Good morning, everyone. Good day, good evening, if you're joining us from different time zones. And welcome to our webinar today. We've got a packed group of experts on a pretty interesting topic that's hitting the market in the last five years and developing very quickly. We're going to be talking about some privacy tech and how clean rooms are affecting all of the future of our businesses. And we have some great experts that I'd be happy to introduce for everyone. So maybe we can go around starting with John, and then we'll do Matt and Jen.
Yeah, fantastic, Max. And thanks for those of you that are joining live or watching the recording. My name is John Ayers. Part of Publicis Sapient, really focusing on go-to-market strategy and really thinking about strategic partnerships, which is really what this is about. This session isn't just Sapient, Salesforce, or AWS bringing partners together, I think is what we're seeing more and more of. How do you bring cloud data warehouses together with first-party data? So this is really just a great open discussion. Thanks for joining. I'll pass it over to you, Jen.
Good morning. My name is Jen Bukic, and I have been in this space for over 20 years. My focus in my role at Publicis Group today is really focused on MarTech and how we weave in necessary privacy strategies as it relates to data. Over to you, Matt.
Thank you. My name is Matt Miller. I work for Amazon Web Services. I work on the AWS Customer Data Applications team. And so my focus is on AWS Clean Rooms as well as AWS Entity Resolution.
And I'm Max Kirby. I'll be your host today. I'm a senior technical architect on the Salesforce side. Before we got going, we were all playing the name game. It turns out a lot of us have actually worked with each other in different ways over the years, and we've been kind of circling and orbiting this very important topic as the web itself becomes more ready for a privacy-first world. Before we get into some questions for our panelists today, we're going to get some stage-setting content through here with John just to kind of introduce the topic to you. If you haven't been paying attention to everything going into privacy tech, we'll make this session accessible for you. But if you have been leaning in on this topic, we will have some tidbits based on our direct experience and some in-depth kind of recommendations for how you can think about this and getting some real value out of it. So over to John, and then we'll go into questions for our panelists.
Fantastic, Max. Thanks. So listen, like if you've joined sessions like this before, there's always this, you know, this is kind of a safe harbor statement. So just want to make sure that you're aware that we are talking about topics that are currently in real world, but also some forward-leaning conversations as we open up the discussion. And then we always like to say, you know, thank you. Everybody's got busy schedules. And as I mentioned, whether you're watching this live or recording, we do thank you for the time. I think, you know, really what this comes down to is this next slide. So we just did quick intros, but I would say that, you know, I'll double down on what I touched on earlier. You know, I've been in the MarTech ad tech space for nearly 30 years, been in the Salesforce ecosystem for over a decade. We've seen parallel lines of ad tech and MarTech never to come together, right? We've always heard them as two parallel lines, kind of like railroad tracks. And that's just not the case. Ad tech and MarTech are merging. And I think what we're seeing challenges, there's a completely different vernacular. There's a different language, if you will, of these two environments. And you see that within agencies and you see it within organizations, regardless of industry. People that might be working top of the funnel media, brand awareness, or down to the bottom of the funnel, if you will, first party data and customer engagement. So when you see people like Jen and I from Sapient, we're a leading digital business transformation company, part of Publicis Group, one of the largest media agencies in the world. We're saying, hey, you know, AWS, Salesforce, like we should really collaborate because this is what we're seeing in the market. This is what clients are looking for. They're trying to solve for the challenges. And we've seen this, you know, as Max kind of alluded to, all of us have been in this space a long time and some of you probably joining are seeing the same thing. We've seen this emerge over the years. So that's just kind of want to just set the groundwork of who's on the call and why we think it's important to come together. So I'm not going to go deep into market landscape, but I think, you know, just to set the stage, just to like lay the groundwork here. I mean, listen, we're all talking about omni-channel personalization and it requires data. And this is something, you know, this goes back years, decades that we're trying to deliver a personalized, relevant, contextual message to someone that cares about our brand. And as digital kind of evolved, that had to be more omni-channel. It had to be more real time. It had to be leveraging data. And that's what we've really seen emerge. So companies are seeking solutions across the board. And I think that that's kind of, I guess, the headline here. Now, we've talked about cookies going away. I'm kind of done with the cookie-less phrase. I really refer to it as signal loss. It's signal loss, however you want to interpret that. But this is not coming, you know, because, you know, it's like still out there. This has been going on. So I think signal loss and challenges to bring relevant contextual information to the right audiences in the right channel at the right time, it's been around a long time. And we've seen these challenges, as you can see here. So I don't have to preach to the choir. I think many of you have witnessed that firsthand within your organization. So as I mentioned, I'm part of Sapient. We're part of Publicis Group. We have traditional agency environments. We have digital and technology teams. We manage massive amounts of media. One of the things we look at this is we take the traditional funnel, if you will, and we look at it in what we call a growth loop. So this kind of infinity graphic. And typically, I think what we're seeing is the left side of the growth loop is how do we solve for that signal loss? How do we manage upper to mid funnel media activation when cookies and pixels and signal loss? And we're trying to find the right audiences on the brand awareness. And you can see some of those channels of acquisition and conversion. And that's what we call the find. How do we find the right audiences? And this is something that media has spent decades unpacking and people like Jen and their team really are experts in that space. And then it comes down to if you move from the lower left to the upper left down to the center, that's the real identity. So that becomes a first party data, an identifier, a unified ID, if you will, a golden record, single source of truth. Now I'm going to click one more. So what we've seen here over the years is first party data and Salesforce sits squarely in the middle of their data cloud. They're now being recognized as the leading customer data platform by Gartner and Forrester and others. Basically bringing first party data to the forefront. So all the activations that you see on the right hand side of customer service, email marketing, web, authenticated mobile experiences or loyalty, all of these direct first party data engagements are now having to come together with top of the funnel. So you have left side, top of the funnel, brand awareness, you have first party data insights on the right hand side that what I would call mid to lower funnel, really brand loyalty. Now why is this important? So this is kind of a typical brand marketing funnel. Well, I'll pass it over to Jen here in a minute. But as we've seen, everybody knows this, we have high customer expectations rising. We have massive regulations that have been in place. Whether you agree with them or not, they are real. And we've seen that from I think 2018 was when I first started really wrangling around GDPR. I happen to be based in California in the US. So California Consumer Protection Act, we're seeing these regulations grow. I also happen to have children who are in the Gen Z demographic. So if you want to have a jaded demographic that they just assume the brand knows who they are, just think about where this is going. So there's a couple, I would say forcing functions of why these two worlds of media and first party data have to come together. So at a high level, you're probably are familiar with data clean rooms and we'll get into it when we open it up. But we thought it was important just to kind of level set. So really a data clean room is just allowing this data collaboration in a privacy protected environment. Just to double down on how do we bring disparate data together? And if you think about a global organization that has multiple data sources that might live in, you know, hyperscalers, cloud data warehouses, you know, Amazon Redshift. How do you bring that into like your first party data environment? How do you collaborate? How do you create more robust data sets? This is what we're really going to get into. Jen, I think what I'd like to do is if you could, you know, so much of this is around privacy, about data privacy and security. So I just if you could maybe just touch on a couple of these points. And in fact, I think I saw a question come in. So this might help with that phrase, right, the water cooler conversation. As we were kind of going into COVID, I would say that that was kind of what was in the kind of broad understanding of what data was doing. It was all about data. I think it's really around trust. And I think that's why clean rooms have moved from, I would say kind of what Matt was more niche to becoming essential. They're just essential to create secure, compliant data collaboration. Now, this goes hand in hand with the emergence of a customer data platform as a category or a need or a topic. I think the CDP concept was introduced over a decade ago, but really found its footing, you know, eight, nine years ago. So around 2016, 2017, CDPs emerging, CMOs are saying, hey, my role of traditional CMO is becoming more chief digital officer. I have to work with technology more than I ever had to. CMOs have to understand the digital space more than they. So all of these things are just kind of coming together. And that's why I think it's becoming, it's a product. It has to be something that is used as a service. Yeah. Jen, starting in Nielsen, moving through the agency space, there has to have been an inflection point where we started getting permission to speak about privacy as something that we needed to dedicate time to, as opposed to something that was a disqualifier for things we might've wanted to do or et cetera. When for you, going back in your career, when did you start saying this is actually important enough that it needs to be a discipline in and of itself?
My immediate response to that, Max, would be the Cambridge Analytica situation. I think that was a big moment in our history from a MarTech and a data and an ad tech point of view. And I think that really forced the issue. I know from a measurement perspective, when we were contemplating things like differential privacy or even multi-party computation, those techniques and methods still had constraints, whether it was introducing noise or the inability to dedupe the same impression, that meant that your measurement results were not as precise as they should be. So for me, to answer that, I think it's that Cambridge Analytica moment, which forced the conversation around technique, which then moved us into a solution that can be scaled and repeated.
Yes, an apropos thing to mention in today's world, Jen, that that changed everything and that we're going to have to keep thinking about these things. I think when you look at programmatic media, as it started to be called after the AOL era, and then all of these movements towards, that was certainly a major inflection point. And then after that, we had a bunch of sudden starts, fits and starts of maybe they will, maybe they won't with different platforms. And then Apple came in. And tying together what you said, John, it was interesting when Apple started to pull away the third party cookies. I actually think Wall Street had something to do with this because they started to realize like, wait a minute, this is a huge part. Wall Street wasn't necessarily as up to date on the latest ad tech and martech differential privacy techniques, right? I think they were probably hovering over the surface a bit there. So the budgets, the abilities of clever data scientists, and also people who were trying to exploit different methods all kind of put together led to this term. So we'll bring it up to the present with my next question. We'll start with, Matt, let's start with you again, actually, because I'd love to get your take on this from an AWS perspective, and then we'll round it around. So the next question is, people today, right? These clean room concepts have come up. They're new, they're interesting, but they're starting to get widespread adoption. You're seeing them pop up. So people today using clean rooms for the different data sharing methods that they enable, what do they make possible for the first time, right? What would be impossible without a clean room in place?
Yeah, so I mean, we do see use cases spanning from enrichment to media planning to media activation to measurement. I think out across that spectrum, the one that would not really be possible without clean rooms is typically measurement. I think if we look at a couple of things, I mentioned just the rise of kind of media networks, whether we call them commerce media networks or something else is sort of relevant, but you see TV companies increasingly have addressable audiences, social media companies as well. And somewhere around the turn of the century, probably a little bit before, as those proliferated, marketers flocked there for the obvious benefits of addressability and just like the richness of the data. But increasingly, there wasn't an ability to measure that that combined first party data from both sides, right? And so if you want to do that today, I think increasingly, a clean room is required. If you want to do it in a way that contemplates the join between what I'll just call broadly, campaign exposure and response data. In some cases, you could do it with a caveat conversion API, but what we're seeing increasingly, if you're a bank or an insurer or an HTLS company, you don't want to necessarily use something like that, because that would require you to move your first party data as well as the marketer. So for me, it's going to be measurable. I think it's the most sophisticated. I think there's the most value to be gained at data. That kind of data is the most guarded.
Jen, from a measurement perspective, same question. What's now possible with these things we're calling clean rooms?
Yeah, so to build on what Matt was saying, from a measurement point of view, think about things like the unified view of a consumer journey, right? There are multiple touch points along that journey. And when is that moment in time for conversion? So it really enables you to securely link that first party data with the platform or publisher data. And the clean room really offers that holistic view of how users engage across really paid and owned. So it enables you and affords you the opportunity to create paid and own media strategies, which is really great. And I think the other possibilities could be something along the lines of improved ROI analysis. We all live and die by return on ad spend. So we shift our budget dollars to the right place. I think that one's a little bit kind of easier to wrap our heads around, right? But lastly, I would also suggest incrementality testing. When we want to test and determine from an experimentation point of view, what those lift studies and A-B testing should look like, how do we analyze the control groups without compromising that individual's user privacy? So yeah, I really think that that's to build on those use cases and what's possible.
Yep. And John, I think you've seen, well, we've all lived through, right? The third party cookie, the DMP era, right? And then now the CDP era. To do these things that are now no longer impossible, right? What do you need in place to pull this all together?
Yeah, I'm actually loving this conversation because it's things that we hear from our clients that we work with. You know, if someone's purchased a product or service and they have a customer complaint and they have an open case and that case is working a problem with a particular customer and that is not shared in a data way back up to media, the media spend, maybe spending time and advertising towards audiences who currently have a complaint with the customer. That's a problem. It's wasted ad spend for one. I would say it bruises the brand. Now, if I take that example of a customer service case into other channels of engagement, how many downloads of a coupon? What kind of life events might be happening when I'm thinking about insurance within my family? When I'm thinking about wealth management, every one of these industries have scenarios that's activity happening, what I would call that direct engagement. Email open rate, email click-through rate, call to action, all those engagements. If we're not leveraging those incredibly rich insights with the direct customer engagement, and I often refer to customer, but I want to be finger quote in the air, this can be B2B, direct to consumer, B2B2C, but if we're not leveraging those very informative insights of the direct engagement to inform more media brand awareness, we're really losing. We're losing out on the ROAS that Matt touched on. We're losing out on those insights for programmatic ad buying. As I mentioned, I think it starts to bruise the brand. You mentioned DMPs, the DMP going away, cookies, the signal loss that we're challenged with, the regulations and privacy security that Jen touched on. These are the areas that we're seeing come to bear. I'll just end by saying the marketing attribution, the way that I often think about this is, if a CMO is asked, if you're a CMO on the call, if you're asked, hey, we need you to cut 20, 30% of your budget for next year, and you're not exactly sure what you should be cutting back on, what levers you should be pulling, where you should add more dollars or less, that's a problem. That's the kind of challenges we're seeing in a market. I think the clean room, this AWS clean room data collaboration is really bringing some robust capabilities to bring those data insights together.
I want to double-click on that a bit, John, because I think that the people who we have assembled today probably have been in a lot of alphas, a lot of betas, a lot of pilots. I think we've seen some really magical things happen with clean rooms where we're like, wow, okay, that's really valuable. Then there's those moments when it's just like, womp, right? It's like, okay, well, it's just a different way of messaging. When I think about where to get started on some of these, one of the safest on data sets. And what we can then do, rather than just keep hitting them with more ads to book a trip, we can say, hey, Matt, looking forward to your next trip, and send him cross-sell, up-sell capabilities. This is really where this needs to go, right? This is where this clean room data insight needs to go. Leverage those insights. Jen touched on it. Matt touched on it. To really what I would almost call future state is refined queries in this data, maybe permission-based query sets that you can go into the data sets and really understand. So you're not just bruising the brand again, you're really leveraging it to drive deeper loyalty. That's where I think this is going. I think this is where it has to go. The last piece I'll end, and this is something we're all challenged with. So those of you on the line, you need to be thinking about who owns this technology, who owns these capabilities. I think one of the things around future state, Max, is what are the roles, responsibilities, and skills that we've had five years ago versus five years from now? And I think those are the things that I'm thinking a lot more about. Where does this technology, where do these capabilities sit? Where have they traditionally lived? And where will they be going in the next three to five years? There's definitely a people process technology. I think, Jen, there's also a policy for where this is going. What do you think?
Yeah, I definitely agree with that, Max. I mean, when I think about the shifts that we're going to see, right, the first generation was really relative in using blank force techniques, like simple query blocking, right? And now we're really going to see clean rooms shift towards more brand reputation trust, ease of use, but also from a privacy point of view and just a regulatory perspective, increased automation and data governance. So think about things like leveraging AI for automated policy enforcement, right? And that ability to really quickly and easily detect potential privacy risks. Those things all help protect brand reputation, which is really what keeps almost every single legal team, GDPR, up at night, right? And so I think we're going to see an increase in that automation from an AI point of view, as well as just really federation. And what I mean by that is really facilitating that cross-industry collaboration while maintaining those strict privacy standards that exist for multiple data sets to come in, and it allows that data to stay within a single source environment and still be used for analytics and measurement, as well as personalization. So yeah, totally agree with the process component.
Yes, I think that's, I see the same thing. I know I didn't pull up the crystal ball, but this notion of dynamic privacy between all of us, we're all mentioning it from a tech lens, from a digital business transformation lens, and then ultimately from a policy lens. And I think to tie two things together, I think that to your point, Jen, if we don't have the attribution, our enterprises are not going to necessarily feel that there is a journey where John is saying, this is more than a media problem, it's a customer problem, but right now those channels are things where we need to understand what's happening. And so to Matt's point, if we can solve that behind the scenes, right, if that can become a service, then tech has kind of unlocked more flexible policy. And I think that will, and I think that will self-reinforce and we'll get more and more dynamic. I love the term permission-based queries. Terrific crystal balls, folks. We should all write some insights articles and make sure that there's some people who are, I don't know, starting hedge funds on this topic. So we'll open it up for questions from the audience now. We have about 15 minutes left. So if folks would like to jump in, I do have access to the chat here so we can see what people are asking. I think there was one question that came in that asked Jen to kind of elaborate on what you meant by signal loss. Maybe what we could do is get some concrete examples of a lost signal and then we'll go to the next question.
Sure. So examples of lost signals would be cookies, right? We've all seen cookie deprecation. We have realized that certain legal teams feel like IP addresses are privacy adverse. So we see loss in tracking for IP addresses. We're also seeing a reduction in things like the collection of email addresses or some form of PI based on the way the opt-in or privacy and the consent collection is occurring. So UX implications as well.
Yeah. Here's an interesting one. I think, Jen, we have to double up on you for this one. Do we think we're going to see a federal privacy act that takes it out of California and the other states that have gone first and establishes what in legal terms are called a federal floor? You can answer or not answer if you'd like on this one. I'll give you the option.
I'm happy to answer. I'm one of those that sits in California so you guys can throw darts at me. That's okay. No, listen, an IAPP report recently came out and what we're seeing was I want to say seven new state legislations from a state point of view. And that's introducing more complexity in terms of how each state is mandating consumer privacy. And because of that now, and that was on pace with last year, so if we're going to maintain the pace that we've seen over the last two years, there's going to be another seven to nine states that are going to introduce new laws in 2025. So that is forcing the conversation to be had around a federal policy to help mitigate the fragmentation across the state ecosystem because it's hard. What we're being tasked with state by state is hard here in the US. So I would like to believe and I would like to hope that the right policy makers are involved in what that definition of a policy plan can look like and that our industry is listened to and we have input into that process. I think that would be extremely important for something successful.
You can just imagine the appellate swirl, right? If there's such a non-homogenous set of laws and you can, you know, these big platforms can be, you can change the venue and bring the exact same suit and have a different result. I think the appellate bench would also be begging for something to make sense of this. Personally, I do think we will see one. We'll have to figure out when, but I think it's to the states for now.
I agree.
This is one that might be good for Matt to lead off on. And sorry guys, I can't see who's asking the question, so I would refer to your question, but we've got one that says, who is really doing real use cases with clean rooms? I love this. So we've talked a lot about the theory. We've talked about some use cases, but who's actually doing these? I think Matt probably will redact some customer logos here, but as far as you're willing to go, like what are you seeing out there?
Yeah, definitely. I think we saw TV companies make really the first foray on the supply side into clean room adoption, using it for both, you know, planning and measurement use cases. I think it was really important to help our industry sort of prove out and validate, you know, these use cases are possible. And from there, I think we're seeing now a proliferation of what I'll just broadly call commerce media networks who are now building their ad monetization capabilities on clean rooms, or at least for certain use cases on top of clean rooms. And so I'll be generic as Max said, but we see things like ride sharing companies, airlines with media networks, things like that, where, you know, they're not media businesses like at their core, like they have like core products and services that they sell directly to consumers. But, you know, based on the richness of their first party data, the way that they're going to build their media businesses and collaborate with advertisers and agencies will be cleaner and based.
Great one. John, we have one that I really like. I think you could take it a number of ways here. What are the implications for first party data when it comes to clean rooms? And I'd love to weigh in on this one too, but what do you see?
Well, the first thing I'll, yeah, no, it's great. Just going back on the use case examples, I feel like there's a little bit of leading the witness here. So the unified audience accelerator, which I'll touch on just in a minute that we've developed across AWS, Salesforce and our Sapient team really emerged because we were working with a very large, well-known QSR, quick service restaurant that is leveraging these capabilities. When we started to working with them and really trying to harvest what I would say data insights to really target the appropriate customer audience segments, that really kind of woke me up of that this has to become an accelerator for us to provide the customers, not just within the QSR space, but across industries. So quite honestly, we saw this six, eight months ago with this very large, well-known QSR and we're seeing that there. Going back to your data cloud, first-party data conversation. So I think what we're going to see and I'm all Salesforce all day long, I'm a big fan, I've been in the space a long time. So Max, I'll pass it back over to you. But listen, I think candidly, I think Salesforce a little bit late to the party on the CDP landscape. The reason why I think they've just dominated is because they see the criticality of first-party data. And where I think all of us on this call can really land that is a no-lose proposition is always fight for the customer that you're serving. And why I say that, regulations be damned. Because if we're really in the business to make sure that our customer's data is protected and that we're serving up