NPAW's Bruno Griner and ESHAP's Evan Shapiro Talk Better Data for Better Business Outcomes at Streaming Media Connect 2025
Streaming Media Connect 2025’s featured a fascinating Keynote Fireside Chat featuring NPAW VP Sales Global | DXI Bruno Giner and Media Universe Cartographer Evan Shapiro. The discussion focused on the importance of data in the streaming industry and NPAW's strategic approach to it, emphasizing the need for democratization and unification of data across departments to improve business outcomes.
Taming the data universe
"The data universe is so extensive, for years and years we've always had users who say, 'I need a data lake, I need data warehouse, I need this, I need that.'" Giner explains that NPAW tries to take a step back from some of the hysteria around data needs and move the discussion towards "what we need data for," focusing "on specific business goals and specific or simple problems that we need to solve."
Noting that NPAW has partnered with and helped some of the largest, most data-rich organizations in the media business—Rakuten, Tubi, Sky, and others—manage and interpret their data, Shapiro asks, “What's the secret to creating a data stack that makes sense for a large organization?"
Some platforms, Giner says, tell NPAW their goal is "to improve acquisition, to increase our user base, or to increase our consumption, to make the best out of the content investment that we have done. Or we need to reduce churn. These simple questions are the most important ones." Others, he says, "need to build ground to understand users' behavior and understand engagement."
Some need to understand "the whys" from each platform: "Why do people like your content, why do people subscribe to your service? Why do people come every day? Why do they stop coming? Why do they churn? Those whys are what we need to answer. From our perspective, what we do is to build a complete understanding of these digital experience from visits, navigations, user journeys, content discovery process, the engagement with the content itself, the preferences of the users, how each segment behaves."
All of these factors, Giner says, help NPAW understand "the whole digital experience in that specific platform with that specific offering. That is what we try to offer with a very quick time to value. We implement this so it gets visibility, and then from that you can breathe and think of your business strategies."
Pinpointing User Behavior and Democratizing Data Access
Segmenting, understanding, and predicting user behavior, Ginger says, are also crucial to improving user experiences and reducing churn. Precise targeting is critical; painting user behavior with too broad a brush is ineffective, but platforms must have access to data to zero in on user behavior in an actionable way.
"It's not just about audience behavior writ large, it is about specific audience segment behavior. If they came in through an Android app or they came in through an OS app or they came in through a native app or something else, how do they behave once they get there? What's the first piece of content that they [choose] and what are the other habits of the people who come in? People invest so much money in contenr. After they watch what happens with that content, what do they do next? The trails of human behavior through these apps is a big part of what produces all this data. You should use data to predict your customer's behavior so that you can make better choices around the content, Identify glitches in your technology where there's too much buffering and you're losing customers in the first 50 seconds in their visit because they're waiting around too long."
Shapiro interjects that when "the time to discovery is too long, you're not satisfying them." Then he unleashes a barrage of questions: "What's the major blockage? I was just talking to a very large MVPD who also does some telecom business as well, so they have millions of customer relationships. But what is the key to having good data health? What are the building blocks to avoid having to corral all these different inputs later in life? What can you do now to improve the data stack. Where are the foundational elements to creating one from scratch that we should all think about?"
"Democratization of access to the data in the organization," is critical, Giner says. "This comes from a unification of the usage of the data." Organizations can no longer rely on "a guru department that people go to request a report and then based on that report take action without a unification of this understanding. This is a huge problem," he continues. When platforms "purchase content, we purchase content for a reason. We have a market view, we understand our audience, we buy this because we believe this will work, but then there's a disconnection from how we promote this content, how we offer this content, how we surface this content in the platform, how we support each of the segments that we are targeting. This needs to be unified. When we purchase content, we need to tag it; we need to have a catalog with good metadata. This is super-basic. After that, our marketing team needs to understand from our audience which categories to promote to which part of our audience, and then how to measure that. All of these actions need to be unified and this intelligence needs to be democratized so that all the departments that are using the same intelligence have access to the data, see the same truth, and be able to measure that as well."
First-Party Data and Advertising
The interview also looks at leveraging first-party data for better ad targeting, and the critical role of data in driving strategic decisions and improving user experiences in the streaming media landscape.
"On the ad sales side, what I have found is those advertising-based platforms who can manage their first-party data well and marry that to some external data can offer first-party data in a clean room environment to enhance the performance of the campaigns in the same way," Shapiro contends. "The thing that most of these platforms really need to do better in comparison to the big walled garden is to manage their content and offer their content to their partners in a safe way that enhances and improves the outcomes of the campaigns of their partners. Are you working on stuff like that usage as well?"
"For ad sales, we have been supporting that for a while," Giner replies. "And to be honest, pretty recently we started to get new requests like, 'Hey, I want to know how my conversion is working with this specific segment that this marketing team has built. Can you measure that?' Yes, we can. Or they say, 'I want to know for which content, this type of ad converts better' and so on. This type of request is coming more and more. And right now there's this concern about regulation as well on how do I make better use of my first-party data? So how do I make better use of the segments that I know? How do I make better use of the leverage that my content gives for knowing this audience and so on. So this kind of cross understanding of audience and content towards ads [is something] we are seeing a lot as well.”
Predicting and Enhancing User Engagement with AI
Giner also highlights the role of AI in understanding user behavior and optimizing content strategies to reduce churn and enhance user engagement.
"Right now we are able to use AI, for instance, to quickly know if one user is saturating on the content that is available to them and that they would like to see. So this is the group of content that would be likable for this guy and he's consuming all that he has," Giner says. "This type of understanding is something that people are really starting to use to measure how their catalog is [performing]."
"So you can predict what someone is going to binge?" Shapiro asks.
"Yes," Giner says. "And also you can look at, 'My power users would binge this content, but they have watched it all, and right now I need to buy more to satisfy this group of users. It's something that people are starting to work on more. And right now there are a lot of trends with how to use AI agents and so on. But as I told you the last time that we spoke, I believe that we need to start with the simple questions and to define [platforms' basic needs] to then be able to apply this type of technology."
Thanks to AI and the way it advances how user data can be interpreted, Giner says, "We can also understand churn." Giner recalls a platform partner that had 5 million subscribers. "They used to have a free content offering in the platform and also SVOD content in a proprietary content library. There were a lot of people who were subscribing and kept just watching the free content. This was a key point. The time that users took to start engaging with the SVOD library was super, super, super impactful in the churn that they used to have. If they could cut seconds off of the time to discover—”
"It really matters, right?" Shapiro interjects.
"Yes. They saw that if they could reduce it to 15 days, they could reduce the churn rate from 12% to 5%, which for them--with 5 million subs--was a huge amount of money. Understanding this type of conversion activation gives us tools to know what to do to improve business KPIs."
Related Articles
Moving streaming video services from 'serial churn' to 'pause and resume' in today's subscription economy.
16 Sep 2024