How AI and Machine Learning Are Impacting Media

New technology and how it is transforming traditional media was the main focus at the recent Streaming Media East conference. There’s no doubt AI and machine learning are impacting media, and according to a panel of experts, it will be a boon to both consumers and the media business.

Nadine Krefetz, a consultant with Reality Software, noted that, “Machine learning and AI are more buzzy than ever.” The reasons why media should be focusing on these protocols include the need to harness huge amounts of data, stitch together all of the complex parts inherent in the data sources and automate the cost efficiencies and time.

According to Gabriella Mirabelli, EVP Consumer Insights and Brand Strategy, Valance Media, AI is the technique that creates human learning using logic. She noted that Machine Learning is a subset of AI that uses, “complex statistical techniques to improve tasks that the technology can do.” Another subset of AI is Deep Learning that “uses algorithms to help the machine train itself.” In sum, AI uses human intelligence while machine learning uses data to predict outcomes.

What AI Can Do
When applied to media, these systems enable smooth processing and understanding of the mountains of data now available to marketers. For Mirabelli, we are advancing into, “an Era of Insight from the Era of Data,” where there is greater complexity and too much data for humans alone to handle. She uses these systems to make buy decisions and to answer unique questions. For Field Garthwaite, Co-Founder and CEO, Iris.tv, these systems can be used to improve the customer experience and, for Thomas Ohanian, Global Sales Executive, IBM Watson, the underlying technology that these solutions depend upon enables speech to text, object recognition, voice and other interactive elements by “reducing human tasks and gleaning insights from content.”

What AI is Doing Now
Sports are a ripe area for AI. Ohanian explained that fewer people are going to events at the stadium, preferring to watch on the glass. AI offers real time insights by using algorithms to assess “intent, reaction and gestures to build interesting and timely sports highlights.” It is an instantaneous way to “pick plays, players, matches and play types through automatic curation which is very sticky,” to reduce the bounce rate and help editors find relevant content quickly.

Nikki Conley, Cloud Solution Architect, Microsoft, offered another perspective. Mistakes can be corrected or monetized. A recent Game of Thrones episode erroneously had a Starbucks cup in the frame. “If we ran through an analysis and saw the cup, we could remove it … or ask Starbucks for advertising support,” instead of inadvertently giving them $2.3 billion in free advertising. AI enables the ability to “look for occurrences in content in post-production with extensive tagging.”
From amassing data to curating content to correcting mistakes to providing an overall better experience for viewers and more profitable opportunities for advertisers, the future is now for AI in media.

This article first appeared in Cynopsis.

Digital Placed Based Media Makes Its Big Move

As if the current television environment isn't challenging enough with flat expected revenue, expanding linear viewing choices, audience fragmentation, cross platform viewing, VOD, DVR, and commercial ratings, there is apparently a new set of challenges eroding potential ad dollars according to the DPAA at its recent conference.

This challenge to media ad dollars may be coming from the most unlikely places - a taxi, an elevator or even a gas pump.  In this ever expanding and developing area of digital placed based media, any place where the consumer journey takes place is an opportunity to reach them via a screen. And digital placed based media is already recognized by brands and advertisers as valuable opportunities to reach their target consumer.  As Jack Haberman at Colgate Palmolive says, "Video is the new day part". That means that viewers are no longer tethered to the home screen… or any one based device... or by any time of day. Prime time can be any time and any place; morning at the gym, lunchtime at your desk, commutation in a cab or train.

Consumers Rule

The proliferation of screens into all aspects of everyday life, both inside and outside the home, presents both opportunities and challenges for media companies, advertisers, brands and data companies. According to Francois De Gaspe Beaubien, chairman of the DPAA, "Consumers are spending most of their time out of the home. People are interacting with screens outside of their homes." That means that there is more competition for attention and more of a consumer driven environment for placing content. And with prescient content owners like HBO deciding to go OTT, it looks more and more like the content delivery chain, like everything else in media today, is fragmenting and morphing. More native advertising, more binging, more multi-tasking is coupled with less reliable measurement, less concentrated viewer attention and less control of the chain by the old television guard.

Who is to Blame?

There is certainly enough blame to go around. Technology certainly plays a large role in this disruption. Jack Myers gave an inspiring talk about how technology is forever changing media habits. He says, “The fundamental culture of our business is changing. Every minute of every day, consumers’ media habits are radically different than they were last year, two years ago, ten years ago and certainly dramatically different than they were 20 years ago.”

But in addition to technology, traditional media’s way of doing business also contributes to the viewers’ changing habits. Ad time within a TV program can be more than 25% of the program run time. Commercial pods are getting longer, encouraging viewers to find ways to avoid ads. BTIG’s Rich Greenfield noted that “Binge viewing drives more binge viewing. We are learning to watch in a certain way. We are training consumers to avoid ads.”

Challenges – Measurement, Privacy and Fraud

However, digital placed media has a few challenges as well. For one, there is no standard accredited measurement across all OOH platforms. The current Nielsen measurement consists of audience studies that are conducted as proprietary research with each study funded by the measured network. Nielsen then takes these studies across all measured nets and compiles into one report that is issued quarterly to the marketplace. According to MediaVest’s David Shiffman “A few years back, the DPAA issued specific guidelines for their members to follow for audience measurement with the goal to help ensure some standardization in what gets measured and how it is measured.”

There is also the privacy concern. Deutsch’s Anush Prabhu noted that “there is a lot more data available today. We can tell your geo location from your phone which is cool but also creepy.” Microsoft’s Natasha Hritzuk added, “Consumers are aware that their data is being collected. It is the lack of transparency as to how we are using the data that is the problem. We need simple conditions that we can opt into and a clear value exchange.”

Fraud is another concern. As Barry Frey, President and CEO of the DPAA remarked, “Kraft has said that 75-85% of their data is fraudulent and they will not use it.” Companies are responding to this challenge. According to MEC’s Shenan Reed, “We are going to 100% viewable which is higher than the industry standard. We will only pay for something that is 100% viewable.”

Ultimately, the DPAA conference was a high note for digital placed based media. As Frey concluded, "We received terrific feedback on this year's Video Everywhere Summit from many of the record 700+ people in attendance. The sessions covered not only issues specific to the rapidly growing digital place based media sector, but those of importance to the overall media and advertising ecosystem as well. It's always rewarding and exhilarating when so many months of planning coalesce into a worthwhile day for so many people." See a video of some of the highlights of the DPAA conference here. (insert link or embed)