--
To demonstrate the power of tweets, Twitter’s ad researchers turned to neuroscience. Here’s what happened.
Twitter’s senior director of market research, Jeffrey Graham is always looking for ways to show the effectiveness of ad campaigns on Twitter — surveys, home visits, data models.
One of the more interesting studies involved two groups of people watching the NCAA basketball tournament on television. One group was permitted to bring their phones and tweet all they wanted. The other had to leave their phones outside and somehow manage without a second screen. Both groups had sweat monitors on their wrists and foreheads, a pulse rate monitor, and eye tracking goggles, to track how engaged they were. In comparison with the no-device crowd, the metrics went wild for the group permitted to tweet. “For people able to do Twitter and TV at the same time, there was a huge lift versus people who were just watching TV,” says Twitter’s global president of revenue and partnerships, Adam Bain.
But Graham felt that Twitter could really make a mark using a technology he learned about in an advertising research association’s report. It described how using neuroscience could get you other unavailable data, stuff from the subconscious reaches of people’s minds.
There was one obstacle. Using the standard method of brain imaging, fMRI (functional Magnetic Resonance Imaging), would be cumbersome and expensive, never mind that tweeting while surrounded by a giant magnet could be hard to pull off. But Graham then found a company in Australia called Neuro-Insight that promised to measure brain activity using a much more manageable technology they call Steady-State Topography (SST).
“People wear a headset,” he says. “It’s sort of goofy looking but isn’t that hard to wear. The respondents wear that, and a visor that beams a rhythmic flickering light. It measures the processing speed of different parts of the brain.”
So Graham’s team arranged a study at Twitter’s UK headquarters. One hundred and fourteen people participated, in groups of around twenty. Videos of the sessions show people putting on the helmets, which look like a cross between a Snoopy-style Red Baron helmet and a polka dot shower cap destined for Katy Perry’s cranium in a music video.
Then, during 45-minute sessions, they alternated between normal Web-surfing activities and using Twitter — reading their timelines, tweeting, and other birdy stuff.
Graham had hoped that the brain profiles of people using Twitter would show the difference between his employer and more static Web use. “When I go on Twitter, oftentimes I really get sucked into it,” he says. “I get this strong anticipation to see what engagement is going to be.” But he admits that he had no idea what the data would actually show.
The results were more than he’d dreamed of. The study first tried to measure a neural signature that tends to correlate with information relating to you—a “sense of personal relevance.” It did this by comparing how participants’ brains activated when either passively scrolling and browsing on Twitter, actively tweeting and retweeting, or engaging in normal online activity. The brain data suggested that passive Twitter use increased a sense of personal relevance by 27 percent. Active use boosted that number to 51 percent. The representative from NeuroInsight told Twitter that in all the testing the research company has done, there’s been only one result as high: when people opened personal mail. (The physical kind.)
The most dramatic results reflected emotional intensity. Reading a Twitter timeline generates 64 percent more activity in the parts of the brain known to be active in emotion than normal Web use. Tweeting and retweeting boosts that to 75 percent more than a run-the-mill website.
The final measurement had to do with memory. Passive Twitter use indicated 34 percent more activity in areas linked with memory formation than normal online use. With active Twitter use, the number rose to 56 percent.
There are endless caveats about what conclusions can be drawn from neuromarketing in general and SST specifically. It’s a big leap to go from measuring electrical activity on the surface of a given part of the brain to postulating that a tweet about a Big Mac might lead you to the Golden Arches. Do people really remember tweets more than what they see on a website? And to the degree that brain technology does reflect thought, the practice evokes unease—after all, it explicitly tries to get to our subconscious, a realm people are not often happy to share.
Graham shrugs off the potential creepiness of analyzing people’s brains to see how they responded to his company’s product. “People were smiling — it’s like a video game, or a 3D headset,” he says. “It’s not scary — we’re not testing them, we’re testing Twitter.”
