Connecting offline data with online data raises the bar when it comes to serving and seeing relevant ads. But what about the "invisible" relationships between consumers and brands: ads that don't necessarily market or advertise the product the consumer purchases, but rather trigger a reminder or an interest?

MediaMath scientists are working to apply artificial intelligence in machine-learning models that the company's chief scientists say will power smarter digital advertising this year -- predicting intent, not just predicting a response to an ad from consumers based on their past history.

TV drives search traffic, but chief strategy officers often don't feel comfortable talking about things their respective companies cannot do. Still, I like to look ahead at the possibilities and ask what if. Like, how could Nordstrom determine whether a Ford Explorer television advertisement created more lift in search for the retailer, or led more traffic to the Nordstrom website then the automaker's?

Findings in a my6sense study suggest in-feed ads showed two times higher click-through and engagement rates than recommendation widgets, and up to 10 times higher CTRs and engagement than in-ad units.