AI-Derived Insight Sharpens Competitive Edge for MVPDs, NextGen TV Broadcasters

(Image credit: iStock)

Artificial intelligence touches far more than the production and workflow sides of the television industry. It is making its presence known at the cash register, as well. 

Guided by insights derived from data pools far larger than any individual person or office full of people could unearth, MVPDs like cable operators have the ability to improve how they shape their channel lineups, acquire and retain subscribers and even exploit opportunities growing out of their own OTT-like services, says Mark Moeder, CEO of Symphony MediaAI.

Symphony MediaAI offers a variety of services to media companies, including auditing licensing, retrans and subscriber fees as well as royalty payments. It also is leveraging artificial intelligence to help cable operators and other media companies compete more effectively for viewers.

I interviewed Moeder to learn more about what the company’s AI can do for MVPDs and whether or not the same tools can help broadcasters as they deploy NextGen TV, which gives them the ability to address individual audience members with content and advertising based on their personal interests.

(An edited transcript)

TVTechnology: When TVTechnology magazine covers AI, it’s frequently from the perspective of production and workflows. For instance, a look at speech-to-text algorithms for closed captioning or other algorithms that can enhance content metadata gathering, such as object or facial recognition. How is Symphony MediaAI applying artificial intelligence to the business of the business? 

Mark Moeder

Mark Moeder (Image credit: Symphony MediaAI)

Mark Moeder: The reality is the consumer is what matters. By better understanding subscriber behavior, you can better understand what matters to the individual. So, as an example, I watch football—predominantly the Kansas City Chiefs.

I watch them religiously every week. I also enjoy comedies. I enjoy a myriad of different things that kind of define my interests. The proliferation of content out there means that my interest is spread across perhaps 500 or 600 different titles that I access when I want for however long I want. 

The dataset on just me is incredibly deep and really hard to get to the heart of now. If you multiply that times 5 million, 10 million or 20 million subscribers, the depth of that data becomes really impenetrable for any human. Sure, you can get surface level inside. So yes, you know 12 million people hop on and watch the Chiefs’ game. But that doesn't really tell the whole story, right?

It doesn't give you an indication that I like sports but that I have other interests and likes. What content can an MVPD bring to me and in what method to really boost loyalty and retention? The data sets on the subscribers can be into the hundreds of thousands of dimensions per individual. AI goes through that data, and our approach is to use it to create a kind of a topological approach. It gives shape to the data.

TVT: Can you explain how the data is shaped?

MM: It starts to cluster users based on their interests, based on their engagement, duration, consumption time, all of these different factors, and starts to create a shape of your subscriber universe. Shape has meaning.

You can glean a lot from the different shapes. If I look at my subscribers, I can see several hotspots, and then I can drill in on that and understand what makes up that hotspot. What are the right conditions that tie this one to that one? Why do I have outliers over here? What are they doing? Were they interested or disinterested? 

Then you start to draw some comparisons and prediction models to understand how to bring these outliers into kind of the larger conglomerates by exposing them to better content or new content that they hadn't seen before, or frankly, by providing them content in a way that they are more apt to respond to—in other words, through digital connections, partnerships with other entities.

Our AI engine goes through the minutiae and raises it up to a consumable method. At the same time, it does predictions to say, you know, these groups of people are similar to those groups of people. 

TVT: How does that translate into a business advantage for an MVPD?

MM: This is especially important to media companies, whether they are MVPDs or traditional media companies, as they try to understand and decide on what content to create, to acquire and to distribute to the market.

All of those are very costly, and frankly, the algorithms and the math used to kind of predetermine the ROI is more art than science in a lot of scenarios. So, tools like AI can take everything that I talked about before to really understand the consumer and determine how they are likely to respond to that content, its long tail value.

For example, “The Office” is still absolutely one of the most watched series out there. I think the tools could show not only do you get a near-term ROI, but based on the characteristics of a piece of content, it actually has a really long shelf life that will attract new consumers and retain existing consumers for years to come. Those types of insights really help you identify the right way to spend your money for your business. 

The other facet is protecting your core subscribers. To do that, you need to make sure the content you're bringing in really speaks to the audience you have.

PLUS: Deep Learning in the Media Supply Chain

TVT: Do most MVPDs have the technical ability to take action based on the data that’s unearthed by your AI? In other words, it seems like to move on this data they need the ability to address individual viewers with content tailored to their personal preferences.

MM: Many do in varying ways. This could inform the decision partly on channel lineups. It could inform their decisions on platform marketing as they market themselves. It could on what content is going to resonate through for acquisition of subscribers as well as retention. 

Then, there is opportunity here for the MVPD to—really not reinvent, but—re-evolve itself into more of a digital offering. That's where the consumer is going. Arguably, that's where the consumer already is.

Absolutely, many do not have the technical infrastructure to do this today, but many are and should continue to look at it and accelerate that investment. That's where you're going to reach the new consumer.

Once you get into that side of the equation, then absolutely you can do individual marketing, individual engagements, everything that we’ve talked about.

TVT: When it comes to offering personalized content and addressable ads, there’s a new player on the block—linear TV stations that are, or soon will be, leveraging these capabilities within an ATSC 3.0 world. Are you also working with broadcasters to make the most of these opportunities by using your AI?

MM: Yes, absolutely. We historically have helped broadcasters but that was from a subscriber management understanding through the general pathway of MVPDs and third-party data. 

With ATSC 3.0, they now have the ability to go direct, and there’s a variety of other things that extend off that. That makes it kind of first-party data for broadcasters. So our subscriber management tools can and do support broadcasters.

We’re having a number of conversations with broadcasters about how they can now really understand who they are distributing to over the air via ATSC 3.0 and creating the same data lake that the MVPDs and others have.

This is brand new to this segment of the industry. They know they should be capturing data to learn how to use those dimensions we’ve discussed, and to some degree perhaps for targeted advertising and to some degree promotional targeting, which will accelerate and bolster their retention strategy.

Obviously, the broadcast industry is at the beginning stages of ATSC 3.0, but we’re helping our customers understand what kind of benefits they could achieve using tools like ours to build deep knowledge around direct consumer relationships.

More information is available on the company’s website (opens in new tab).

Phil Kurz is a contributing editor to TV Tech. He has written about TV and video technology for more than 30 years and served as editor of three leading industry magazines. He earned a Bachelor of Journalism and a Master’s Degree in Journalism from the University of Missouri-Columbia School of Journalism.