Data Collaboration: Privacy, Clean Rooms, and Interoperability

[music]

Dana McGraw: Don't just collect everything you can collect. It doesn't make any sense. Do what you need to do to make the consumer experience the best one possible and to maintain that relationship of trust like that's kind of our guiding light independent of everything else.

Kailey Raymond: Hello, and welcome to Good Data, Better Marketing. I'm your host, Kailey Raymond, and today we're diving into data collaboration. The tools that make it possible, the role of data clean rooms, and why interoperability is key. In the realm of adtech, striking the perfect balance between personalization and privacy is a nuanced art, one that is mastered by placing the consumer at the heart of every decision. This consumer-centric philosophy ensures that data processes are conducted with respect for consumer expectations and privacy regulations, which in turn builds trust between the consumer and brand. The evolution of technology and the shifting regulatory landscape demand innovative approaches to data collaboration, and Disney is at the forefront of this change. Joining me is Dana McGraw of Disney Advertising. We discuss how data collaboration is evolving, the importance of interoperability, and Disney's revolutionary audience graph.

Producer: This podcast is brought to you by Twilio Segment. 92% of businesses today are using AI-driven personalization to drive growth. However, successful AI-driven engagement is only as good as your data. Be it targeting your top accounts with relevant ads or delighting customers with personalized experiences both online and in-store, Segment has helped thousands of companies like Intuit, FOX, IBM, and Levi's be AI-ready by laying a foundation of data that they can trust. Want to get your data AI-ready? Learn more at segment.com.

Kailey Raymond: Today, I am joined by Dana McGraw, Senior Vice President, Data and Measurement Science at Disney Advertising. Dana leads teams responsible for data science, analytics, audience modeling, in addition to attribution, advertiser performance, and insights. She leads data science and advanced analytics for advertising sales, covering Disney's portfolio of linear, streaming, digital, and social properties. Previously, she's held several positions with Yahoo's media group and analytics roles leading interactive agencies. She's also coached women's basketball for seven or eight years. So Dana, I'm excited to learn about that. Welcome to the show.

Dana McGraw: Thank you so much for having me. I'm excited to be here.

Kailey Raymond: Tell me, Dana, a little bit about your career journey. Of course, I just mentioned, you know, you started in basketball. Now, of course, you're leading a large group of analysts and data scientists and people that are running models in the background. Walk me through how you got from there to here.

Dana McGraw: It's definitely not a straight line. It's been a little bit of a winding road, but I think that's a great way to get where you're going because you get a lot of experience with a lot of different things. But yes, I played basketball in college. And then while I was in grad school, I was sort of in a graduate assistant position coaching basketball. And I had loved basketball since I was little, since I was, you know, three, four or five years old. And I wanted to keep doing it. Wasn't really good enough to continue a playing career by any means, but I fell in love with the game and I fell in love with coaching. And so, I finished my master's degree and I wanted to keep doing it. So I coached basketball, Division II, Division I level for seven or eight years. All different places, moved all around the country, which was also an amazing life experience. And then, I kind of decided that that wasn't the direction I wanted to go for the long term. I didn't aspire to be a head coach. I really liked being an assistant coach. And I felt like, you know, pre-30 years old, maybe not having aspirations to get to the pinnacle of your career wasn't what I wanted.

[laughter]

Dana McGraw: So I actually stopped coaching and I moved to LA. But during the course of coaching, I fell in love with the analytics and kind of the math of basketball. And so I sort of made it part of my job at every stop to really be analyzing the data and thinking about, you know, these are our two best players, but they're not our two best players together and so how do we think about that math and how do we move things around? I moved to LA and I just got lucky. I stumbled into a job in search engine optimization at a time when nobody knew what they were doing in search engine optimization. [laughter] So not having experience was fine. And so I ended up getting a job at a small search engine company. And there was one guy there who was focused on analytics and he left and they sort of just were like, you seem smart. You want to try? And I was like, yes. I definitely do. And so that sort of set off this journey. And as you mentioned, I got a job doing sort of analytics type stuff at Yahoo Sports. And so that was my first foray really into this side of the business, the publisher side and thinking about the advertising side and that kind of thing. And so, that was my journey to get here in a winding road. And I still have that same passion for sports. So it's great that, you know, I get to work on ESPN. It's one of my favorite parts of my job because I still have that passion.

Kailey Raymond: That's amazing. And obviously, I'm talking to you on the heels of a groundbreaking women's tournament, which saw record-breaking viewership, nearly 19 million viewers for the final between Iowa and South Carolina. I have chills thinking about it. I played basketball in high school, not college, so I never made it as far as you.

Dana McGraw: I mean, it's like for those of us, you know, younger than I am. But for those of us who were doing this early on, even my college games, they were played at a smaller school, but there was nobody there. I would just scratch and claw to find any game of women's basketball I could watch. I was obsessed with the Olympic teams. Any game that was ever on TV was generally just the final four at that time. And I was probably the only one watching. And so, just to watch the growth and just see what these women are doing and really what ESPN has poured into women's sports and the support. Yeah, it makes me super proud to work here and watch just the commitment to women's sports at ESPN. And we're seeing advertisers too, the commitment of advertisers to women's sports. And they're really understanding kind of what it means for their brands, which is awesome because that's how the game will keep growing.

Kailey Raymond: Totally, yeah. This is such a cool intersection that I'm sure you're feeling right now of like the ESPN win, that momentum, a lot of the thrust that they put behind this, you know, sport is finally paying off. And yeah, I'm excited to see what happens in professional sports I'm attempting to get a bunch of liberty tickets for the season so, maybe I'll root against you. Are you a Sparks fan?

Dana McGraw: You know what? I'm a basketball fan. But yeah, certainly I watch the Sparks because they're on more. I get to see them. And with all the streaming, you can kind of be a fan of any team, which is kind of the amazing part of where we've evolved in coverage of sports and really any kind of media is the ability to stream. You can find your local passion wherever you want to, which is pretty awesome.

Kailey Raymond: I love it. Well, I'm going to, for the sake of everybody, we're not gonna talk basketball the whole time. I promise, we could. We're going to talk about data. We're going to talk about some of the applications of it within advertising. So I want to pivot and I want to ask you about some of these trends that you're seeing within this space, in particular as it relates to the consumer experience in the Adtech space. What are some of those big trends you're watching?

Dana McGraw: Yeah, I mean, I think if you think about Adtech advertising in general, what we're seeing with consumers is exactly what I just mentioned with regard to streaming. Like just the change in the way audiences are consuming content. And so the way that we think about advertising, the way that we think about what the Adtech needs to be to support that advertising and all these different platforms. And really, it's about obviously, consumer first for us. I work for the Walt Disney Company, so the consumer is always our North Star, regardless of what part of the company that you work in. So thinking about how the advertising experience is as beneficial to the consumer as it is to the advertiser, and kind of how those two things can come together. And I think we're seeing more and more that growth. So kind of the right ad at the right time on the right platform to the right consumer. And that matters. It matters for the brand, and it matters for the consumer because consumers, generally speaking are great with ads if it's the right ad experience. And so I think that's where we're getting is not this, you know, historical one size fits all just ad experience, but a really tailored ad experience that works for both the brand and the consumer.

Kailey Raymond: Yeah, I totally hear you. You're really talking about that deep level of personalization that really attracts people to be brand loyalists, which I think is just really important. One of the things that I'm curious to hear your take on is, it's no secret that there's an increasing interest in data privacy. And I know that you're saying consumer first. That's obviously a really big part of it from consumers and advertisers alike. So talk to me about this mega trend of privacy and consent and how that's impacted Disney's strategies for advertisers.

Dana McGraw: Yeah, I mean, like I said, for us, the consumer is our North Star. So we've always erred on the side of is this what the consumer expects from us? Regardless of what we've asked or what we've asked them to consent to, going beyond just that, is this what they expect from the Walt Disney Company? Is this a viable expectation of how data is being used in our relationship? The whole company is based on a relationship of trust with the consumer, right? From any aspect of our business, that is what the Walt Disney Company represents to people. There's an affinity, there's a trust there. So even for me working in the data space on advertising, that's exactly what leads us as well. Don't just collect everything you can collect. It doesn't make any sense. Do what you need to do to make the consumer experience the best one possible and to maintain that relationship of trust.

Dana McGraw: Like that's kind of our guiding light independent of everything else. And because that has always been the way that we've conducted ourselves and thought about this, really for us, you know, all of this change in the privacy market, it's great. We're all for it because we want to maintain that relationship and respect the consumer the right way. And so, we've designed everything that we do around that idea. So even as these things change, sure, we're super mindful and we're very, very aware and making sure that everything we do is the right thing to do. But it's not some inhibitor for us because of the way we think about it in conducting business.

Kailey Raymond: It's really powerful. I think that it really shows that Disney, at the very center of the decisions that they're making, is thinking about the consumer at every step. So no matter what the change is that the consumer is demanding, if that's your North Star, then it's easier to say yes. Very cool. Data collaboration is obviously an evolving field. And I know that you're building some really cool technology to keep up with what we're talking about. So, you know, privacy, consent regulations. But of course, making sure that those consumers, like we talked about, are those North Stars. Can you break down what some of those solutions are and how they actually aid in improving the consumer experience?

Dana McGraw: Sure. And it's multifaceted for us. I mean, historically, you know, you're thinking about collaboration termed use kind of loosely, but you're thinking about third-party data onboarding and all of those kinds of things, right? And with the sort of evolution of clean room technology, for instance, you start to think about data collaboration where you are able to gain insights from one another, but data doesn't exchange hands, which I think is a lot more comfortable for all of us as consumers as well, right? So there's two parts. There's the technology to be able to do that and gain some understanding and insights from one another, again, to optimize the experience for all parties without data exchanging hands. But there's also this notion, and I was kind of referring to it loosely when I was talking about how we've always conducted our business when we think about the consumer. And so we've spent the better part of 10 years building what we call a pseudonymized audience graph, such that it's not identifying sort of individuals, but it's the backbone of everything we do.

Dana McGraw: So with that audience graph, you start to be able to reconcile things when you're collaborating. And so there's this notion of somewhat overused term at this point, probably because of us, but interoperability. And so, how do we take this audience graph that we've built, that we feel really confident about the way that it's built? How do we use that to be interoperable with, you know, whether it's measurement, partners or brand partners or whatever it is in this clean room space with this technology so that we're able to collaborate, we're able to leverage what we've built. And really, it's about accuracy. And that accuracy is what leads to the better ad experience, right? The accuracy of the measurement to understand like, okay, this was the right audience for us. These are our high value consumers that we wanna reach again, or we wanna reach more people like them. And the accuracy on the front end of was the ad served to the right consumer who has interest in this particular thing. So it's really one that backbone of the audience graph. Two, the kind of clean room technology and other technologies, frankly, that will emerge over time that allow us to collaborate and understand one another to deliver a better experience without data exchanging hands and moving around.

Kailey Raymond: I like it. You're keeping it safe for consumers. But there's this thing that we talk about here at Segment, which is the privacy and personalization paradox. I think that sometimes advertisers can fall into the trap of thinking that you're giving up, it's a trade-off, that you're giving up one for the other. But to your point about, you know, if you're collecting maybe the right data or not all of the data, like the things that actually matter with the consumer in mind, then I do think you can have both. It's like a myth, this paradox. Like you can have privacy and personalization. And one of the things that leads that is first-party data. So I'm wondering this intersection of data clean rooms and first-party data. Talk to me about how these two concepts are related and how they work together in improving targeting, but also respecting people's privacy.

Dana McGraw: Yeah, and I'll speak from the publisher perspective for sure. So for us, obviously, we take a lot of pride in our first-party data because we have the relationship with the consumer we do. If you don't have that relationship, if you don't have that kind of trust, you're not going to have a wealth of first-party data in our content. We have world-class content that people want to see. And so that combination is sort of perfect. But like for me, I get so excited about it, almost as excited as we're talking about basketball. But I get so excited about what that means for us and what it means for our consumers. So that wealth of first-party data and knowing the way that we treat it, as I mentioned, we pseudonymize. And so there was this notion of, you know, as recently as a couple of years ago, you would hear, no, if it's not one-to-one targeting, then it doesn't work. It's not possible. And I've never believed that and we've never done it that way. So, you know, there were probably a lot of wealth brands, agencies or whatever who felt that really deeply and maybe thought like, well, Disney is not doing it the right way for us because they're doing things in a more aggregated way. The reality is, when you get to this place with clean rooms and all of these other things where we're able to say, no, this is our first-party data.

Dana McGraw: This is what we're able to tell you about this consumer base. And by the way, we've created interoperability with our audience graph, measure it, and then come back and tell us if you still think the only way to do it is this one-to-one way that you've been talking about. And the outcomes start to speak for themselves. So when we're able to create that interoperability for measurement, leveraging our first-party data, and leveraging our audience graph, then it's like, oh, these two things are not at odds. You can do this in a super respectful, compliant way and also achieve the goals that you're trying to achieve. And so that really, I feel like for us, has helped us a great deal is getting to that place with the technology where we're able to ensure accurate measurement to say, "Listen, this is what we need." You can look like model this, you can do this in other ways that maintain that trust with the consumer and you're still getting the outcomes that you want to get.

Kailey Raymond: Yeah. That's exactly the outcome, obviously, that you're hoping to achieve is making sure that it feels like it's completely tailored to that one-to-one individualized person, but making sure that you're not breaking that level of trust that you've been developing with them over years. I'm curious about like, cookies, pixels, things like that, they're going away. And so how we think about matching and resolving that data, what's the solution for identity resolution in this environment?

Dana McGraw: Yeah, and that's that interoperability piece. Another thing that I get super excited about, because I'm kind of like a proud parent, like we have a fantastic team, and they've done a really, really good job. And we're all like-minded in our approach to the consumer and the right thing to do. So creating this audience graph that we have, which you know, it's- based on over 400 million unique identifiers, resolves to 110 million households and this is US, obviously. But when we can leverage that and we're able to resolve identity for a brand or for a measurement, we're able to take that and create an interoperability. But what we do is we create what we kind of loosely call a synthetic ID. And so that synthetic ID is not replicatable.

Dana McGraw: It's not revealing identity of any kind. But we're able to, whether it's a LiveRamp ID or an Experian LeWitt ID, or in the case we work with the Trade Desk in there, UID2, we're able to, on our own through a clean room, resolve that identity, create that resolution with an ID that we're creating on the fly that's not exposing any data whatsoever, again, to get back to the accuracy point. And so, that audience graph I mentioned in the beginning, it truly is the backbone of everything that we do because anything that you think about that you wanna kind of connect the dots or you wanna resolve, it's that audience graph, but it's the ability to create that synthetic ID out of that audience graph and that interoperability with any other ID solution that you could think of. That's what's really getting us somewhere with regard to how accurate we can be.

Kailey Raymond: That makes perfect sense. As a marketer who is using multiple channels every single day, what we're talking about with interoperability, the ability to actually activate some of what you're talking about across a lot of the different channels that are important to us is incredibly important, keeping in mind that you're activating those really robust, complete customer profiles and hitting the right people with the right person, right ad, right time, right channel.

Dana McGraw: And we hear it loud and clear, whether it's from agencies or from brands or whatever it is. They don't only buy from us. As much as we would love for them to only buy from us, we get that. Marketers are buying ads other places, they're reaching audiences other places as well, and they need to be able to measure across all of us. And so, if you think about those measurement solutions I mentioned and those identity spines, they power a lot of the various measurement vendors that are out there that are using those identity spines. So that interoperability, if we're feeding into that and allowing them to resolve to a single identity that isn't ours but is interoperable with ours, it starts to solve that cross publisher problem.

Kailey Raymond: That's awesome, that makes perfect sense. I can't believe we've gone this long into the conversation without anybody uttering the phrase AI. So I'm just gonna say, I'm gonna get it out of the bag.

Dana McGraw: We did well, we made it this far.

Kailey Raymond: We did okay. 2024 is different than 2023 in that way. I think that people can now wait 10 minutes before saying the word AI, where it was the first thing that you actually said last year. So, progress. But I do know that your teams do work on advanced modeling and machine learning, and so I'm wondering about how you're incorporating some of those tactics into your Adtech solutions.

Dana McGraw: Yeah, so we have a product we call Disney Select, and it's essentially 2000 off-the-shelf audience segments that you could buy based on psychographic, behavioral, you name it, right? And we use extensive machine learning and modeling to create those, to the earlier point about one-to-one versus not one-to-one. And we've been doing that for a really long time. So that's essentially when we started down this path, that's how we started it, with that kind of machine learning, that kind of advanced modeling, and that will continue. We'll advance it, we rework it all the time. Our audience graph, if you think about, is essentially a grandiose neural network of sorts, and all of that modeling comes into play. One of the things that I think is super interesting that we've advanced in the last year and a half, two years in the clean room is the ability to do the data collaboration that we talked about and sort of overlap audiences with the brand, for instance. And then we're able to use machine learning within the clean room environment to lookalike model against the audience overlap to find those who at whatever we set our level of confidence at, who look the most like those high-value audiences but are not currently being reached by that particular advertiser. So that's kind of some of the ways that we're using it.

Dana McGraw: And then, we talked about this in our tech and data showcase a little bit and has gotten more traction than we even anticipated, but something that we're calling Disney's magic words. When we're thinking about kind of context and emotion, if you think about our library of content and all of the different places within our content where an ad might be served. And again, even a more advanced version of right ad, right time, right audience. And so how do we think about the context of a scene, the emotion of a scene? We've talked about it with regard to food. Like whoever the cast is in a restaurant or there's something happening there, and then what is the next ad that gets served that's in the kind of right context for. If you think about emotions, there's a particularly happy scene or a particularly dark scene. What ad should come next and how do consumers react in terms of advertising outcomes to the right ad at the right time? So we're using sort of machine learning, AI-type technology for that.

Kailey Raymond: Wow, it's like the closest you can come to Smell-O-Vision, I guess.

[laughter]

Dana McGraw: Yeah. I hadn't thought about it like that, but yeah, that's what we're going for.

Kailey Raymond: You can use that tagline, you know, to CC me at the back of it if you want to.

Dana McGraw: Got it.

Kailey Raymond: [laughter] That's really interesting. I was talking to the head of data at the New York Times a couple of years ago, and they're doing something similar where they're taking emotion and they're using AI to build models around that to serve the consumers the right audience based off of what they might be reading about and kind of feeling. It's such an interesting way to build context within the environment that you're serving up an ad. So I think it's brilliant and I would imagine increases conversions for advertisers, which is only going to be good for you.

Dana McGraw: Yeah, that's exactly what we're going for and just thinking like, what is the right amount of time between that scene, that particular context, and the ad? And what is the right ad tone type? Do you wanna come out super happy out of a dark scene? Or do you... You know what I mean? What is the right cadence and how do you bridge that gap? So as we're in beta with this, that's all of the kind of stuff we're testing.

Kailey Raymond: I'm curious to see what those combinations are. If it's like sad equals happy ad, or if people wanna dwell on that emotion a little bit. Obviously, Disney is extremely advanced in what they're doing right now as it relates to data management and data collaboration. And I'm wondering if you could share some of the biggest challenges along this journey to build data infrastructure that meet the needs of clients exactly where they are as it relates to making sure that we're putting consumer first, making sure that we're building that trust and maintaining that privacy. What was hard along the way?

Dana McGraw: I think in the beginning there was this notion of cookies are going away, but then it's sort of, are they really? And so...

Kailey Raymond: Yes. [laughter]

Dana McGraw: We were full force in that direction, but bringing people along who weren't quite ready to like, no, but this is the pixel I've always used, which is human nature. There's nothing wrong with that. We all have some of that. Change is hard. In the beginning, it was that. No, this isn't just kind of crazy talking. This is the way this is gonna go and so, come kind of get on board with us. That was the beginning. When we first embarked on thinking about the clean rooms, we thought it would sort of be insights, activation, measurement in that order. What we kind of found was insights were first for sure in collaboration, which makes sense for obvious reasons. Activation and measurement had to happen at the same time because no one wanted to activate the campaign and not be able to measure it, which again, it makes sense. As we advanced and we moved in that direction it was... And then, there are cases where we're being asked for measurement without the other two in the clean room. And so, we had to kind of think of them not linearly anymore. We had to think of them as kind of all happening concurrently. It was a pivot for us, not a particularly difficult one, but a pivot in the way that we were thinking about it.

Dana McGraw: I think the other thing was not every brand, not every agency has an army of data people or engineers or whatever the case may be. Sort of on our side, one, pivoting to be more consultative. Here, we can actually walk you through how to get yourself set up and we'll talk you through it. We'll consult you on how to structure the data. I think, maybe we didn't anticipate how much of that we would do, but we really enjoy it. That's been kind of one of the things that we had to change about how we engage and how do we give something to you that will get you set up quickly. It doesn't have to be the whole universe of what you have, but there's a way to do this relatively quickly. I think those have been some of the pain points. Once we really engaged with the holding companies and the agencies, and once you get one going, then you start moving really quickly.

Dana McGraw: So for us, it was kind of a step function. The first, let's call it five to ten, that was hard. It was really kind of moving mountains and convincing people to get on board, but from there, it was super quick because you start to see what's going on, you start to read about it in the press, and you start to see it wasn't that difficult, and so from there it grew exponentially.

Kailey Raymond: Well yeah, when you're consulting people on the way and you're really positioning yourselves as thought leaders and showing them, I like this concept of breaking it down because I do think that some people can get into the trap when they're thinking about their data infrastructure and how to actually use it. They're thinking about the end use case first as opposed to the little one to start with, and so being able to show them that there is a maturity curve that they're going to have to adapt to and that it's okay to start small and that there are infinite use cases for you to apply as long as you are doing it right I think is an interesting way to go and probably the way to make sure that people feel comfortable and confident in actually getting what they want out of this. And often I think that they're at real-time personalization, everything, and it's like, yeah, that's going to... It can happen. It'll take a bit.

Dana McGraw: It's true. It's like, okay, here's a small use case. That's why we started with insight, right? Here's just a use case where we can tell you more about where your high-value consumers live with us, but we had to make sure that the workload on the brand or the agency for that small use case was not so substantial that it felt like that wasn't enough of a payoff, and so that's where we had to be really consultative and make it much, much easier.

Kailey Raymond: Totally. That ROI immediate, the gratification that they're seeing, you need that dopamine hit if you're investing in a new technology to feel like it was the right thing to do.

[music]

Producer: This Podcast is brought to you by Twilio Segment. According to Twilio Segment's, State of Personalization Report, 69% of businesses are increasing their investment in personalization, and with the world going cookie-less, first-party data is the most reliable, trustworthy, and compliant way to create tailored experiences. Imagine suppressing consumers that just purchase that pair of shoes from your ads in decreasing acquisition costs by 5X or sending the perfect text, promoting your other beauty lines and converting 50% of customers. That's the power of Twilio Segment. Learn what's possible at segment.com.

Kailey Raymond: We're gonna ask you about the namesake of the show. Do you have a definition around what you would consider good data?

Dana McGraw: That's a great question. There's two parts. Good data is the data that the consumer feels comfortable with you having and using. That's part of what good data is. Good data is structured in a way that it's functional, it's replicatable, and it's useful for the best experience for both the marketer and the consumer.

Kailey Raymond: Nice. I like that. I like that you're always starting with the customer. It's something that's really interesting, and I think unique about the Disney approach. I've talked to a lot of people on this show, and more than most, you are front and center. The customer is obviously the number one. The ability to activate, of course, is what I, as a marketer, am keenly interested in, but I need to trust it to begin with. I think that's always where I start. I'm wondering if you have any examples of how your advertisers are using that good data to execute on some of their customer experiences or ad strategies that you'd be willing to share?

Dana McGraw: Sure. There's one use case that comes to mind immediately that we've talked about publicly but with Indeed. It's a great company. They do great work in the job search space, amongst other things. They're one of those that was willing to come along with us early on and be innovative and allow for things to break or maybe not go as well as planned, but they're one of those where we work together collaborative in a clean room environment. It's where we were able to test and do lookalike models on their behalf, and they saw the ROI. They saw the outcomes, and so it's this amazing iterative process with them. It is a wonderfully collaborative partnership between the two of us. What about this idea? What about that idea? They're so respectful too, even with their own brand and how they want to think about the consumer and so it works really beautifully, but we're allowed to innovate and we're allowed to generate these amazing outcomes for them because they trust us, and they're willing to say, okay, we know this one might not work, but let's try it and see if it does. And thus far, things have worked beautifully, but it's that again, the consultative approach, the collaboration, and the trust with one another to understand everything isn't going to work all of the time and that's okay, but we will use that to then optimize the next solution, and I think those are the best relationships in how we work together in data.

Kailey Raymond: That's funny. You two are a match made in heaven. Wow, I used match, and I didn't even mean it, but Indeed, it actually works really well. I talked to their CMO. She was on this show a couple of months ago, and they take a really human approach to their marketing, and so I do feel like, you know, Disney and Indeed are kind of like this beautiful kind of collaborative match made in heaven for this, so it makes perfect sense to me that that would be something that that relationship would be one that would be great. I'm wondering if you have anything that surprised you, that you looked at the data that you gleaned and you said, "Whoa, I wasn't expecting that the data would have showed me that."

Dana McGraw: It's interesting. I think, we think of data science and that it's like math and engineering and that kind of stuff, but it's really kind of a social science. Ultimately, if you think about it. A lot of what we see is a reflection of culture and society as a whole, which is super interesting part of the job for me anyway. And I think we all like to think of ourselves as being tremendously unique and special, but when you start aggregating [laughter] things and grouping things together, sometimes it can be a rude awakening, but no, I'm actually like these 50 million other people. [laughter]

Kailey Raymond: [laughter] I am a sheep, it turns out.

Dana McGraw: Yeah, exactly. It's those kind of insights, but one example was this was last year, going back to basketball, but totally not on purpose. It's a real example. It was...

Kailey Raymond: Sure, Dana, sure. [laughter]

Dana McGraw: There was a show that one would say was certainly a male-targeted, male-directed television or streaming show, and they were gonna have a premiere and they wanted to drive tune-in, and so they wanted to run some ads. So we, again, were able to collaborate and see their typical audience, and then we started looking for where that audience lived, and we're like, you're not going to believe this, but this audience wildly over-indexed for consumption of college women's basketball. And so they literally moved money to some of the women's tournament games last year that were on ESPN, and they drove tune-in to this definitely male-targeted and kind of male-dominated...

Kailey Raymond: I love it.

Dana McGraw: Show, but it's those kind of things. I mean, the kind of data we have, like we talk about the fan of pick a sport, the hockey, NHL. Oh, okay, well, we also know that they over-index for, just examples, driving SUVs or interest in domestic beer, like all of those kinds of things, and you start piecing those things together, but then you see like a wild card somewhere over here, like, wow, okay, those things maybe I would have conjured in my head for this audience, but then you see this other place where they're consuming a particular set of content or a show or another sport. And so, we run a lot of analysis around this particular content, potentially on Hulu, and they over-index for boxing. It's so interesting what you start to learn, but it's not, like two of them, it's millions of them. So then you go back to this idea of, like we're all kind of similar in groups of, you know, in the groups that we do at the end of the day. So that's...

Kailey Raymond: Everybody thinks they're like somebody unique, you know. But...

Dana McGraw: My humbling part of the job, you're like, wow, these people in my group also log into this exactly 2.3 times a week. You know what I mean? It's just really interesting. I wonder if we do it at the exact same time. This is very interesting.

Kailey Raymond: That's really funny. You know, as a woman who identifies as a lesbian, I would say, like, Subaru would make a lot of sense for me if you said that they wanted to advertise for women's basketball, but shocking that a male-dominated show would be interested in advertising to women's basketball. Also thrilled to hear it. That's awesome.

Dana McGraw: Yeah, and, like, listen, they were as surprised as we were, but they trusted us, and they did it, and it worked. Those are the kinds of things that like our whole team really gets so excited about when we start to see those things.

Kailey Raymond: A win, win, win. That's great. Do you have any brands that you admire that you're like, wow, y'all are doing it well?

Dana McGraw: Yeah, there's so many. There really are. If you think about the advancements even in the consumer package goods area, and like they have things down to such a science because of the way that their margins work and all of those things and the way that they think about data and what they do and how they're advancing. If you think about the retailers that are growing into retail media networks and the way that they thought about kind of the data they have and what that means to grow an entirely different line of business for them than the one that they originated. Like, I really admire that kind of innovation, particularly when it's keeping, not to kind of keep reiterating this, but when it's keeping the consumer at the core and they're really innovating based on who their consumer is and so all of their innovation that kind of spawns from that, I find that super interesting and really admirable. Two, because admirable on behalf of whoever's in charge that's letting them take a swing that absolutely might not work, but it might. So there's someone, you know, inside of their business that is trusting people to kind of innovate and maybe fail and fail fast, but innovate and try and create something different for their consumer.

Kailey Raymond: Yeah, totally. I was blown away when I learned that Kroger has like a multi-billion dollar data team or a company spin-off. And I was like, oh, it makes perfect sense to me, now that I know that, but yes, to like these really innovative approaches to making sure that these industries are top.

Dana McGraw: Super innovative, very smart. They do a great job.

Kailey Raymond: Yeah, absolutely. I'm wondering what changes do you see on the horizon in the next 6, 12, or more months as it relates to data collaboration measurement?

Dana McGraw: Yeah, I think more growth in that space. More comfortability around to your point machine learning, AI, those kinds of things and what that means to sort of advance. I think more eyes on outcomes and optimization than ever before, just because, you know, if you think about where a marketing dollar goes and what you're trying to achieve, but for us, too we're trying to be really thoughtful about top of funnel and bottom of funnel, because you don't wanna get so focused on the bottom of the funnel, you lose the top. 'Cause if you're not feeding it, you're gonna be in a lot of trouble eventually, maybe not immediately, but eventually. So, I think a lot of movement towards outcomes into, we talk about it a lot, but I don't know that we're there yet in fully moving ourselves to letting that lead. And then obviously I think, you know, there'll be continued focus on privacy, consumer interest, and certainly changes to come one way or another in that space, which is gonna force us all to be more thoughtful and innovative, which I think is a good thing. I don't think that's a bad thing at all, but I do think that we'll all be more focused on that area and have some things to think about.

Kailey Raymond: Absolutely. I've heard this shift towards outcome-based advertising recently, and I'm curious about it to learn about what you all are thinking about it, what advertisers are thinking about it, the guarantees involved, all that, and it's definitely gonna be an interesting world to see where that lands.

Dana McGraw: Yeah. It's complicated, which again is not necessarily a bad thing, but it's complicated if you were to think about buying against outcomes because the publisher would need to have some visibility, because how could we ever forecast, right? Today, we can enable outcomes, but we're not in that place yet, but certainly, as we drive towards, I don't even want to say real-time, but more real-time optimization. But again, we've got to be really thoughtful about not forgetting to feed the top of the funnel, so we still need to think about intent and awareness and that kind of thing, and not just the end funnel purchase piece, because that can be a little bit short-sighted.

Kailey Raymond: 100%. Absolutely. I agree with you, and as somebody who's worked in B2B their entire career, I think that a lot of the time, often B2B marketers lean almost entirely towards the bottom of the funnel and kind of forget about the top of the funnel, so it's a nice important reminder of yes, you actually need to do all at once to make sure that you are successful in actually building your funnel and converting people through it.

Dana McGraw: We're in a good position as a publisher because of our scale. Our scale is so great that we can really easily see the top-of-the-funnel impact, right? But because we've grown in terms of addressable so much over the last several years, we have better eyes on the bottom of the funnel as well now, and so we're really working ourselves to a place to be able to be consultative about what that balance looks like.

Kailey Raymond: That's a great place to be in. My last question for you before I let you off the hook today, Dana, is what advice would you have for a team that's looking to improve their data collaboration processes?

Dana McGraw: Yeah, it starts with the team that you build. I mean, I said it a couple of times. I could not be more excited and proud of the team that we have, and we've been together for a long time. Our core group has really been at this together for about 10 years, but it's a diversity of skill sets because I think it's easy to have just math folks or just engineering folks that you have on the team, and some of the social science stuff we were just talking about can get lost, you know, no ill intent whatsoever, but our team is, we have someone who came to us as a former FBI analyst. We have someone else that was working as an actuarial and insurance, and we have someone else that was a social media manager, so it's that, starting there, and then really it's the structure of the data, unfortunately. It's the really arduous process of having clean data that kind of matches and makes sense, and then it's just, you know, you talked about it earlier, taking the small step, 'cause sometimes if you look up too high, it's really overwhelming, so you need to connect the dots on the small steps and know how they're going to kind of bridge together, but just kind of one small step at a time, and that tends to snowball into something you didn't think of.

Kailey Raymond: Absolutely. Take it a step at a time, and I love that advice of build diverse teams. Diversity of thought really does impact the outcomes and really builds innovative approaches, so it's always a good reminder to hear that.

Dana McGraw: I think too, like, as a leader, you have to be really self-aware. I know what I'm not good at, and I tend to surround myself with people who are good at the things I'm not good at, and hopefully, vice versa, hopefully, there's some things I'm good at that they need to lean on me for, but I think you have to be really honest with yourself about where your areas of opportunity are and surround yourself with people who can fill in those gaps.

Kailey Raymond: Yeah, no. That's the sign of a good leader, is you know your strengths, you know your areas of opportunity, [laughter] and you know, who might be the person that would be able to really lift that side of your team up, where that gap might be for you. Dana, wealth of information. Thank you so much for sharing all that Disney is doing in the world of data collaboration today. I learned a lot, and I really appreciate you being here.

Dana McGraw: Thank you so much. I'm so happy to have done this. This was a great conversation.

[music]