At OMMA Programmatic on Tuesday RTBlog heard diverse views on how marketers, agencies, and third parties are leveraging programmatic media processes. The overarching view of programmatic is
that it has the potential to make the planning, activation, and optimization of advertising and media more effective and efficient. That’s at a 10,000-ft. view.
Therefore, programmatic
should minimize risk and maximize ROI, according to George Musi, SVP/head of analytics & insight, Optimedia. Speaking at OMMA Programmatic, Musi said the beauty of programmatic is that it
can help identify risk throughout the media planning process. But he also seemed to caution practitioners to return to the fundamentals of advertising and media. Reaching the elusive consumer along
his or her increasingly circuitous, non-linear route to decision-making is more complex than ever. The fundamentals like reach, frequency and ROI remain important. Reaching consumers at the right time
and place, with the right message matters.
While agencies have taken a beating in recent months for a lack of transparent processes, media rebates, and a host of other issues, Musi, an agency
guy, naturally believes the agency’s role is that of a “trusted advisor.” Further, he believes that role needs to be reinforced and fortified. “Agencies need to provide counsel
on market mix modeling” and other areas, especially where programmatic is concerned.
“When market-mix modeling is integrated with multitouch attribution, it will value the message,
the medium, the channel, the consumer, and context,” Musi said. And while integrating word of mouth into a digital attribution model isn’t easy, it must be done because word of mouth is
powerful, and it matters—a lot. The right analytics must also be integrated into market-mix modeling.
Further, Musi argued that a major shift has occurred as programmatic has taken hold:
the shift is to machine learning, which can help uncover insights from all the data available.
In fact, machine learning has the potential to identify a lot of knowledge very quickly. And when
integrated with media-response modeling, programmatic, and activation platforms, that's a lot of potential.
While algorithms can understand past behaviors, making quantified and
validated guesses, they still can’t account for context, which is why people are still needed. For example, the “machine” can’t spend $100 million on Facebook; you still
need people to do that. Thank goodness. The beauty of machines, however, is that they can learn from their mistakes and from outputs. Machine learning aids with the accuracy of forecasts. The goal? To
go from data to decisions.
The complete potential of programmatic offers a decrease in the time from opportunity to execution, according to Musi. “Opportunity is a finite thing,
there’s a shelf-life to it. …Ultimately, our goal is to take analytics and data to empower people through the process. We want to create insights to enlighten them.”