We find ourselves in a dizzying world where advertising can be bespoke, targeted and personalized down to an amazingly specific degree. Consider, after all, the futuristic scenario where an ad for tortilla chips gets sent only to left-handed, glasses-wearing bike riders in Pennsylvania. And then whittled down a little more so that it’s only sent to those viewers who have never before eaten a tortilla chip (and who may now suddenly be feeling a little snackish).
What are the benefits and downsides of creating that kind of targeted advertising? And what role are machine learning and artificial intelligence playing to improve advertising and promo insertion to customers?
At the Tuesday afternoon session “AI in Media and Entertainment: During the Broadcast,” a panel of media executives looked at the ways that these technologies are improving ad insertion.
Right from the start, the panelists addressed head-on the mystique that sometimes surrounds AI. “I think there’s a great deal of education that needs to happen about AI, especially when you have Bill Gates saying AI is the most terrifying thing [and will] ruin the world,” said Geoff Wolinetz, senior vice president of client relationships with FreeWheel, a Comcast Co.
Instead, he said, “The benefit of AI has everything to do with the application and nothing to do with the concept.”
The panelists touched on the ways that AI and machine learning can give media companies useful insights, including a deeper understanding of forecasting and feedback on the best ways to personalize an ad experience.
But for all the promise of an automated AI world, an actual human being is always going to be part of the scenario. “When you go into machine learning and you start applying datasets, you’re going to find out that when you develop an algorithm and apply it to a dataset, you’re going to be wrong,” joked Christopher Curley, partner lead of video distribution and telecommunications at Google. Human intervention is needed to create clean datasets, which can then be used to guide companies on ad and promo insertion, he said. “You need a human to go back and correct those datasets. That’s not changing any time soon.”
The true benefit of AI, said Wolinetz, is that it gives humans the opportunity to focus on more strategic thinking. “There are things that we can optimize [with AI] so that we can turn human attention to more strategic things,” he said. “We need … humans teaching the machines to get better and smarter.”
Jarred Wilichinsky, vice president of video monetization and operations at CBS Interactive, agreed. “It’s a great tool for any company department manager to use to automate tasks … so you can do more strategic thinking.”
Media companies also need to be aware of the tricky balance between using data to optimize advertising and the demand for consumer privacy.
In January the American Data Dissemination Act was introduced in the Senate in an effort to create a national consumer data privacy law that would impact the collecting of data via the internet that contains personally identifiable information. That bill followed others related to data protection and transparency. Despite the challenges that additional regulations might entail, the panelists
reiterated the importance of smart data governance and privacy protections. “The challenge exists to protect consumers and extract data,” Curley said. “We very much respect the relationship we have with our consumers that use our products.”
And what role are AI and machine learning playing in today’s biggest advertising buzz topic, advanced advertising?
“It’s where we are today,” Wilichinsky said, adding that savvy media companies still need to ask whether this combination of advanced advertising with AI makes sense from a monetization standpoint. The answer ultimately lies in whether your advertising efforts are reaching the right customers in the right way.
The panel was moderated by Chris Pizzurro, vice president of global sales for Canoe Ventures, and included Joel Hassell, CEO of Canoe Ventures.