Michael Capretta, VP of Global Research, Insights and Analytics at Warner Bros. Discovery, traces the evolution of the media business from cable TV's dual-revenue model through the streaming era, using that history to argue that media research organizations must embed data scientists directly within insights teams rather than separating the functions. Drawing on nearly 22 years at the company, he shares practical lessons from his own team's experience bringing data science in-house and what that shift meant for speed, accountability, and analytical depth.
Key Takeaways
- Embedding data scientists directly within a media research or insights team, reporting to the same leadership, produces better results than housing them in a separate tech or data division where competing priorities deprioritize research requests.
- The recurring subscription revenue model that cable TV pioneered in the 1970s and 1980s is the direct ancestor of today's streaming business model, and the structural tensions it created, such as leakage and fragmented measurement, are still unresolved.
- Media consumption is behavioral. Changes accelerated by events like a pandemic are likely to be permanent, which means organizations should plan for audience behavior shifts to be irreversible rather than cyclical.
- Turning a media researcher into a data scientist works only when the individual is already analytically inclined and self-motivated. Forcing the transition through training programs alone produces mixed results at best.
- The hardest part of organizational change is not adopting new ways of working. It is letting the old ways go, including protecting legacy revenue streams that still fund the business but can crowd out investment in the future.
- Bringing even one data science function in-house, even if the team leader cannot evaluate the technical work directly, can meaningfully accelerate delivery to internal clients and free research teams to spend more time on analysis and interpretation.
Questions & Answers
- When helping C-suite leaders visualize opportunities in a data-driven environment, how do you approach that?
- Capretta said leadership is generally forward-thinking and understands where things are going. The harder challenge is figuring out how to get there inside a large, mature company. He used the analogy of painting an airplane while it is flying: the company is changing direction while the existing business is still running. The priority is meeting viewers where they are rather than forcing them back into old viewing behaviors.
- Are generational shifts in viewing behavior happening faster than in earlier eras?
- Yes. Capretta noted that just two or three years ago Disney+ and HBO Max did not exist, and now they are major players in streaming. He also observed that the shift is no longer limited to younger viewers. Adults 50 and older are also viewing content very differently than they did just a few years ago.
- Can you give examples of legacy thinking inside your own organization and how you have dealt with it?
- Capretta pointed to his own focus on linear TV networks, TBS, TNT, and truTV, which are the part of the business that is not growing but still generates significant revenue. That revenue funds the company's investment in new priorities. The tension is real: protecting existing revenue long enough to fund the transition, without over-allocating resources to a declining model. At some point, he said, you do have to let go of at least some revenue to put resources toward where viewers are heading.
- In a matrix organization where insights and data science are in separate functions, how do you bring them together despite the organizational design?
- Capretta acknowledged limited direct experience with matrix structures. From his experience with centralized research functions, the benefits included quality control, vendor savings, and cross-divisional learning. In a matrix, he said it requires a lot of communication and becomes very challenging when things are moving fast. He noted that buy-side sales agencies seem to make it work, but did not claim to have a clear formula for the matrix context.
- Is internal data science better than a strong external partner?
- Capretta said a close external relationship can work, but internal is preferable because it aligns accountability and priorities under one hierarchy. The turning point for his team came when he brought a data science manager in-house. He noted that external data support is still necessary, but even adding one embedded data science role made a significant difference in speed and output quality.
- For teams without an in-house data scientist, can you describe the role and what it changed for you?
- The core change was prioritization. Tasks like consolidating multiple data sources into a single report had always been hard to prioritize alongside other work. When someone was hired specifically for that function and held accountable for it, it got done consistently. Initially it created some added pressure on the rest of the research team, but within a few months the data pipeline the new hire built made the whole team's job easier.
- How did having a data scientist change your own role and ability to focus on interpretation?
- It allowed the team to spend more time on analysis and insights and less on data crunching. Capretta described being able to deliver curated, filtered data to internal clients in sales, finance, and scheduling much faster than before. He sees the curation and filtering as a core part of the research team's job, not just passing raw data on.
- What content trends are you watching?
- Capretta highlighted live sports as a continuing anchor for the linear TV bundle, with significant changes coming on the sports rights side in the next few years. He also noted strong audience appetite for content that reflects or speaks to the current moment, citing Severance on Apple TV+ and Search Party on HBO Max as personal examples. He also acknowledged that well-executed unscripted television remains very popular as an escape.
Session Notes
Setting the Stage: Why This Moment Matters
Capretta opened by situating the talk in what he described as the most fascinating time to be in media research. The seismic changes of the past 10 to 30 years have created real opportunities for both businesses and researchers, but capitalizing on them requires stepping back to visualize those opportunities clearly and thinking strategically rather than reactively.
Cable TV as the Original Disruptor
To frame today's disruption, Capretta started with cable TV in the 1970s and 1980s. Going from three broadcast channels to 50, then 100 or more was genuinely groundbreaking. But the more consequential innovation was financial.
- Cable built a two-revenue-stream model: advertising (mirroring broadcast) plus subscriber fees paid by cable operators to carry content.
- The subscriber revenue stream was notably not dependent on viewership or usage, a structure that sounds familiar in today's streaming subscription economy.
- Audiences and subscribers both grew year over year, and Nielsen supplied common, transparent metrics, average audience, rating, share, and reach, that every competitor could access.
Because the business model and metrics were settled, what differentiated media companies was content. The competitive question of the era was simply: what is the next big TV hit?
The Familiar Defense of Legacy Players
Broadcast networks initially dismissed cable as low-brow and unserious. Advertisers paid less for cable than broadcast, a gap that persists today. Capretta's observation: legacy can be a hard thing to let go of, and that dynamic repeats itself in every era of disruption.
Internet, Fragmentation, and the Rise of Frankenmetrics
The internet era arrived gradually, then suddenly. A few key forces compounded each other.
- Content costs rose as cable nets competed to produce and acquire originals. Costs were passed to consumers through higher cable operator fees.
- Around 2014, subscriber counts began declining, first in younger demographics. Media viewing is behavioral; once viewing habits shift they are unlikely to reverse.
- Cable bills approaching $200 per month accelerated cord-cutting, especially among younger adults who no longer treated a cable subscription as automatic.
- In 2007, Netflix launched streaming. Industry reaction was largely dismissive. One executive famously compared Netflix to 'the hapless Albanian army.' Netflix's actual strategy was to build a deep content catalog and pay for it aggressively, which suited studios and networks looking for cash.
Measurement fractured in parallel. When the C-suite asked how many viewers watched a show, the answer became complicated fast. Mobile, desktop, pirated, hotel, bar, and out-of-home viewing all existed but were not uniformly counted or monetized.
The viewing that wasn't fully monetized became known as leakage.
Each company began building its own proprietary metric formulas, what Capretta called frankenmetrics: internally coherent combinations of data that were no longer transparent or comparable across companies. Explaining them to the C-suite was hard. And the common, shared measurement language that had made the cable era legible was gone.
The Talent Gap: Media Researchers vs. Data Scientists
As data complexity grew, organizations faced a structural talent problem. Capretta laid out the tension plainly.
- Data scientists were expensive, often 20 to 50 percent more expensive than a comparable research director title, and many lacked the institutional knowledge of how media data is actually used across sales, finance, programming, marketing, and PR.
- Media researchers with strong contextual knowledge often lacked the technical skills to wrangle new data streams independently.
- Efforts to retrain media researchers as data scientists produced mixed results. Success occurred mainly when the researcher was already analytically inclined and personally motivated to develop those skills.
- PhD-level data scientists were sometimes dismissed by media research teams for misreading viewing data, illustrating the depth of the disconnect between technical skill and domain knowledge.
Is it easier to turn a media researcher into a data scientist, or turn a data scientist into a media researcher?
Neither path is clean. The more practical answer, in Capretta's experience, was structural rather than individual.
The Structural Fix: Embed Data Science Within Insights
Capretta's central argument is that separating data science into a standalone tech division, where it fields requests from research and sales and finance simultaneously, guarantees that programming and marketing requests land at the bottom of the queue. Revenue-driven requests from sales always win priority.
- Insights teams benefit most when data scientists report into the same division, with the same goals and accountability chain.
- A common objection is that insights leaders may not have a data science background themselves and cannot technically manage a data scientist. Capretta's view is that this discomfort is manageable and the benefits outweigh it.
- When he brought a data science manager directly into his team less than a year before the talk, the group's ability to deliver data to internal clients, sales, finance, and scheduling, improved substantially and quickly.
- The change freed media researchers to spend more time on analysis and interpretation rather than data wrangling, which is the actual goal.
It was the first time I hired somebody to do a job that I couldn't do myself, and that can be kind of scary.
Behavioral Change Is Permanent
Capretta returned repeatedly to one principle: media consumption is behavioral. Changes that occur, even if accelerated by a one-time event like a pandemic, are unlikely to reverse. This has direct implications for how organizations should invest. Trying to maximize linear TV performance while deferring investment in where viewers are heading is a losing strategy at some point. Linear TV networks at Warner Bros. Discovery still generate billions in revenue and remain important to fund new priorities, but the direction of travel is clear.
Closing Thought
The hardest part isn't taking on the new ways of thinking. It's letting the old ways die.
Transcript
Read the full transcript
uh well again thank you to um thank you to everyone um who is uh been listening in uh this afternoon um and those of you who are joining us uh now uh for the first time today uh just a few reminders i think most of you by now are veterans of these events um but just remember we're at the mercy of tech um and so if something goes wrong we will try to troubleshoot it but just a little grace again you you know we encourage you to interact we encourage you to ask questions just be respectful um and and you know at the end there'll be opportunities for networking engaging and interaction after um our speaker um and michael i'm going to let you introduce yourself as well as pronounce your last name because i i'm afraid i may not do well at that um but before i do that i just let people know that um i'm gonna be monitoring the chat for you michael and um so as you know however you want to do it do you want to take questions as you go along or would you prefer we keep them to the end um probably doing a discussion at the end will work best for us today yeah wonderful so when we get uh when you get done just let me know and i will help field the questions that come in for you okay yeah we'll probably run it'll probably be about you know 30 minutes worth of stuff from me before we um get into uh q a which i'm looking forward to wonderful okay so i'm gonna stop sharing i'm gonna let you go ahead and share your screen okay i share my screen all right i'm gonna do it so hopefully everyone is seeing my screen now right so i'm michael yeah so i'm michael capretta i have um i'm the vp of global research and insights and analytics at uh warner warner brothers discovery which is what we now are as of just a couple weeks ago um so i've been with the company for nearly 22 years and have done a wide range of research um work for the company during this time and had stints on the new side and the corporate research side and currently head up a group that does audience forecasting and modeling for what we call the tnets which is tbs tnt and truetv um so you know i want to thank you know billy and bill for having me today this is great i'm glad that i uh you know finally got a chance to do to do one of these um i've been uh really impressed with them so far um so i'm just going to jump into my um my talk and and you know there's never been a more fascinating time to be involved in media research in the media business um the seismic changes we've seen in the in the past um 10 30 years and especially the past 10 um have created possibilities and opportunities both for businesses themselves but also for us as media professionals and today i'm going to try and put some context around the changes we've seen and take a step back and allow us to visualize those opportunities and to remind ourselves why we're fortunate to be in the throes of this pivotal time and think about how we can increase our strategic thinking and borrow a phrase work smarter not harder so while we don't have time to get into the whole history of television i'll i'll start with a previous disrupter in in media cable tv of the 70s and 80s to go from three channels to 50 60 then 100 or more was truly groundbreaking if you're of a certain age you might remember the very day when the cable truck came through the neighborhood hooking up everyone's cable or you might remember the day when you found a channel that played music videos and only music videos and it blew your mind and wait you know all they do is show videos all day long it was great and for the tweens of the day with too much time on their hands there was nothing better but business but as a business it was truly groundbreaking too cable tycoons of the day were making big moves and laying the groundwork for for the business for decades to come one key component of the cable tv business model they stumbled onto had huge implications even if they didn't necessarily know it at the time they smartly invented two revenue streams as the foundation of the business there was the ad supported business model just the same as broadcasters hadn't been doing um you air programming people want to watch and you sell advertising during the show easy peasy secondly subscriber revenue from cable operators they were so motivated to grow their subscriber base that they were willing to pay for the content something that they weren't doing with broadcasters at the time i suppose you could have called it itc right indirect to consumer um so that's interesting think about that so the key to the business was a recurring subscription revenue model and it wasn't dependent on viewership or even usage and that sounds kind of familiar today lots of companies are banking on a recurring subscription model to be successful fitness gyms have always done it uh maybe there was a wine in the month but at least then you were you know getting something tangible in return but now customers are being asked to sign up for a recurring subscription for everything from guitar lessons to online greeting cards to videos of people riding stationary bikes because paying 2 000 for the stationary bike wasn't enough for cable business was great audiences were going up year after year and subscribers were climbing every year and two streams of revenue were coming in and there were a lot of things about it that were easy even if it was overlooked at the time first we understood how an audience behaved they woke up and turned on the tv uh or radio they got in a car they listened to the radio kids come home from school turn on the tv families watch tv together at night so it was like the american dream okay sometimes maybe on the weekend they would swing by blockbuster and pick up a couple movies one action movie one rom-com one kid's movie and everyone's covered right but we also knew how to count all of this viewing the industry all agreed on the metrics and nielsen supplied it average audience rating share reach it was all transparent and everyone had access to competitors performance too as long as we protected the two streams of revenue all was good and since the business model was clear the metrics that got us there were clear there was no differentiation between companies and in terms of metrics what differentiated media companies was content and media companies could devote resources and energy developing or acquiring the best content so what took up our headspace at the time was what was going to be the next big tv hit so who knew that who wants to be a millionaire was going to be a national sensation which for a fleeting moment it was who knew that survivor would break open the floodgates on unscripted tv that were still drowning in 22 years later and the playing field was even enough in competing on content that the old guard broadcasters felt threatened tv cable tv was disparaged it is not serious and low brow content cheap production and not worthy of people's time um and that's a common defense whenever your new way of doing things disrupts the status quo right and we've heard the same thing today about new content streams that have sprung up cable tv cared about how they built a 24 hour seven day a week schedule they curated the programming for their audience they built brands cable tv worked hard to be broadcast replacement cable networks couldn't believe advertisers would not pay the same price for what was sometime the same programming on cable versus broadcast when old media companies were running down the cable nets that sort of worked the gap between what advertisers pay for broad for broadcast versus what they'll pay for cable still exists today but legacy can be a hard thing to let go of okay so there's something important and that sounds pretty familiar to today's needy business too so we'll come back to that the legacy can be a hard thing to let go uh but soon hit tv shows could come from anywhere the sopranos iron chef the closer the shield were all buzz worthy shows that generated as much conversation as broadcast shows and as the internet came to age in the late 90s early 2000s it was clear it would fundamentally change the business that was coasting along we thought we thought pets.com or stamps.com we're going to make a scritch and if you were smart enough to buy amazon.com because you thought it was going to make you rich i guess in that case you were right uh but most didn't and the internet bubble burst and then not much changed initially we sort of went back to status quo some internet companies got their hands on traditional media like aol capturing time warner and other media companies made sure to solidify their space in the online world with a lot of success as espn did with sports and cnn did with news but largely things remain unchanged while that specter of this newly connected world was bubbling under the surface i remember a former boss telling us we couldn't be tv researchers anymore that we had to be media researchers we were given reading material like uh who moved my cheese and the fish philosophy all dealing with adapting to change and it was great advice uh it put us in a mindset of welcoming change those working in media that have developed adaptability of thrive it was a it is a valuable skill to develop it remains more valuable than ever today you might hear it called change management there's a whole niche industry built around it and eventually the disruption came the content we were all fighting over started to get more expensive cable was now producing a mountain of original content acquiring content from studios and broadcast nets was competitive among those that needed more content to reach more eyeballs and when it got expensive the costs were passed on to customers via cable operators who were paying more in subscriber fees to keep quality content flowing from the cable nets so there's a big chain reaction there um around 2014 or so there started to be some cracks for the first time subscribers began to decline not necessarily in the households those still looked pretty strong but in the younger demos the decline was beginning to reveal itself those that wanted to paint a rosy picture could still do so they could say well those are just young people that don't know any better when they grow up they're gonna get a cable subscription it's still the best deal around but it was an early warning sign especially considering that media viewing is behavioral once it evolves it's unlikely to reverse itself and that's an important thing to remember and eventually customers started to push back on cable bills reaching 200 a month viewers started to cut the cord slowly at first but for a struggling young adult living on their own for the first time it was no longer a no-brainer that you needed a cable subscription for those too young to remember this ancient era it went something like this you make a call to a cable operator sit on hold for 45 minutes set an appointment for a six hour window that the cable guy would show up and they would always show up a half hour after the window closed or as soon as you're in a five minute errand but eventually the cable guy would arrive with huge nine pound cable box for each tv that you had uh hook everything up just in time to host a melrose place viewing party then you get the first monthly bill and the 95 a month cable package you thought you were buying was really 180 a month once fees equipment taxes and something called a regulatory recovery fee were included it wasn't exactly customer friendly and customers started to push back the number of ways to consume content was multiplying viewing was fragmented in ways we'd never seen before even if viewers weren't quite watching full-length feature film on their phones we could easily picture a time when they would be able to or to watch a live sporting event wherever they were in 2007 netflix announced that subscribers could directly stream movies to their homes and streaming was born at the time it was thought consumers weren't ready for this even our home internet connections may not have been ready for it but what netflix knew was that it was better to be too early than too late i remember the reactions inside media at the time and it was fairly dismissive saying things like they can't do what we do well they don't want to do what cable networks were doing they were playing a different game media execs weren't exactly shaking in their shoes one even famously comparing netflix to the hapless albanian army part of the netflix strategy was to have a deep catalog of content and they came with a big checkbook and cable tv was starting to see some of those cracks in the armor and they could use the cash and if netflix wanted to overpay for content the studios and networks were happy to let go and this had implications on audience measurement as well when c-suite asked how many viewers watched a show the answer was no longer so easy we were counting digital viewers we're recounting the digital viewers the mobile viewers pirated viewing were we counting laptop and desktop blackberry and flip phone etch a sketch did we count viewing in bars restaurants hotels airports that tv above the gas pump at the gas station so it was getting ridiculous right it was everywhere the viewing was fragmented but it wasn't always fully monetized that wasn't something media companies were accustomed to not monetizing every eyeball so when you could count all the viewers across all of these platforms it wasn't apples to apples because not all of it was monetized the viewing that wasn't fully monetized became known as leakage so now leakage wasn't just something middle-aged men were dealing with but middle-aged media companies as well and what was the best way to capture all this leakage depends all right so it's one thing to put a depends like joke in there but it's not like it was a whole slot i was like committed to this joke so let's move on all right media researchers did the best they could and uh started blending metrics together and the best name for it was frankenmetrics to describe what we concocted um they made sense to those that created them but explaining it to the c-suite was a little harder these frankenmetrics were no longer common and transparent across companies each company had their own little black boxes that contained our secret metric sauce and if we were media researchers and no longer tv researchers what did that mean and how did you become one it meant learning new metrics and parsing the non-traditional tv viewing how do we measure who is watching bootleg adult swim shows on youtube is youtube going to give us a rating for that and can nielsen measure it for us so nope and no as with many industries data analysts and data scientists became increasingly necessary to sort through the unregulated metrics that were being concocted sometimes on the fly and the truth was there were many tv researchers that didn't have the skill set to pull off this transition it is not easy and i've seen media research struggle with the most basic statistical analysis i've been there myself but there were also a lot of data scientists that couldn't grasp the nuances of what influenced viewing and it raised a question for us is it easier to turn a media researcher into a data scientist or turn a data scientist into a media researcher well data scientists were expensive you might have an opening for a research director let's say but you soon find out that hiring a data scientist or a data director or even a data manager might be 20 or 50 percent more expensive than what than a media research director and they wouldn't even be equipped to do most of the job functions the person they were replacing was doing so status quo wins and you end up with another media researcher so efforts were made to turn media researchers into data scientists and that got mixed results at best the the truth is it is usually only successful when the researcher already has a knack for data and more importantly is self-motivated to do it so this media researcher you want to shape-shift into a data scientist they have to be self-motivated because after trimming down all areas of the organization no one could afford to continue doing the job they were currently doing um or really the two and a half jobs they've now been saddled with since trimming trimming down staffs that meant they were doing it on their own time or even better they were finding ways to automate their traditional job which freed up time to explore the skills for the new world they were entering part of the secret was to not tell too many people they've partially automated their job that report that took four hours every week to produce don't tell your boss it now only takes you 15 minutes so next company started to build out data data decision science teams to navigate the flight of data to support the non-traditional tv viewing data the instinct of the company was to have the data team report up to a tech division which would then support the media research that reported up to a different division again mixed results the data team is flooded with a wide range of requests the requests on the sales side or the ones directly impacting revenue were always going to correctly be prioritized first they usually didn't have time for programming or marketing requests that fell to the bottom of the list and the media researchers understand the context of data they're providing and it usually isn't black and white is the data for finance or is this for marketing is it for content developers uh is it addressing an roi question is it a pr claim so there isn't you know an outcome for the data that works for every group the institutional knowledge of how the data would be used in in a media company was important and sometimes overlooked as we became enamored with all of the new data streams that were available the nuances of how the data was being used mattered and i've seen data scientists with phds some of the smartest people you'll ever come across get laughed out of a room of media researchers by misinterpreting viewing data it was a stark revelation of the disconnect that existed so all this is to say and here's kind of my hot take and where we'll bring this together not just for media for for a lot of fields that are pulling together data and insights is that if you are in an organization that is still structured with an insights division that does not have data scientists embedded in it you're probably still struggling with marrying these two disciplines media researchers are going to most benefit from the data science support when working under one unit reporting up to the same division with the same goals and accountability data science will be best used when embedded within this organization this is challenging it's hard um the way that people move up in an organization is to be good at their current job and get promoted to supervise the people without doing the job they used to do in a truly embedded organization a data scientist might be led by someone that is not really a data person or didn't have come up in the industry as a data person they may not know sql from python and that can be difficult for people on both sides of that dynamic um this increases in the rate of change for the n string is is going to continue right media consumption is behavioral as i mentioned before changes that occur for viewers even if they're accelerated by something like a pandemic are likely to be permanent in nature you get the feeling that we're always on the cusp of another seismic change uh whether it's crypto or nfts or web 3. my colleague of mine told me had i was obligated to get those words into the presentation somewhere right every research presentation has to have them in there now um but the need for for a marriage of data science and media research will be more important than ever um so everyone has to continue to embrace change and that's no big revelation uh and everyone's on board with that um but in conclusion you know the hardest part isn't taking on the new ways of thinking it's letting the old ways die so thanks everyone for um letting me talk with you guys today i'm happy to have a discussion or take some questions and billy's going to moderate she didn't like my she didn't like my joke that was in the middle of the presentation i loved it in fact um yeah in fact michael thank you so much it was this was a very refreshing presentation um it showed your skills as a storyteller that everything was visual and i think nowadays when we see so many slides they're just full of words and charts and so forth especially in the research world it is so refreshing to see this type of visualization and you talked about that at the beginning you know visualizing opportunities so thank you so much for for presenting this to us at this time if anyone has any questions for michael michael and i will just start talking if anybody wants to jump in with anything that would be great um i guess i will kick off the first photo we're letting people kind of digest what they just heard you talked about at the beginning seismic changes and obviously in the media space you you've kind of profiled what some of those look like um we've all seen it especially as uh you know we've moved to streaming content and that's obviously really changed your world right um so when you are in research and you're helping people internally especially at c-suite visualize opportunities how do you approach that especially when a lot of this is very data-driven information um that's not the i don't know that that's the hard part honestly i mean i think that the leadership really is pushing in in in these directions and they're very aware of of um what the future holds and are you know very good at you know very forward thinking about and understand where things are going um when we're asked to figure out how to get there that's the challenging part right and and getting there getting there can be the challenging part especially when you know that i work in a company that is um you know a mature and very large company and it's it's that's when it can be hardest to change directions and um you know sometimes there's a there's an analogy of you know we're we're painting an airplane while it's flying right we're really trying to um change the company and um we prioritize how we're you know how to how to meet viewers where they are and not just um try and force them into a an old way of viewing content that they're not on board with anymore right we have to be able to you know as i said you know you know embrace the change that's happening are you seeing um quicker shifts across generations so like for instance you know our generation versus the millennials versus the next generation that's coming up behind them are things changing at a how fast are things changing compared to maybe you know 50 70 years ago even 20 years ago oh yeah i mean it's yeah it's incredible how much you know quickly things are changing right now and you know in just the past two or three years things are you know the landscape of viewing looks very different you know just to think that you know two or three years ago that you know disney plus didn't exist and hbo max didn't exist right and now those are among the major players in the in the streaming streaming wars immediately so um you know changes like that have you know a huge impact um and the impacts in terms of cord cutting and in older viewers adapting you know to to change in this changing way to view content the way that you know younger viewers take it on so easily but you know we're seeing it now in in some of the older you know older demographics of you know you know 50 plus you know where is also yeah um viewing things very differently than they did just a few years ago well and to that point then you also talked about and you said that this was really important and i wrote this down because i love it legacy can be a hard thing to let go of but it's the only way to go forward right so can you give us some examples or talk to things within your own organization that you have found um even though they sound very open-minded they know they have to be progressive where you've encountered something like that and how you've how you've dealt with it yeah i mean look i still focus you know most of my current position is based on linear tv networks right now right and that's uh we all understand that's the part of the business that's not growing it is shrinking um but it still generates a lot of revenue you know billions of dollars for this company so it remains very important to keep that business as healthy as we can so that it can fund the new areas of the the company and the new um priorities that the company has to to move forward so um you know i think we are all in uh on board with with the changes that are happening um um of course when there's revenue involved it can be it can be difficult and it's a you know it's a hard um decision to make when it's when it's time to let go of revenue to view to put resources into um where we think viewers are going to be in the future um so it's some point you do have to let go of at least some revenue to continue to fund that if you're just gonna try and continue to maximize linear tv networks and linear tv viewing it's it'll be a tough task okay i just realized i had the chat button open but the q a there are questions coming it's i apologize megan i'm going to uh to go ahead and ask your question um with large organizations oftentimes being in matrix of functions where insights and data science are separated by the functional organizational design how do you think these two can come together in spite of organizational design um that's a good question so um so in a matrix organization yeah those are that can be really tough you know i can only you know speak from the experience that i've had and most of my career has um been working for a you know a centralized research function um which for for uh our company and for in my experience worked really well um that you know we were um you know reporting up to a central research chief that could make sure that we had a lot of quality control that we were having savings across organizations in terms of the vendors we were working with um in terms of making sure that the researchers across that you know whose work was concentrated across different divisions were learning from each other along the way those were all some huge benefits to having that centralized organization for us if you know in a matrix organization which i you know only kind of had a little bit of experience with it took a lot of communication um and um uh you know ends up being you know really challenging when things are moving fast it can be done i'm sure and i'm sure there's a lot of places where it works well um but i don't i don't have as much experience directly with it do you know of or have you seen a company that you felt was more effective in this area um you know i think the there's some of the sales agencies um tend to be organized this way and it seems to work for them um and it works in that in that kind of area of the business on the buy side of the media business um so that seems to be where it works best we have another question coming in um and first off he wants to tell you a great presentation it brought back a lot of great childhood memories for him um and then he wanted to ask the question do you find having the data science as functional as function internally versus as an external partner better or can you develop strong external support as well um sure i mean i think you can it would take a real close relationship but i you know i i do like um when it's internal you know as i said i think when um it's laddering up and having the same priorities is a is a big um is a key component to this right that um uh it keeps accountability in line under you know under one hierarchy and that can be um harder when it comes you know from an external group but absolutely yeah you can have um strong external support um and i certainly do right now from a lot of divisions in this company with relationships i've built over the years and and it's worked really well but um less than a year ago when i was able to bring some data science knowledge directly into the the group that i oversee that that to me was when we were able to make really um [Music] big steps forward right and that was kind of the you know the inspiration for my presentation in the first place that i saw the change that it had for me and my team when we were able to bring it in house and um you know as i mentioned in the presentation it was the first time that i hired somebody to do a job that i couldn't do myself um you know and that can be kind of scary for me um it's probably scary for the person that i hired to do that job also that you know they don't know how to do my job um and so you know that was that disconnect i talked about a little bit but it um but the benefits were have been humongous so far and really set us up for a lot of success there's a lot of um still external data support that we need but even bringing just um some of that function directly into our group makes a huge difference and michael for those of us who don't have interaction or a data scientist in our companies um for access to how do can you kind of explain like the role that they play kind of how they change things for you the approaches that they take and how that role that they they serve yeah i mean it allowed um in some ways it was just a you know a prioritization you know i wouldn't you know i couldn't ever you know prioritize spending that much time wrangling all these different data sources together to um that we wanted to bring into one report for example it was just hard to prioritize that when we bring someone in and we make that their primary job function um you know well they're held accountable for it now and they're you know spending their time doing it and it was you know us as a department um saying that that was a priority and carving out a position to do that um so for me that's that was the biggest difference um and it meant um you know squeezing some other members of the group too when um that um weren't part of that kind of data science function some of their at least certainly initially some of their um job felt a little busier and a little tighter because there weren't as many um headcount resources being being put to the work that they were in charge of but it you know only if you know a few months later i think we're on the other side of that and already um making the job easier for our team of media researchers based on the flow of data that we've been able to set up with our data science manager in the group if anyone else has any questions please feel free to put them in the q a box um in in in doing that so the prioritization and having someone be able to really be focused on that i'm curious as to how that changed your role in your ability to spend more time in the interpretation and the activation of the information yeah absolutely it's um and that's the goal right is to give us more time on analysis and insights and less time on on crunching numbers or wrangling data and it's allowed us to turn things around to some of our clients whose you know jobs are to do those same things you know we have clients that we're accountable for on the on the ad sales side or even in the you know in a finance function or even in the scheduling side of the business um that we're able to provide them with data much easier than we had been before um and so it you know not only benefits our group that we can spend more time on the analysis and insights and pass that on to our clients um in terms of hey this is what we're seeing in data and and here's the data that you know that we've that we've been able to um pull together and you know and pass on to them in a you know in a simplified and kind of paired down way so that it's not dumping data on them that we're you know we are curating and and filtering that data to what is important that's you know i see that as part of our job um so yeah it has definitely allowed us to to spend more time on the insights and analysis than the data crunching absolutely and then listen thank you for that um and unless anyone else has a question i have one more and it's kind of a fun one i think i'm just really curious what you're seeing in regards to trends regarding content um yeah you know i don't know it's um it's interesting it's it's hard to say it's um you know i know what i like on on tv and and um and i watch a lot of sports and i watch a lot of live sports and that's gonna you know continue to be something that you know holds together a tv bundle right and we'll be big changes on the sports side you know coming up in the next few years for sure um but i also watch a lot of the um a lot of content that is um um a little more um futuristic or kind of that kind of speaks back to the times we're in um i'm watching severance on hbo no on uh sorry it's on apple plus it's not even hbo max show uh it's not apple plus you know i love the the most recent season search party that was on hbo max for fans of that show it was one of my favorites um so those things that kind of speak to the times that we're in are things that i like but you know there's you know there's something for everyone right there's there's such a you know mountain of content right now that um there's you know something for everyone out there right and a lot of people are finding that they like to escape into the you know um silliness of you know some unscripted television and you know that is um uh very prevalent right now right and and so and can be really well done at times michael thank you so much uh you know i i think others are commenting in the chat thing that they really enjoyed this presentation as well like i said um you're it was very refreshing and i love the storytelling aspect of it and thank you so much it's a very interesting space that you're in great well i'm glad everybody enjoyed it yes i didn't want to flood people with more data and numbers and charts and thought it might be a good um yeah a good uh something a little different from what they sometimes say
