Marketers, Welcome To The World Of Emotional Analytics
Major media agency MediaCom announced today that Realeyes’ emotional tracking with facial detection and analytics is now an integrated part of its content testing and media planning.
Focus groups, online questionnaires and standard video recordings are so — you know — last century.
A new announcement indicates it may be time to add emotional analytics to your metrics.
Major media agency MediaCom and emotion measurement firm Realeyes said today that emotional measurement will become a regular part of the agency’s content testing and media planning. Both companies are based in London.
The Realeyes technology will now be integrated into MediaCom’s central content hub, so that producers and planners can regularly include emotion analytics in their reports. Emotion analytics are expected to impact decision-making during creative testing, media planning and post-campaign analysis.
WPP-owned MediaCom has a staff of 6,500 in about 100 countries and annual billings of more than $33 billion. Its client list includes Procter & Gamble, Shell, Sony Mobile and Coca-Cola.
“Tools such as Realeyes allow us to get behavioral information upfront, so we can optimize and measure content before launch,” MediaCom executive Palle Finderup Diederichsen said in a statement.
The announcement suggests we may have crossed the Rubicon in terms of employing emotional measurement as a standard practice. At this point, Realeyes CEO Mikhel Jaatma told me, emotional analytics is “faster and cheaper” than traditional online surveys or focus groups, and it presents direct affective responses instead of reactions inferred from, say, questions or discussions.
Realeyes employs standard webcams — usually in remote panels of users viewing ads or other material in their homes or offices — to detect facial expressions and then provide emotional analytics.
Jaatma said that his company’s technology accounts for emotional differences between, say, the reactions by a native Japanese and a native Italian to the same ad.
“The beauty of the emotional approach,” he said, “is that it’s the same set of emotions” for everyone, although “the intensity varies.” The idea is that the facial detection can spot similar facial expressions, although the analysis is tuned for the culture.
Jaatma said that question-based surveys are around 65 percent accurate in predicting long-term sales, while Realeyes has achieved as high as 75 percent accuracy for an unnamed client. In predicting social activity — such as shares, comments and views — his company claims about 80 percent accuracy.
Realeyes says its tests range from $3,500 to $7,000 and take 48 to 96 hours, while a standard pre-launch test from Nielsen/Millard Brown takes $10,000 to $15,000 and upwards of 150 hours. Jaatma said Realeyes can offer its services less expensively because “everything is done by software,” including facial expression capture and analysis, compared to human writing of questions or leading of focus groups.
“If there is more value and it’s easier to use, [emotional analytics] becomes the new standard,” he predicted.
Emotion tracking is also being offered by a variety of other companies. Jaatma pointed out that US-based Virool uses his company’s technology for its video-based ad business.
His company’s main competitor, Jaatma told me, is Affectiva, which focuses on the US market, while Realeyes is more oriented toward other countries. There are also a handful of smaller vendors, he said, such as Sticky.
The MediaCom/Realeyes announcement comes on the heels of Apple’s acquisition of Emotient, which Jaatma described as “a few years behind Affectiva and us” in terms of overall software and service capability. He predicted that Apple’s new acquisition will be focused on device interaction, instead of analysis for ads and marketing.
Now that emotional analytical systems can reliably detect emotional expressions, Jaatma pointed to the “next challenge:” better determining which emotions drive actions like making a purchase, signing up or voting.
Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.