If you can’t judge a book by its cover, can you judge a movie by its trailer?
One company says that you can. Innerscope Research claims that by looking at viewers “emotional engagement threshold” during a trailer, they can predict just how well it will do at the box office. Here’s how:
The gist here is that if a trailer doesn’t reach a certain threshold of what Innerscope calls “emotional engagement” then it probably won’t crack $10 million on opening weekend. If it exceeds another threshold, it will make more than $20 million. Fast Company explains a little bit more about how the study was done:
The study’s findings were gleaned from Innerscope’s proprietary biometrics database, “the largest of its kind in the world” including over 20,000 people, Marci says. Innerscope showed 40 discrete movie trailers to more than 1,000 respondents from 2010 to 2012. The trailers were part of “masking content” to subjects, who were reacting to other stimuli such as ads, music, videos, etc. for Innerscope clients. They were wearing “biometric belts” that captured their skin sweat, heart rate, breathing, and motion responses. Each trailer was shown about six to eight weeks prior to the film’s release. Later, Innerscope looked at publicly available data and, voila, found strong correlations between movie trailer emotional engagement and box office. (Innerscope will not release the study report publicly but has agreed to share it with Fast Company for this story.)
The story is full of buzzwords like “neuromarketing” and “neurometrics”—which really means using tools like fMRI, EEG and eye tracking to gauge a subject’s attention level and brain activity. But there’s a problem here. There’s not a lot of information available about what, exactly, is being measured: Innerscope’s “engagement threshold,” for instance, doesn’t have any units on it. And while Fast Company writes that one entertainment exec found that “the biometrics work is more sound, practical, and actionable, and does not overpromise” than other work in this field, reporter Kevin Randall reported in an earlier story about some of the drawbacks of what they call “neurocinema”:
Independent filmmaker and neurocinema pioneer Peter Katz told Fast Company about the studios’ frustration with sketchy focus groups full of viewer respondents who “don’t really know or can’t articulate or even remember how they feel about a movie or scene.” On the other hand, Morgan Spurlock’s upcoming film, The Greatest Movie Ever Sold, makes fun of studios’ growing reliance on marketing, including trailer testing via fMRI brain scans, as a way to supposedly flop-proof films and bolster the odds of a blockbuster.
Here are some similar techniques being applied to the Avatar trailer by another company called MindSign:
And here’s Wired explaining even more about how the techniques work:
Even neuroscientists argue about how to interpret brain signals from people, so the chances of being able to make something meaningful out of an fMRI readout during a movie trailer might be low. Neurocritic sums up what’s wrong with a lot of coverage of neurocinema. He points to this article from CNN first that summarizes a neuromarketing study:
For the experiment, researchers at functional MRI research facility Mindsign Neuromarketing, based in San Diego, California, scanned the brain activity of a subject while she watched two scenes of his movie. Analyzing the data from the scan, they were able to pinpoint the exact moments when her brain was lit up with fear.
. . .
During Katz’s experiment, researchers analyzed scans to identify the exact moment during each film scene that the viewer’s amygdala — the part of the brain linked to several emotions, including fear — was activated, and to what degree.
Neurocritic points out the problems here, which apply broadly to many of these kinds of studies:
Let’s see where they went wrong scientifically: (1) Calling one subject “an experiment” — was he making a movie just for her? (2) Saying fMRI can pinpoint the exact moment of anything — there’s a significant delay between initial neural firing and the peak of the hemodynamic response, which is estimated using a procedure that is not trivial for something as complex as an emotional response. (3) Using amygdala activity as a proxy for fear and thereby committing the cardinal sin of reverse inference (one cannot directly infer emotional state from the observed pattern of brain activity) — didn’t they learn from the op-ed neuroimaging fiasco in the New York Times (“This Is Your Brain on Politics“)?
But that isn’t keeping companies from shelling out the big bucks to watch our brains while we watch their work. Film producer Peter Katz puts it this way to Wired:
The same tools that are applied to making films scarier can be applied to making them funnier, or more dramatically moving. Film should be an emotionally engaging experience. This technology can be used to heighten, no matter what the intended emotional effect. The best-case scenario features packed multiplexes where individuals rarely check their text messages … they are completely enthralled … because the filmmakers have done their homework.
More from Smithsonian.com: