• by Ben Kartzman , Op-Ed Contributor, May 10, 2017

Artificial intelligence  (AI), or machine learning,  is a hot topic and has become an increasingly vital component of data-driven marketing.  

Essentially, machine learning is the programmable ability employing software algorithms to gather data points at unprecedented scale, which then inform insights and market intelligence.  

When real-time bidding (RTB) and open exchanges came into vogue, machine learning was in its infancy. The foundation of RTB was driven largely by third-party data from data-management platforms (DMPs), on which information was often imprecise and outdated.  Even when accurate, the data painted a far from comprehensive picture of a particular consumer’s behavior, desires and habits.  

We remember all too well the annoying cyber-stalking that plagued the early days of retargeting, when a consumer was inundated with vacation ads after having just booked a getaway online.  

But with the emergence of brand marketer first-party data and improved technology, the consumer feedback loop is constantly improving, and consumer targeting has become more precise, leading to greater levels of engagement.  And with the rise of dynamic creative technology, marketers can now change creative elements in digital ads on the fly.  The age of truly personal marketing has arrived!     

Marketers are now delivering the most beautiful, relevant and engaging brand content to prospects and customers across all devices and formats.  AI-driven platforms now analyze thousands of data points: audience segments, weather conditions and media placements for every impression.  Decision engines apply structured logic to data inputs to deliver the most effective creative messages. Auto-optimization then employs machine learning to optimize those rules.  This level of sophistication continues to be beyond the capacity of humans.

Just ask the Tennessee Department of Tourism Development, which crafted over 2,000 personalized vacation video spots tailored to consumers’ online behaviors, likes and interests.  In only two weeks after the launch of this campaign, TDTD saw a 46% increase in Web traffic, and 12,000 trip itineraries were created.

Beyond creating more compelling media messaging and better offers through optimizing creative, machine learning can also have significant impact beyond paid media. Insights can even inspire and inform new product development. For example, marketers may develop specialized product offerings based on the discovery of new audience segments.

In 2017, look for such AI-driven advancements as  new “ambient” reporting to measure performance in an even greater combination of data points including time, location and weather in addition to traditional first- and third-party DMP segments.  

Another innovative tool in development is  content tags, which will scan and analyze the creative content within ads and identify every individual visual element.  By cataloguing all of their ads' visual components, marketers can more easily and efficiently spot trends among the best performing ads and improve the auto-optimization by reconfiguring individual ads. For instance, you may notice the top-performing creatives always feature red cars as opposed to black, or that "no APR" ads don't perform well.  

 As the era of the Internet of Things blossoms, AI will play an even more central role in everyday lives.   Certainly, brands can take advantage of this construct within the realm of programmatic digital advertising now and going forward.  The benefits of AI for brand marketers are becoming real, clear and powerful.