The Daily: Data and targeting strategies for 2025

On today’s podcast episode, we discuss how Google’s cookie choices affect targeting with insights on alternative identifiers, data privacy, and strategies for 2025. Tune in to the discussion with our Principal Analyst and host Yory Wurmser, Chief Brand Officer of CP Skin Health Group Echo Sandburg and Senior Director First Party Data Strategy of the Eli Lilly & Company Moitree Rahman.

Subscribe to the “Behind the Numbers” podcast on Apple Podcasts, Spotify, Pandora, Stitcher, YouTube, Podbean or wherever you listen to podcasts. Follow us on Instagram

Episode Transcript:

Marcus Johnson (00:00):

eMarketer is your trusted partner if you would like for actionable data and insights on marketing, advertising, commerce, and more if that's possible. But did you know eMarketer also has a division focused on B2B media solutions? "What," I hear you say. It's true. You can partner with eMarketer today or tomorrow, we'll still be here, and connect your brand messaging with our powerful audiences. You can head to emarketer.com/advertise to learn too much more.

Moitree Rahman (00:35):

It's not just one, not just website or desktop or mobile. There are a variety of social media sites and channels are there. So we're really double clicking on all of them to understand the full portfolio for channel strategy and making the right decision based on that.

Marcus Johnson (00:56):

Hey gang, it's Tuesday, November 26th. Welcome to this special edition episode of the Behind the Numbers Daily, an eMarketer podcast. In this episode, eMarketer principal analyst Yory Wurmser hosts a panel about data and targeting strategies for 2025 from the November 1st eMarketer Summit. This panel features Echo Sandberg, Chief Brand Officer of CP Skin Health Group and Moitree Rahman, Senior Director First Party Data Strategy of the Eli Lilly and Company. Yory and our expert guests discuss how Google's cookie choices affect targeting with insights on alternative identifiers, data privacy, and strategies for 2025. Enjoy.

Yory Wurmser (01:44):

Hi everyone. Thank you for joining us today for Indestructible Data and Targeting Strategies for 2025. I'm Yory Wurmser and I'm principal analyst at eMarketer covering media, advertising, and technology. We recently published the results of a survey we conducted with SNAP in July and one stat caught my eye. 61.4% of marketers in the survey said that they were looking to augment or measurement strategies with better and faster media mix modeling. We'll cover that in the next 30 minutes along with Google's cookie non-deprecation and strategies for targeting and measurement for 2025.

(02:20):

Joining me are Echo Sandberg, Chief Brand Officer, CP Skin Health Group at Colgate-Palmolive in Scottsdale, Arizona, and Moitree Rahman, Senior Director of First Party Data Strategy Lilly USA in New York City. Before we begin, let us know in the chat where you're tuning in from. Also, if you have questions, please feel free to drop them in throughout the session. Our emcees, Susie and Jeremy along with a special guest will return after and take your question. So let's get started. Thanks Moitree and Echo for joining us today. I'm going to just start with the most broad question, which is how does Google's recent clemency for cookies affect the data ecosystem? Does it really make a big difference?

Echo Sandburg (03:07):

Hi Yory, thank you so much for having us today. I would say that with Google's recent decision to no longer prioritize rolling back cookies, it does relieve some of the concerns regarding the immediate future, but ultimately it doesn't change too much for us. We've been preparing for this cookie-less future working off of the perspective that identity is the new marketing currency and investing in a people-centric world identity strategy so that we can thrive in that potential cookie-less world. And this has involved us working to build this comprehensive views of our users through zero-party data like brand surveys, first-party data like our CRM database and even second and third-party sources. And so now I would say that we'll just continue to develop, grow, and maximize our 1P identity and data assets, further leverage 1P unknown, all without losing out on our current cookie-based targeting options.

Yory Wurmser (04:12):

Yeah, it makes a lot of sense. Moitree, have you had sort of a similar approach?

Moitree Rahman (04:16):

Yes. And thank you again, Yory and the eMarketer team to having us. Really a pleasure to talk about all the things that's going on in the advertising world from the data perspective. So this discussion has been going on for a few years. [inaudible 00:04:31] do it? Is it going away? It's not going away. I think in a situation like that, it does give us the opportunity to really think out of the box and really innovate at this at that moment.

(04:42):

So we had been very diligently thinking about reducing our dependency on external identifiers and frankly the things that are probably not durable and not quite credible in all aspects. So we were thinking what are the things that we can actually build and innovate on our side that is going to be actually more credible in terms of building that relationship with our customers and we're not depending on external cookies and identifiers. So with that, we have invested so much of our effort in building first-party data strategy, which I'm leaning at Lily. So... and again, whether it goes away tomorrow or not, looking at our comprehensive data strategy more from an internal perspective has been the key driver for us to have a more innovative mindset. And of course this will complement with whatever we have externally in the market because that does impact our digital targeting strategy and all of that. So we are not losing sight of that completely, but it is more about how we can have a more robust strategy internally that complements with external identifiers is really the key for us.

Yory Wurmser (05:58):

Have you seen that translate concretely into the external targeting strategy?

Moitree Rahman (06:03):

Yes, for sure. One of the things that has... Because we have now more robust view of our first party data, we kind of have more insight into our customer groups. So taking that insights and translating that in our external targeting strategy has also been a key innovation I would say because we're not necessarily depending on probabilistic models and all of that, but now we know more about our own customers through our own direct relationships with them. So that definitely has enriched our strategy when we think about targeting. So that's how I'd say it's been nicely translating with what we are seeing on our end and how we are quickly demonstrating that in our external targeting strategies.

Yory Wurmser (06:52):

How have customer data platforms helped you adapt? What has their role been?

Moitree Rahman (06:58):

Yeah, so customer data platforms have been a very critical add to our Martech ecosystem, especially when you think about in today's world where privacy is paramount whenever you are dealing with customer data. So we had been looking into expanding our Martech system by bringing in really those tools that are what we call privacy by design. And that really allows us to bring several different sets of data in one place.

(07:29):

So we are not looking at our patients or our customers from a disjointed view, rather we are bringing in that digital view, offline view, the business outcome data and all of that in one central place. So we know that what are the different stages that they're moving in from, connecting the data from disparate places in one place and central place I'd say. So customer data platform has been a key part of that. It has also helped us in how we shape our personalization strategy when we are taking the data back and designing their journey whenever they're interacting with us through our website, through our different platforms, through variety of digital touch points. And CDP has been an instrumental part of this building audience strategy, helping us orchestrating those personalization tactics and all of that. So I would say it has definitely been a very big part of this changing landscape and a critical part of our Martech ecosystem.

Echo Sandburg (08:33):

Yeah, I completely agree, Moitree, with what you're saying. And we've really seen that CDPs have offered that ability to unify the data across touch points as you're saying, and really centralize it all into one platform. As the capabilities evolve in this space, we can better align our messaging with each of our consumers or customer's journey across email, SMS, web, improving that overall engagement we're delivering by giving those relevant and personalized messages and content. Another space for us, it's really redefining how we review our own media audiences. We're able to redefine our targeting to factor our messaging and our offerings to fit what people are truly needing, whether it be skin care concern or a specific item that they need to complete their skin regimen.

Yory Wurmser (09:22):

Makes sense. I know these more robust CDPs can also help in creating these MMM platforms, the media mix models. How do you foresee your use of MMM evolving in the next couple of years?

Echo Sandburg (09:39):

Yeah, I would say we definitely see our use of MMMs evolving. Today, we have this practice in place that we're using across each of our distinct brands really to help us better understand the impact of our marketing actions and what is driving the most effectiveness and efficiency, which I believe it's important to understand. It's not just efficiencies, it's the effectiveness of the spends that we have.

(10:03):

So we're specifically using that data for high level media planning and budgeting. It is of course, critical to supplement this work with more modern ways of understanding our marketing mix, particularly with digital and support that with more real-time learnings, with multi-touch attribution, which gives us that ability to help measure in much greater detail like how our targeting strategies or a creative is performing. But even with that, those methods don't necessarily allow us to see that bigger picture and the impact from other efforts or if we didn't have that marketing tactic in place. But I do think it's exciting when you think about machine learning and AI processes that are really paving the way for what a new era of attribution could look like for really that full customer journey. And that's for us a quite exciting space to further explore.

Yory Wurmser (10:59):

Yeah. For sure. And as part of that move to MMMs, I know that some of the larger platforms offer some MMM offerings as well. I think your companies are probably big enough that you have your own way of measuring all of that, but are you finding that first party data in some of these larger platforms make them more attractive as places to advertise?

Moitree Rahman (11:24):

Yeah, so for first party data, the availability on our partners side is also very critical. So we know that there is a richness in that data, and it's not just some probabilistic models we have, but we know that our first party audiences and they have a clear interest as well as availability and scalability in the external ad platforms that we are seeing. So that is definitely a ripe place for us to look at as we're building the audience strategies. MMS is a very robust tool within our measurement toolkit for sure, which does help us understanding which channel is really performing well and how can we rethink about in terms of our budget strategy, the channel effective strategy, and also now adding on top of it our audience tactics. So that is a very important part of the whole equation. It's not just how the channels are performing, but are we able to leverage other variables within our first party audiences that is going to actually drive a higher effectiveness?

(12:33):

So both of these are important part of the equation. One is when we are looking at building the audiences from a targeting perspective, are we getting the richness and the robustness in our partners' platforms? Are there intersectionality between those two datasets that we have about our patients, the things that we know as well as the publisher partners and how they are also tracking the interest and engagement on their side of the same group of audiences? And is there a nice little happy metal between those two and which publishers are actually really garnering the highest ROI?

(13:12):

So those are the things we look at in MMS. And of course really good way for us to actually assess the effectiveness and looking at our channel strategy, they help us quite a bit. They're also becoming very robust in terms of how we're incorporating digital touch points as well. So the previous world was much more about linear TV versus digital, but right now we're also looking at, "Okay, within the digital there are a variety of channels." It's not just one, not just website or desktop or mobile. There are a variety of social media sites and channels are there. So we're really double clicking on all of them to understand the full portfolio for our channel strategy and making the right decision based on that.

Echo Sandburg (13:56):

Yeah, I agree. I touched on this a little bit before, but really it's a combination here and leveraging that MMM data heavily to initially decide where and how we would want to participate in different channels and also leveraging specific measurement and verification to ensure that we're helping to improve and we bring in our agency partners here to help us improve the sources of data verification and our measurement to let us manage our ad strategies down to smaller detail events and really prioritize how we can report on it and compare our data to benchmark data and performance to make sure that we're getting the most out of our investments.

Yory Wurmser (14:35):

Yeah, I mean MMM is... notorious is probably the wrong word, but it is very high level and it seems to me that you're able to... because you're developing, you're able to go down to more mid-level type of analysis.

Echo Sandburg (14:47):

Absolutely.

Yory Wurmser (14:48):

So obviously a lot of this relies on first party data. MMM has a nice advantage in that it's aggregated, but still first party data is becoming more important. How are you ensuring that you're protecting privacy and getting consumer buy-in to use all this data?

Echo Sandburg (15:06):

Yeah, it's important. Brands of today and tomorrow are built through the delivery of personalized consumer experiences. So we know that in today's marketing age, we really have to master that rapidly changing approaches to acquiring, retaining, managing, and mining that valuable data that's going to help inform those meaningful experiences that we're all looking to create. And this has to be done in real time and it has to be done in a privacy-safe manner. So it's super easy, but it's important for Colgate as a broader company at CP Skin Health, we absolutely take the privacy aspect very seriously. We know that it's important to treat consumers data with integrity, and that includes proper consent gathering, proper data handling, all the way through from capture storage through activation. And so further to that point, it is important to be very transparent. Part of how we seek to do that is really clearly communicating about the data that we're going to use, offering very simple opt-in or opt-out mechanisms and for our privacy policies to be readily available. So we have that on sign-up forms or the website that our customers can easily access.

Moitree Rahman (16:26):

Yeah, I would add on top of that is that transparency piece is so critical because oftentimes that can become like, "Okay, I don't know how they're using the data, but if we give the control back to our consumers and in terms of opting in and opting out, if there are certain communications they do not want to participate in whether it's email communications or maybe not being tracked anymore and they have the opportunity to actually opt out, that takes the control back to them." So we are not necessarily... just once the data is in our hand, we're not necessarily just using it, but also allowing them to not give us the permission to maybe not contact for certain things for certain programs and brands. And I think that is very important when we want to establish that credibility and transparency with our consumers. And for that one, adhering to the privacy compliance is also very critical.

(17:27):

Having the right consent architecture and control architecture around using the data is actually very, very critical. And as we are looking into building our own data architecture behind the scene, the consent piece is actually which drives how we're activating. So it's not like it is consent is sitting in one place where probably yep, we're just tracking box, but it is actually where we start. We need to make sure that we are looking at all of our audience strategy based on people who have given us permissions. It's not the other way around. And that is also elevating some of those tensions around how the data usage and ensuring privacy compliance is in place and also customer have the trust in the brand because that is very key for us.

Yory Wurmser (18:17):

I mean, it's not just getting permission to use that data, it's also just being able to access all the data that you have. And I know that many, especially the bigger an organization is, the more siloed sometimes that data becomes. How are you trying to remove those silos and really bring together the full force of the knowledge you have about your customers and about how they're interacting with you?

Moitree Rahman (18:39):

Yeah. So it goes back to the different channels where we collect the data. And oftentimes if you are only looking at channel strategy, not consumer strategy, the silos happen because essentially you start looking at specific platform, specific brands and programs, but essentially this is one customer right? You don't want to look at the customer from different angles because disparate data is going to lead into a disparate or fragmented view of customers and that will lead broken experience and you don't want to do that. So we want to make sure wherever there is silo and wherever there is a consent, that is, again, I'll probably say that more because I want to make sure we have the right permission to actually bring the data in. So if someone who has given... opted in on our site, have accepted whatever the first party data agreements we have on our website, we have the consent that yes, we can actually store and use their data to enhance their experience.

(19:46):

We also want to be very transparent into what specific use cases their data will be used. And once we have the consent in the right place, that gives us the permission to actually relook at and reevaluate our data architecture on the behind the scene. And that's where we have done quite a bit of work over the past year or so to bring the data in one place and create an integrated view of our customers. The enterprise data lakes helps us quite a bit. So as opposed to looking at brand specific view of customers, we are shifting into more enterprise data strategy and enterprise wide market decision. So that allows us to do those data joining wherever the permissions are in place and we have the right control architectures in order to make sure that we are looking at the joint and integrated view of our customers.

(20:39):

And they do not live in silos because again, going back to data silos, leading to fragmented strategy, to broken experiences is where we want to make sure that we are actually using... making a responsible use of data by bringing the data in one place. CDPs, as mentioned before, is a very good tool for that very purpose as well. But before the data can get used through CDPs, there are a lot of other upstream work that we need to do to remove a barrier that currently exists between the different datasets and data groups. But essentially that is really where we are starting off. And also promoting a culture of data sharing and collaboration is also very important, but making sure we have the right consent and control architecture in place so that not everybody has access to that data, but it's really protected in an environment and then also used for the right purpose and it should be connected back to enhancing the customer experience.

Yory Wurmser (21:41):

And what role do things like data lakes, data seas, data oceans play in all of this?

Moitree Rahman (21:48):

Yeah. So going back to the earlier point about where do we actually bring that data in because it's not just data living in a certain system, but having building that enterprise data warehouse where everything is coming in and having a really robust identity map is critical because it's not just like some data sitting side by side and magically you are going to be seeing everybody like all the views coming together. You have to have their robust identity strategy in place to enable that even within a data lake architecture. So that's how we're looking at disparate datasets and then joining them through a common identity structure, bringing the consent architecture, bringing all of them together in order to enable that common view of customers.

Echo Sandburg (22:37):

And I think it's important if you are working in an organization that's multi-brand or enterprise level to think about your tech stack strategy and your approach there so that you can build that, those opportunities to have unification and consistency across the work that you're doing that Moitree was speaking about.

Yory Wurmser (22:54):

Speaking of tech stack and technology and no conversation in marketing can go without talking about generative AI. How is generative AI affecting how you're approaching analytics and targeting these days?

Moitree Rahman (23:08):

Yeah. So generative AI, I mean it is becoming... it is embedded in all parts of the decision making. So when we think specifically about from a data and analytics perspective, a key thing for us is how can we get the insight faster? And when we think about the variety of data we currently get, whether it's structured, unstructured, whether it's like text, whether it's qualitative data, quantitative, the variety is so much is that it's hard for us to actually get to that insight at a speed to respond to our customer demand. So what we are looking at is that how can we embed this generative AI capabilities with large language models for us to actually sift through that at a rapid speed so we can actually get the insight faster, run experimentations faster, and see where there's an opportunity for incrementality in business outcome. So that's how we are experimenting quite a bit in the data space as we're looking into dealing with variety of data. Data quality can be an issue as well where we're seeing how these newer advanced capabilities can help us get to decision faster.

Echo Sandburg (24:25):

Yeah, I think this is a really exciting space in general. Part of our organization at the enterprise level, it's a space that we're really leaning into and our approach by and large has been on building an internal AI suite that we can be comfortable and a safe place for us to all work in. And it's really been about continuing to iterate and develop and test and learn what can be done with these tools and the power of these tools. It's really, really been an exciting space and I can't wait to see the future when you have the right parameters in place to unlock the ability and the future potential with AI.

Yory Wurmser (25:07):

Yeah, for sure. And I mean, one of the things I've noticed in my own research in this area is just that it just opens up the analytics to a broader group of people just by making much simpler just to interact with these datasets. And also, as you mentioned, the increased speed of insights. You could start doing things like more iterative incrementality studies, faster analysis on these really complex new models. So a lot of options are opened up.

Echo Sandburg (25:37):

So many options and clearly a lot of oversight needed, but the ability to get from A to B much quicker and then to work with that data in a different way, it allows us to spend our time in the right places.

Yory Wurmser (25:48):

Yeah, for sure, for sure. I'm sure we're all going to be talking a lot more about generative AI in the next year. One area that I've noticed some impact from generative AI is in contextual advertising and contextual targeting. What have you seen in this area? Has contextual targeting actually moved forward as a result of some of these innovations? Are you seeing it as a more appealing option for your companies?

Echo Sandburg (26:11):

Yeah, I mean, first of all, we like to keep that mentality of test and learn and maybe what worked in the past, what worked today and vice versa. But we have been bringing more and more contextual advertising into our marketing mix, and we are starting to see some positive results in that space. For us, we're leveraging it in places through our partnerships like a partnership with consumer publications. And that really plays an important role in driving attention and consideration. And we're ensuring that we're choosing partners that align with our target audiences to avoid some potential pitfalls that could be around the space of contextual. And when we think about how this could be further innovated, there's been some examples that we've used with our audiences, with sponsorships that have helped us increase visibility through placements in areas like sunscreen portals for our SunCentric brand LTMD that we're going to see great responses to.

(27:13):

And we'll definitely be leveraging newer opportunities as we head into 2025 to take these efforts to the next level with things like interactive native articles, shoppable ads and units and quizzes that can be placed alongside relevant content. And if you think about why is it not more widespread, potentially it is less accurate targeting, you're not using that personalized data, so it may not be as effective in reaching your ideal consumers. However, I would say with privacy and cookie concerns in this ever evolving space, I believe that it's worth marketers to continue to explore as part of their broader media strategy.

Yory Wurmser (27:56):

Yeah, I'm really curious to see its impact on brand safety, both in terms of negative targeting so you have a better sense of what not to target against, but also targeting places that are blocked for brand safety reasons now but actually it could be good places to advertise. I think that's one area where generative AI is going to really help with that. I think, why don't we wrap up here? Echo and Moitree, thanks again for a really great conversation. I really appreciated the time that we just had and the conversation we just had.

Marcus Johnson (28:31):

Thank you for listening to this special edition episode of the Behind the Numbers Daily, an eMarketer podcast. Tune in tomorrow for the Reimagining Retail Show where host Sarah Lebow will be discussing this month's November retailer rankings.