Causeway Solutions
ConnectWithCauseway, Media Consumption
November 25, 2024
While consumers are binge-watching content, Netflix, Hulu and other streaming services are keeping their data in silos. How can marketers develop successful ad campaigns? Learn about new media consumption models that will help refine your marketing and advertising programs – including streaming services, TV and social media. Find out how the Causeway Solutions Media Connection supports marketers in key industries, including healthcare, sports, travel, restaurants and more. Listen to the full episode.
Podcast Episode 12, published November 25, 2024
Thérèse Mulvey, Vice President of Strategy: Welcome to Connect with Causeway. I am your host, Thérèse Mulvey, Vice President of Strategy and Insights at Causeway Solutions. Joining me, as always, is Lauren Kornick, our Manager of Strategic Partnerships. Hey, Lauren.
Lauren Kornick, Manager, Strategic Partnerships Hey, Thérèse.
Thérèse: We're also really excited to welcome back our colleague Tim Duer. Tim is our VP of Enterprise Insights, and he is going to talk to us about some new work we're doing to better understand media consumption. Tim and his team have developed some new models that allow us to look at how consumers are likely, and you'll see that that's very important later, watch, stream, subscribe in today's world of endless choices. So thanks for being here, Tim.
Tim Duer, Vice President of Healthcare & Strategy Glad to be back, Thérèse and Lauren.
Thérèse: Tim, let me set the stage. Media consumption is a huge topic, and it's continually growing as our options grow. I started my career in television a million years ago, and it had a very clear way to measure viewership. Since that time, a lot has changed not only in media, but more importantly in data. But the funny thing is we have so much more data than we ever had before, but less really clear insight on viewership and what people are doing in terms of what they're watching.
So, our content options have grown astronomically, but the way you can measure viewership is all over the place. The leading streamers don't share it. What used to be an even playing field in the days of Arbitron, which probably none of you have ever even heard of, and Nielsen are no longer. That is the situation that's bringing us on this journey and what we want to talk about today. The options and the information out there are a really interesting topic. With that, I'm going to turn it over to you, Tim.
Tim: Thanks, Thérèse. I think you hit it on the head. My background is not a million years ago of television, ironically in healthcare, and I've migrated to this place. But in some similar ways, healthcare is also in the same boat where you have lots of bits of information, but it doesn't always come together the way you think it should.
When we're talking about media consumption, you're right, there was more of an even playing field when you had other third party groups measuring data, sharing with others, and everybody could talk about where advertising dollars should go or where viewership was for certain audience segments that were created. Now you have all these platforms that run independently. Whether that platform is a social media platform, whether it's a streaming provider, they're all having countless amounts of data, but they keep it in-house. It's only known to the keeper of platform, which is really becoming a one-sided conversation.
There's something Causeway Solutions has always talked about as being important in democratizing data. I think when the company started 10 years ago, democratizing data meant presenting data in a way that everybody could understand it because there was a big difference between the data nerds and the not nerds, if you will. And that was it.
Thérèse: Exactly.
Tim: Somebody had to help bridge that gap, and that's where Causeway Solutions filled that void. And now I see democratization of data is just making sure how do you fill the gap of those that have the data and those that don't? And you don't have to share, so how can we help companies work together to be on the same playing field?
Lauren, what do you think about that? You've been with the company longer than myself, but I think that democratization of data is kind of a unique role we play, and it seems to be evolving what that means.
Lauren: It's what Thérèse mentioned in her intro. A lot of this information is kept in-house. Nielsen's still around, but it just produces minutes. We used to get breakdowns on ratings for the magic 18 to 49 demographic. We don't even get that anymore, so we have to figure that out on our own. That's where our “likelihood data” comes from, and that's what you mean with democratizing this data. Other people don't have access to this, so we're going to connect our likelihood data to consumers who are just as interested as we are.
Thérèse: When a lot of the streaming service started, they were unique, so you could assume that the viewers of those different things were aligned with whatever. But now they're becoming so mainstream, it's that much more important that you do have those demographics. And who are these people? The Netflix viewer may be almost everybody. But within that, they're really going to be different.
Tim: Another reality is these platforms have changed. When they started, their goal was just to get membership, whether that was a membership to Netflix or somebody being on Facebook. No matter what it was, the goal was to get the count of members to be as high as possible. But somewhere along the way, the game changed. Appropriately, they became more advertising platforms. Well, now you need to share data, but there was never a history of sharing that data. So it really changes the conversation. It changes the intent. So to fill in the blank, that's where we decided to take on a bit of a task.
We've talked about it before when I've been on the show, and I know in other episodes, Lauren and Thérèse, you've talked about it, what Causeway's bread and butter is applying predictive or prescriptive analytics and taking that likelihood that both Lauren and Thérèse have already talked about. Because these aren't exact data points. We don't know all however many million people it is that are on Hulu at any given moment. We don't know, but the platform does. But instead, what we decided to do is take on a task of, could we build models to say, who are the most likely subscribers to these services? Who is most likely to be on Facebook? And without going into details for those that have listened to previous episodes, when you do a predictive model, we're taking a very large sample of 6,000 or more people nationally, asking them, are you on a platform? Is this something you subscribe to? Do you use it regularly? Many different versions of the question.
And then what we do is match that back to the respondent. And then using machine learning, we find a way to figure out what is the likelihood that other people would answer the same way. And that's why as painful as it might be for some listeners, we will constantly be using that “likely” term because we don't know that they're the exact subscriber or user of a platform, but we do know they're most likely to do it. They have the best fit to the model that we've built. It is not an exact data point, but the upside is we've now got the ability to get these groups and connect them together, which is really eye opening when you can start connecting different things. As we get through the conversation today, we can talk about that a bit more. It's challenging because you're not sharing the data between platforms, but the upside is if we can get even a dotted line connection between things, it's a much better state than where things are presently.
Thérèse: So Tim, what is available in terms of the kinds of information that we're building and that we look to have? I know that this is relatively new for us, and I know this is something we're really committed to, so the best is yet to come, but I would be interested in what is our philosophy. What's the most important way to approach it?
Tim: I'm going to cover the big picture version of what's available. And then I'm going to pass it off to Lauren because she is such an entertainment industry consumer. I think she'll be able to do an even better job filling in some of the gaps than I can. To start at the high level of what's available, we built really unique models that look at who is a social media user based on specific platforms, such as Facebook, TikTok, Instagram. We have some of those pieces that are specifically available. But then we went a bit further with social media, including who identifies that they have been influenced by an influencer? Which is interesting because so many things are using an influencer as a direct sale. For social media, it’s not just which platforms do you use or are you high frequency user, but I think something really interesting is that extra piece.
Then when that carries over to the streaming platforms, we did the same with some of the mainstreaming platforms, Hulu, Netflix, Apple TV, Max, Paramount Plus. These are things we built, and then we also did a version of, do you subscribe to the service like this that has ads? Separating who is the most likely to be an ad-based streaming platform subscriber as opposed to the premium package that doesn't have those ads. I have interesting conversations with our clients about how we’re going to use this information. You start to be able to paint a picture of, one, who's a high-frequency social media user versus who just happens to have a Facebook account because those are two very different things.
And then in the world of connected and programmatic TV displays, knowing who is likely to be a streamer only is a lot different than knowing somebody who both streams and they’re watching live TV, as opposed to somebody who's only watching live TV. Saying somebody has an account is really just the tip of the iceberg of the conversation. The important action items become rolling all these modeled universes up and figure out, okay, so now, what can we do with it?
Lauren, I'll pause there and let you fill in some of the cooler stories of what we're seeing from some of these audiences.
Lauren: What's cool about what we have versus what used to be out there in the past or what people are filling in the blanks for today is that you can get age and gender, those kind of common breakdowns, that old 18 to 49 ratings demo. We can use that as comparisons as the baseline, but we can do so much more. We can find so many more extra insights on these likely viewers, kind of like what you were saying, Tim, with the influenced by influencers universe that we've crossed this streaming audience with. For instance, we can guess that someone on Instagram or TikTok might be more likely influenced by an influencer, but what about a streamer?
And one of the things we found is that certain streamers, so think Max or Hulu, or even Max with ads or Hulu with ads, they have just as high of a rate as being influenced by a social media influencer than a regular Instagram user. So that's the kind of insight you're not going to find just anywhere, or even just in the old ways that we used to report ratings or how people watched. This is something that you can only get by the process that we do and all the work that we've put into it and all the algorithms that we've put into it, and all the studies and research and insights that we've put into ourselves.
Tim: One of the cool takeaways is, comparing this back to traditional media first, it's easy to say you should put an ad on Facebook. Or if you're doing Netflix with ads is where you want to be in a streaming platform because they're the most popular services, but there's also a price tag that goes along with that. If you always go to the most obvious choice, there's a price tag with it. And I think Thérèse, as you alluded, you began back more in the TV world. Well, there was a different price tag that went to different shows. You weren't just advertising to be on TV. The reality is everybody could tell you that you should put an ad on the Super Bowl. So, 123 million people watch the Super Bowl. Odds are your audience is watching it. But guess what? There's a price tag that goes along with that.
Thérèse: You think?
Tim: I don't know what it's going for this year, but yeah, it's not a cheap placement. That's why some of these new models we built are pretty interesting because we can now profile an audience that we've created for any client and add this as a new way to view the audience. Did you know they're much more likely than the average American to be on X, or Pinterest, or Reddit, or Paramount Plus. And so maybe a platform that might be a middle of the pack platform from overall membership could be where your specific audience really pops. So now you're going to get much more bang for your buck by aiming there with your placement. And again, I think that's just something that's been lost in the transition from more mass media to more micro media in these digital platforms.
Thérèse: I agree. And one of the things I've always loved that we do is we tell you who not to target. To your point, most people can't afford the Super Bowl and most people don't need to reach everybody on Facebook. That's why it is so important to figure out, where is the person that I am more likely to need? And some of it is logical, but again, it gets back to the exact same conversation at the beginning that we still don't have that data. It's always nice to have something to help you understand where that's going to be, especially things like having these ads. That's a great option for viewers, right? But isn't there a certain type of person who's more likely to say, "Sure, it's worth it to me to spend $10 less a month in order to skip those ads"? That would be important to know if you're an advertiser, that you may have that captive audience by having that information.
Tim: And I think one of the really important things, I've said this over and over again when we've talked to others about these new media propensity models, is it really is an extra layer of identification. It's not where we start. We don't start by saying, "Let's find TikTok users," and then we'll say, "Now, who needs a cardiologist?" It's the other way around of... Bad example, but maybe a great example. I don't know.
Thérèse: Well, you never know.
Tim: You never know. But the point is, we do it the other way around of where I can say... From my healthcare background, I know around 20% of adults in the country we've identified who might be looking for a primary care physician. Well, if we start there and then we narrow it down to those of the people in the Philadelphia market where I live. And then I want to say, now tell me where they are on the audiences because I've already started with my audience I want to find. Now, show me which platforms they're on. Are they on LinkedIn? Are they on TikTok? Where do they over index? And it's really important to use that as the last kind of round of filter, the call to action, not the first thing. You don't want to start and try to work your way backwards because I think that's the easy answer of just tell me people on Instagram. Well, that's not the answer. It's find the audience, and then see where they are.
Thérèse: The other piece that I'm really excited about is the commitment, right? Because almost all data is really interesting over time. If you're me, you think that. Lauren knows how much I'm like, "But I want to trend it." But even in television, a lot of the ratings that were estimated for media buys were based on what happened last year. So same thing with this. Look at how much it's changed in terms of the options, or look at 10 years ago, Facebook was a place where a 30-year-old was. But now it's much more likely to be older people that are on Facebook and some of the other platforms tend to be younger. All of this is going to continue to evolve, and I think that's another piece of why this is so important to be able to see those evolutions, the ebbs and flows of that audience and maybe be able to uncover what some of those drivers are that influence what people are going to choose.
Because as we know, options are just... It's not like that's never going to go away. Newspaper was dead, TV's dead, but ultimately, really all the media is still there. It's just there's more of it. So the more we can understand about this new media, I think the better we can offer that to the marketplace to help them to make those decisions.
Tim: Any models that we build as a company, we try to make sure they all have a shelf life, meaning we're retraining them at no more than every once a year. We're still working out the kinks on this one of how often we're going to retrain, but it's certainly not going to be an annual train. This is probably going to be something we have to reconstruct every six months, or maybe even every quarter because of the constant evolution of these audiences, to your point of some other things we build are much more about your overall belief system. What's your likelihood to hold a certain belief or take an action? Those things tend to be a little more static.
Thérèse: Lauren, I know that you are very much on top of the industry. I'm curious whether you have seen anything or read anything about advertisers asking for more visibility to the streaming services, to the social networks of the world to get this information?
Lauren: Oh yeah. It's been a battle for the last couple of years. Everyone wants to know the basic information of which shows are doing well. Nielsen is trying to get ahead of the game, so they report minutes by week, but they're a month behind. And you have other companies, Samba, Luminate, that are also trying to report on how shows do. We've now got all of the streaming services are reporting their minutes to Nielsen so we can get that weekly report. And Netflix even does their own data drop where they report all of the minutes for their shows for the year. Problem is though, that's all you get. You only get minutes. And we've kind of mentioned this before, that, okay, that's nice to have, but we don't really get any more insight about who's watching Wednesday or who's watching The Boys. You can kind of make some guesses here and there, but you're never really going to know, or you're never really going to know what are their interests they have.
Thérèse: So are you saying that the minutes are just based on, say Netflix in general, not a particular show?
Lauren: They'll break them down by show. So they'll have the normal Nielsen report where it's like, oh, this is the show that won the week, and here's how many minutes we're watched. You don't know who was watching them, but this is what was watched.
Thérèse: Interesting. We talk about how much data we have, but it's really not that helpful if it doesn't tell you what you need to know. More data isn't really the answer, and I think that's where it comes back to what we're doing, right? It's more insight. That's what we're really trying for with this new initiative.
Lauren: Yeah, if I could get different commercials rather than the same truck commercial I get every time I watch Only Murders in the Building, that'd be great.
Thérèse: Really? And do you have anything against trucks?
Lauren: I'm not the demographic to be buying an all-terrain truck.
Thérèse: I understand that.
Tim: It’s interesting you bring that up. For better or for worse, most of these platforms, and this is streaming just as much as social media, they're working with one hand behind their back. An advertiser might be only using some demographics to choose an audience. And I'm not sure what demographic is pushing Lauren pickup trucks, but maybe there's something there.
That's why some of the audiences we build get really interesting, because you start adding those other layers. There is a prescriptive or predictive nature to it, but you're hitting much closer to home when it does hit. And I think that's really important that a platform might say, we're going to push an ad just to an age group or a gender, but they're really missing, what's the motivator?
And so we're talking about these media models, that's really what stands out to me. It's not saying somebody might be a streamer or might be a live TV viewer. The key is who's the audience that's only a streamer? Because if that's my audience I want to reach, I have to make sure I'm putting it into where my audience is. That's where I've got to go, as opposed to maybe I have to go to live TV and I'm going to spend up on a Wheel of Fortune buy because that's where my audience is. And I think that's what's gotten lost in the last 20 years that everything shifting to the digital. It felt like it was going to be much more precise, but I think what we're talking about is it hasn't actually been more precise. There was probably more precision when you bought a show as part of your ad buy.
Thérèse: Yes, you did because you could get... It wasn't just the show. It was the demographic, and it was even sometimes the break. You knew whether or not you wanted to buy a header after Wheel of Fortune. You knew what the next audience was going to be. You knew how it had changed. You knew what it was in March versus November. It's really interesting to think that that long ago, there was so much better data totally on viewership in general than there is now. And I think part of that is just that's how it started, and nobody has really changed it, which is kind of interesting that it hasn't changed more. And it will be interesting to see if this podcast will probably be a catalyst for American Association of Advertisers to get on this topic.
Tim: I think it's interesting that there's been such a change that it was all about measuring click-throughs. And we've often referred to them as vanity metrics. You can measure the thing that can be measured when it comes to a digital. And because of the near real-time nature of it, it's very exciting to see what's doing well. But maybe what got overlooked in all of that is the importance of starting off with a really smart audience, and placing your campaign in just the right place for the right person kind of got lost because it just came measure this click as opposed to looking at the long-term play. You can't see an immediate impact.
Thérèse: Yeah. The whole click-through thing is another soapbox that I have in my closet that I promise I will not take out right now, but we can talk about that some other time.
Tim: While buying a Wheel of Fortune buy versus a football game meant you were at least picking your audience, and that's what's gotten lost in some of the digital. Either there's no picking of the audience and you're just letting the platform run with it, or you're using very specific terms about somebody's digital persona, if you will, and letting Facebook or wherever other platform do the work for you. And maybe that's this conversation where it's all sweeping back to know more about where your consumer truly is going to be as opposed to just letting the platform tell you that they'll do it for you.
Thérèse: It’s data versus insights, right? I think everybody is like, "Well, you have data. That's great." It isn't necessarily great if it doesn't tell you anything. And that's what needs to happen with people saying, "Well, we have all this information that you didn't have before." Okay, but what can I do with it? Now that you've explained exactly what we're doing, are you seeing this as a multi-year commitment? Would we at some point have a view of the media landscape as it evolves?
Tim: It's a heavy question. I think for us as Causeway and what we're going to do, we're going to stay on top of this. And just like you said in the general space of it's not about the data, it's about the insights. And I think that's what we're trying to pull together now too. I will say in the conversations I've had internally with our team, and along with speaking with clients, that there's clearly an appetite for an improved use of data to help us tell the story, help us find smarter audiences, but also help us figure out where to speak to them, and where they're going to be and what they want to see. And I really am excited about that as the future. I think the question becomes how that's going to look in the overall landscape from the measurables. And I know as a company, we've been looking at a lot of different things there.
How do we measure some of the viewership pieces and some of the other missing aspects of the story? But I think this is exciting as a starting point for us to say, well, let's be a bit more predictive about who's likely to be on these platforms. You got to start somewhere. And some of it is accepting the fact that we're going to use that term likely, but I would rather be able to say, with some likelihood, here's what we think it's going to be in it. Yes, there's a probabilistic piece to this, but that seems better than waiting for perfection in this case. So maybe it won't be the exciting breakthrough next week where we've got the way to measure everything in immediate landscape, but I really value the fact that we're starting down this path now to fill in some of these gaps.
Thérèse: We're also committing to being transparent in terms of explaining exactly what it is and what it isn't, and that's really important as well. And I know we are sharing this information so that people can see it and see the kinds of insights we're beginning to get.
On this podcast, we always ask for input. This is a conversation where there could be some really interesting input in terms of what people think about this idea, what people think about this conversation. It is an idea, and it is starting and it's new, but we're open. That's one of the reasons that it was so important to us to talk about this on Connect with Causeway. So, I thank you both. Any parting remarks?
Tim: I'm going to go back to where I started, and I think Thérèse just said it as well, I really like revisiting this idea of what does democratizing data mean in the current landscape?
Democratization of data means a lot of different things to a lot of different people, and that's okay, but the reality is I like being able to serve the purpose of trying to blend it all together, make it more accessible, whatever that means in today's landscape.
Thérèse: Thanks. That's great. Lauren, any last words of wisdom? Any ads that you'd like to see?
Lauren: To echo what we've been talking about, it's still kind of a wild west out there, and we are getting better. I hope that the industry is going to get better at figuring out this information as well. Two years ago, all we really had was survey data on this, so we could give you some insights based on whatever we were surveying about based on those representative people that we surveyed. But now that we have models and universes, we can better define these likely streamers or likely social media users. I can't wait to see a quarter, a year, two years from now, what we're going to have or what the industry is going to have. It's going to be exciting.
Thérèse: Awesome. Well, thank you both for sitting down with me today. It's been a great discussion, and it's obviously an area that has a lot of opportunity for better understanding. As always, thank you guys out there in the world for listening to us. Please feel free to send us your questions, and please be sure to continue to Connect with Causeway. Thanks for being here, everybody.
To learn more, visit Causeway Solutions to get started!