Monday, October 15, 2012
Creating Online Ads We Want to Watch
For millions of TV watchers, the commercial break is an annoyance of the past, thanks to the fast-forward button on their digital video recorders. Consequently, advertisers are turning to the web, where popular sites such as YouTube and Hulu force us to watch a brief commercial before playing the video we really want to see.
But assuming a captive audience on these video sites may not make online marketers more effective at reaching consumers. Harvard Business School Assistant Professor Thales S. Teixeira notes that it's just as easy for viewers to tune out the 15- or 30-second ads preceding online videos by simply turning away, opening another browser window, or chatting with someone. "It's not at all hard to avoid an ad online even if you can't technically skip it," he says.
Teixeira argues that a viewer's attention cannot be purchased by an advertiser but must be gained by the ad. Thus, he is helping advertisers to make online video ads so riveting that users want to watch them. His experimental research looks at the emotional components that both attract and hold the attention of viewers.
The stakes for advertisers are huge. Americans were exposed to over 5.3 billion online video ads in July 2011 alone, with Hulu generating the highest number of video ad impressions at more than 963 million, according to data from comScore. So Teixeira's research certainly has major implications for marketers, who spent $1.42 billion on online video advertising last year in the United States, an investment expected to jump 52 percent to $2.16 billion in 2011, according to analytics firm eMarketer.
To help advertisers up their return on these investments, Teixeira conducted a series of experiments in the HBS Computer Lab for Experimental Research. The results showed that the key to success in grabbing and holding the attention of viewers lies in evoking a carefully timed mixture of surprise and joy. (The findings will be published in an upcoming article in the Journal of Marketing Research, "Emotion-Induced Engagement in Internet Video Ads," coauthored by Teixeira and fellow researchers Michel Wedel of the University of Maryland and Rik Pieters of Tilburg University.)
The researchers paid 58 adults to watch a four-minute sitcom clip, followed by a series of 30-second consumer product advertisements. Out of 28 ads, 14 were chosen because they were decidedly provocative; the researchers expected them to evoke either joy or surprise in the viewers at different points in the brief plot. The other 14 ads were meant to be emotionally neutral; these were played between other ads as emotional buffers—sort of palate cleansers for the soul. Participants could watch an ad until the end or skip to the next one by pressing the space bar at any time.
In each case, a camera recorded participants' facial expressions as they watched the ads, while an infrared eye tracker unobtrusively measured their eye movements. The researchers parsed the camera data with the help of computer vision software that evaluated emotional response based on variations in facial features.
Meanwhile, the eye tracker detected where participants focused on a particular image in the ad. (In this this video clip illustrating a participant's eye movements, the large dots indicate actual focus.) "It so happens that there are two basic types of eye movements: one is fixation, and the other is where the eye is moving from one position to the next," Teixeira explains. "We don't notice it, but we only have visual information when the eye is fairly fixed."
Combining the expression data with the eye-tracking data yielded important results. "We found that people's attention patterns on-screen were different depending on the emotion they were feeling," Teixeira says. "So during the ad, your eyes move differently on the screen, depending on whether you're surprised or you're joyful."
As expected, the researchers found that the more attentive viewers were, the less likely they were to skip to the next ad. More importantly, they found definite patterns in the emotional elements that commanded the most attention. In short, evoking surprise proved the most effective way of capturing attention, while evoking joy was best for retaining attention.
"The findings showed that advertisers should use a quick element of surprise at the beginning of an ad, followed by a longer period of joy, in order to get the most 'attention' bang for the buck," Teixeira says.
This finding flies in the face of focus groups of the past, which found that viewers responded most positively to ads that had a big surprise or punch line at the end, according to Teixeira. But in the focus groups, viewers did not have the option of skipping the ad the way they do in real life. A surprise at the end of the ad has no effect if the viewer doesn't bother to wait for the end of the ad.
Video-ad designers also have to extend the opening surprise long enough to provide maximum impact. That's no easy feat. The data indicated that it's hard to evoke a feeling of surprise for more than just a second, even for companies known for their marketing genius.
The experiment showed that Budweiser was the most successful of the advertisers, with many participants exhibiting more than a second of surprise while viewing a Bud Light ad. In the commercial, office workers have to put a quarter in a communal "swear jar" every time they utter an obscenity. Rather than curb their cursing, they start swearing a blue streak after learning the pooled money will go toward a case of Bud Light. Thus the ad features a series of bleeped-out curse words. It's surprising at first, and then it's just funny—evoking small moments of giggly joy, culminating in the big moment when everyone is rewarded with a beer bash.
Teixeira notes that Budweiser has garnered similar surprise-inducing success by featuring barely censored nudity in a recent web-only commercial. "No matter what you're doing, you'll stop when you hear profanity, even bleeped-out profanity—or when you see almost-naked people running around an office," he says. "These are just a few among many ways to garner attention quickly in the first few seconds of an ad."
After getting the surprise element right, the ads must deliver on joy. The researchers found that ads retained attention better if they delivered several snippets of joy in succession than if they delivered a sustained period of joy. "It's like feeding them little bits of chocolate, one at a time, as opposed to handing them a whole chocolate bar," Teixeira says. "People adapt to almost anything, and adaptation is a constant in human nature, even when it comes to something good. So you want to alternate [the emotion] to really get them hooked."
Of course, capturing attention is only part of the advertising battle. Converting attention to sales matters, too. Teixeira now is looking to discover the factors that make the former lead to the latter.
"Attention is the biggest bottleneck, and we need to get through it," he says. "But after you've secured attention through good ads, the next questions are: How can I persuade? How can I communicate? How can I get people to change their evaluations of brands and products rather than simply be entertained for 30 seconds?"
In collaboration with colleagues at MIT, Teixeira is conducting experiments in which participants view online ads in their own homes, monitoring them with webcams. ("With their permission, of course," he says.) The idea is to gauge how consumers react to advertisements in a natural setting.
"There are two competing hypotheses," he says. "One is simply that if you get people's attention, you're likely to have some opportunity to persuade. The other hypothesis is that through emotions, people reduce their guards against persuasion. The advertiser can work to put people in a better mood, which makes them more open to suggestion--and that could lead to an increase in persuasion."