What: Portada talked with Acuity Ads Chief Strategy Officer Seraj Bharwani about how machine learning is expanding brands’ reach to specific cultural audiences.
Why it matters: Machine learning is paving the way to a new, more effective approach for reaching specific cultural audiences and will be a featured topic at Portada Los Angeles on March 15. But digital platforms, like Google and YouTube need to do more to protect brands from inappropriate content, Bharwani says.
Seraj Bharwani has his eye on decision sciences, specifically: machine learning and computer algorithms which he says give brands powerful new tools for penetrating deeper into multi-cultural market segments than ever before.
It used to be that speaking the language of a specific ethnic market segment was all advertisers could do to put their messages before those audiences. But new technology has changed the rules of the game, Bharwani tells Portada.
“Computer algorithms that can process vast amounts of user data to assess user preferences and propensities by aggregating their social, search, shopping and viewing behavior in real-time are now being used to create what we call live personas of in-culture audience.”
Portada: Why are traditional in-language targeting tools insufficient to reach cultural audiences today?
Seraj Bharwani: There appears general consensus among sociologists, anthropologists, agency planners, and the like that defining the US Hispanic population by language is simply too limiting. This audience lives a rich life defined by their unique core family values, cuisine, music, festivals, celebrations and other characteristics more indicative of their preferences.
Another unique aspect of this audience is their rapid, scaled adoption and prolific use of digital and social media. These online behavior signals in aggregate provide robust cultural cues that allow us to reach the other two-thirds (40 million) of the US Hispanic population that is bi-cultural and bilingual with more relevant advertising.
Acuity Ads Chief Strategy Officer Seraj Bharwani will be a featured speaker atPortada Los Angeles on March 15 when he will provide insights in how machine learning tools are expanding brand marketers’ ability to reach multi-cultural audiences.
Portada: What role do decisioning technologies, i.e. Machine learning or Artificial Intelligence play in updated efforts to target specific cultural market segments in the US?
S.B.: Computer algorithms these days can process vast amounts of user data to assess user preferences by aggregating their social, search, shopping and viewing behavior in real-time, and are now being used to create what we call live-personas of in-culture audience.
By incorporating what keywords people type, who they follow, which videos and content they watch, which mobile apps they download and check into, etc., the “learning” algorithms can predict (in real-time) if a given PERSONA is sensitive to the cultural context and will respond to specific ads, products, or brands with a culturally relevant message.
Computer algorithms that can process vast amounts of user data to assess user preferences and propensities by aggregating their social, search, shopping and viewing behavior in real-time are now being used to create what we call live personas of in-culture audience.
Portada: Have the sources of consumer data changed as marketers have begun to use machine learning tools to reach specific cultural targets?
S.B.: Publishers like Facebook, Google and Amazon possess substantial consumer data individually within their respective lanes – social, search or shopping data respectively that could provide selected “In-culture” cues. What advertisers need are in-culture personas aggregated from a full spectrum of consumer behaviors across social, search, viewing, and shopping behaviors.
Portada: Several major brands recently pulled their advertising from YouTube after it was revealed that viewers were exchanging child pornography links in the video comment sections. How can marketers better protect their brands?
S.B.: This is not the first time we have heard of the brand safety scandal on YouTube. Advertisers need to be clear about why they are buying advertising on YouTube/Google. If an advertiser wants great audience REACH, then the environment within which the ads run may not always be brand appropriate or brand safe. As sophisticated as Google happens to be with its pattern recognition algorithms, it is unrealistic to expect the platform to police every video and comment posted on the platform. And Google isn’t about to shut down the commenting altogether like the PBS News Hour as it would severely curtail the engagement potential of the YouTube platform.
If on the other hand advertisers want fully brand-safe and brand-appropriate environment for their ads, their best bet would be to buy access to the YouTube audience through premium publishers like Disney, Vevo, Buzzfeed, and others who have captive channels on YouTube to ensure control over the quality of the media environment. There is a good chance that advertisers would sacrifice media efficiency (premium environments command higher prices) and REACH.