study.)

After viewing the entire movie, all 122 participants were asked to recall the title and plot of each of the trailers they had seen. They also rated their enjoyment of each trailer, and whether they planned to see the movies that the trailers promoted. Because participants were seeing movies they chose themselves, the trailers were presumably the theater’s best guess at what upcoming films would most interest customers—and bring them back to the theater again.

Six months later, all participants were sent a surprise follow-up survey that again evaluated their ability to recall each of the trailers.

The participants also received a free ticket to a movie of their choices from the ones they had seen in the trailers. This offered researchers a direct way to test customers’ “willingness to pay” for any of the movies—an important measure in marketing research.

Ultimately, the researchers analyzed the electrical activity of the cap-wearing moviegoers and calculated the level of synchrony between them. This synchrony is called Cross-Brain Correlation, or CBC.

As it turned out, the higher the average CBC for each movie trailer, the more likely all participants—whether they wore the caps or not—were to recall that trailer right after the film as well as six months later. A higher CBC also increased how likely participants were to report being willing to pay to see the associated movie, as well as how likely they were to actually use their free ticket to watch it. Given that only about half the participants were wearing the special caps, this means that the synchrony of a smaller group can be extrapolated to make assumptions about the audience as a whole.

Why does CBC correlate with enhanced recall and enhanced willingness to pay? The answer comes down to a variation on the maxim from Anna Karenina: all happy moviegoers are alike, but bored ones are bored in their own way.

“The idea is that content that is very engaging is going to make all brains look alike,” Cerf explains. “Whereas content that is boring makes our brains go in different directions. When we are bored, you might start thinking about your kid, while I might think about my class. The stimulus is not there to drive our brains in a similar fashion.”

What Were Moviegoers Reacting To?

Cerf and Barnett wanted to make sure that a higher CBC really did indicate enhanced engagement, rather than something else.

Perhaps, they thought, CBC was being driven by collective confusion, rather than collective interest. That is, perhaps the moviegoers’ brains looked similar not because they were all interested in the same stimulus, but because they were all confused by the same stimulus.

To find out, the researchers also analyzed the movie trailers’ visual and semantic complexity in a separate study. They determined that trailers with fewer prominent objects to focus on—as well as fewer words, sentences, questions, unique speakers, and overall speaking time—had a higher CBC, making it unlikely that viewers were confused by the content.

Neural Activity Predicts Box Office Success

All of which is well and good, but how did the results track where it matters most: the box office?

Quite well, it turns out. Not only did a higher CBC correlate with greater viewer recall and willingness to pay, but “it’s nicely correlated to box-office sales as well,” Cerf says. “You can use it to predict general-population behavior. We picked a random set of 122 people to test. Nothing is unique about them. So it could have been any large group of people, making it possible to generalize to a large population.

Average CBC for each trailer, the researchers found, was a strong predictor of the associated film’s average weekly ticket sales, opening-weekend revenue, and total lifetime revenue (which are all correlated to one another. For example, the film whose trailer garnered the highest CBC, X-Men: Days of Future Past, made $111 million on its opening weekend, while the film with the lowest-CBC trailer, Mr. Peabody and Sherman, made just $32 million on its opening weekend. Looked at another way, average weekly ticket sales were more than four times as high when CBC was in the top quartile relative to when CBC was in the bottom quartile.

Frank Elavsky

To no one’s surprise, Cerf and Barnett have seen a great deal of interest from Hollywood and others in their CBC-measurement method. 

“Since the paper came out, we got emails from Paramount, Viacom, and a number of companies that make trailers,” Cerf says. “We now want to do the same thing for TV commercials, following a number of studies we ran during the Super Bowl in the last 3 years. We also got a call from Las Vegas asking if we can apply the same tests to the audience of live performances, theaters, circuses, etc.”

Cerf has several similar studies underway to determine how measuring CBC can be useful in other milieus, such as music, politics, and sports. “We can see, for example, who finds a political candidate engaging and who does not,” he says—something he discovered while monitoring and comparing the neural activity of people watching the 2016 presidential primary debates.

In the meantime, he emphasizes that the findings apply to more than marketing.

“What we are looking at is not just a way to measure engagement with content, but a way to understand how some people are able to tap into the collective consciousness,” he says. “Think about Picasso shutting himself in his studio, painting something on a canvas, and somehow capturing with a brushstroke the brains of hundreds, thousands, millions of people. That is what true artistry is—to be able to access the brains of your audience in your mind.”