 All right. Good morning. Hello, everyone, and welcome to our next EDW session called NBC Universal Parks and Resorts Cloud Success Story, which will be presented by Lea VanZelm and Anchor Jane. Excuse me. Lea is the VP of Analytics and Targeting at Universal Parks and Resorts. Anchor is the SVP of Strategy Lead of Cloud Platforms at Merkle. And all audience members are muted during these sessions. So please submit your questions in the Q&A window on the right of your screen, and our speakers will respond to as many questions as possible at the end of the talk. So let's begin our presentation now. Thank you and welcome Lea and Anchor. Thank you, John. I really appreciate it. Good morning. Good afternoon. Good evening, everyone, depending on wherever you're joining from. My name is Anchor Jane. I lead the global cloud practice at a company called Merkle. When you think about Merkle, think of us as a customer experience transformation company. We help the best brands in the world create competitive advantage through people-based marketing. We believe in marketing to people, not to proxies. And we believe the future of marketing is personal, informed by data, powered by technology, and delivered through creativity. Now I'll bring Lea VanZelm, my colleague, to introduce herself as well, please. Hi, everybody. Thank you for joining us today. I've had the pleasure of working with Anchor and his team for a number of years now. And you guys all have heard of Universal Studios, I'm sure. The Universal Studios lot was founded in 1910 and followed by the park in Hollywood. And the location is the largest location. We do have a global presence with parks in Singapore and Japan. We're building out one in Beijing as well to open later this summer. As you guys know, we offer very immersive, exciting, fun adventures for families and couples alike. We're very data-driven, both in terms of the storytelling that our attractions offer, as well as the experience that we're building before, during, and after the trip. Therefore, the data-driven, tech-enabled personalization has been critical to us and advancing our marketing and a big effort that we've built with Michael over the last year. Awesome. Thank you, Lea. So just a little bit of background setting here. As you might have heard, John speaking about the title of the talk here, the webinar here. It's definitely centered around cloud. So when we talk about cloud, I want to just clarify that we are not talking about any software as a service type of an application, which could be Salesforce or Adobe or SAP when we talk about marketing. Those are the type of applications that come to mind. Instead, what we're going to focus today is about cloud computing resources. Think about storage, think about compute, think about AI, ML. Those type of resources which are offered to you through a browser over the internet. So those are the cloud computing technologies we're going to be focusing on, such as Google Cloud Platform or Amazon Web Services or Microsoft Azure. So that's going to be the focus of the conversation in light of how NBCUniversal leveraged these cloud computing platform technologies to shape up the experience of their end customers. So I hope that gives you a little bit of context. And now the number one question one may be wondering is that why are we even talking about cloud? What is so important about cloud or why is cloud relevant in today's market? So this is my favorite slide, which I kind of talk about, which is talking about the value proposition of public cloud platforms. So number one value proposition that I believe that cloud platforms bring to the table, which definitely is relevant in today's digital age, is the agility. The agility that it brings to your business, the agility that it brings to your development initiative, the agility that lets you spend less time on the infrastructure alone. Think about scenarios where you have to buy hardware, software from different vendors and install it. Essentially with public cloud computing platforms, you don't have to buy any hardware or software. These are all available to you over the cloud platform itself. So focus more on building the solution rather than managing or procuring infrastructure underneath. The second point here is the cost. Obviously cost becomes a huge play when we are talking about setting up complex global nature type of solutions. So with cloud platforms, basically you pay as you go. That's the kind of cost model that these cloud platforms bring to the table. And it shifts your pricing model from CAPEX to OPEX. So meaning you can start small and continue to grow. So that's another big aspect from cost perspective that why public cloud platforms are becoming so appealing. Security is another big aspect. A lot of times people wonder whether my data would be secure if I were to move all this data into public cloud platforms. That is yes, the cloud computing platforms have become so mature. They offer a lot of encryption kind of techniques. So you can encrypt data both at rest as well as in motion. If you are worried about regulatory compliance such as HIPAA or PCI or any other FedRAMP in public space, these public cloud platforms are actually regulatory compliant as well. So a lot to cover on the security aspect. And then the scalability, like I said, you can start small and you can grow over a period of time. Performance boost you can expect because you can keep adding more and more resources if you are seeing performance is not up to par. So essentially it gives you the avenue of starting with some sort of a minimum viable product. And then keep adding functionalities and grow the functionalities over a period of time as you grow your application and solution in a true global sense. And then reliability is another big one, which is how public cloud platforms kind of give you inbuilt disaster recovery or high availability or the state of the art infrastructure that they bring to the table across the globe. So your applications are much more reliable, much more secure, much more cost effective and much more agile. So that would be my, you know, five point value proposition or six point value proposition that I usually bring to the table just to ground everyone that why are we talking about cloud, why cloud was attractive to NBC Universal and why we as here at Merkle are so passionate about public cloud platforms. So with that kind of a background, I would like to bring Leah back into the discussion here to talk a little bit about what challenges, what was the business need, you know, from NBC Universal, which prompted them to look at cloud and bring Merkle in. Thank you, Ankur. Yeah, I'm going to talk a little bit about what our business need overall. And how we use data and then talk a little bit about where we you know, our infrastructure challenges, you know, that led us to this cloud solution are. So as you know, we have a park, there's actually three parks here in Orlando and parks around the world. We also have hotels on property in some of the locations and so, you know, our goal is to always identify the highest value guests who kind of fit with with our brand and will really have the opportunity to enjoy our brand by visiting and it's our business goal is to drive attendance and drive sales, drive, drive bookings of hotels, longer stays, larger parties. And the way we do that from a marketing and sales perspective is of course identifying prospective guests who are for travel and have a propensity for theme parks and then deliver delivering a marketing and experience that is persuasive enough and personalized enough that they decide to book before they get to Orlando and they're really committing to the to the travel to this day. It's not an easy business, though, because there's lots of competition for discretionary spending, right? It's not just the other theme parks. It's also the beaches or staying home and doing stuff at home. So, you know, we really believe that that the experience and immersing the guest in our experience before they come is part of the value proposition. For your marketing space, you know that there are changes all the time with compliance requirements both from governmental entities from platforms like Apple and browsers like Chrome. So we're also very cautious with how we use secure and comply with privacy requirements related to data. You can go to the next slide here. We have, I think, what many of you have probably experienced or maybe your internal clients and partners have experienced, which is we started with an email database. And, you know, that was our CRM platform and over time that's that need for data and personalization across channels has changed what we need from our data platform. So now we've got the need to tie anonymous website behavior with email to media advertising activity and then follow that person through their visit so that we can really bring in all of their preferences or an unstated preferences to their in-park experience. But when you start with an email database and then you add on a website analytics data platform and media platforms and so on, you know, you end up with these silos of systems and data platforms, reporting platforms. So we really saw the need to modernize and integrate all of that data so that we can get the true 360-degree view of the guest and know how they're engaging with us across different channels, able to not only interact with them, but also analyze our performance or marketing based on those different interactions. Given that cloud platforms are newer and the marketing technology ecosystem is constantly evolving, it's very hard to stay on top of that, you know, being on in the client side or in industry because it is changing so fast. And you can imagine for us, especially we've got so many attractions and ticket booths and gates to keep open. All of those operational needs are very intensive from a technology perspective. So we turned to Merkel as a partner to help consider where we want to go with the marketing experience. What's the most efficient way to get there? We started on this effort with some very small wins in mind and we'll talk you through that because we took a very, an approach of building for now, proving out a use case and then building on that use case. And it was the best thing we could have done because we finished up some of the work just before COVID forced the shutdown of parks. And, you know, we had a stake in the ground. We had proven out our first use case at that point and we're able to then build on it during the, you know, very difficult last year that all of us have experienced. Awesome. Thank you, Lea. So as Lea mentioned, you know, Universal Parks and Resorts, you know, they brought in Merkel to kind of help them and Merkel is not new to Universal. You know, we've partnered with Lea over a number of years so far. So definitely the partnership between Merkel and Universal is something not new which started with this project. It evolved over a long period of time as well. So the approach that Merkel took to kind of help with the challenges and the business need that Lea described was number one, you know, we lead with cloud. We feel that cloud should be the underlying platform to give you all the benefits that I talked about, the agility, the cost effectiveness, the reliability, the robustness and the performance and everything. So any type of solution that we are, you know, bringing to the table, any offering that we are bringing to the table here from Merkel's standpoint, we lead with cloud. It was inherent choice as the underlying platform to build the solution for NBC Universal. And then the second point that Lea was describing, the approach that we took was that let's build the solution for now, meaning let's not boil the ocean, let's not create, you know, or think about futuristic needs. Absolutely, the architecture should be scalable for future. It doesn't mean that we're going to create, you know, some sort of a technical debt by building a point in time solution. But let's focus on what sort of outcomes we want to achieve now and then create room in the architecture, use the technologies, use the platforms in such a way that they can scale for future needs if need be. Meaning like, for example, we started the solution with Hollywood. That was our MVP. We failed the solution to bring in Orlando. That was the further extension. Now we can take this solution and bring on international parks such as from Beijing or from Tokyo and so on. So that's the kind of scalability this platform kind of provides. And then, you know, for complex global projects of such nature which cut across multiple service lines, absolutely there will be some sort of course correction. We cannot expect that everything will go as planned. So the whole idea is that the methodology should be that if we need to adapt, if we need to do some course correction, let us be agile enough, let us be nimble enough to kind of bring in that sort of a, you know, course correction into the play here. So that's again a core, you know, tenet that we bring to the table while thinking in terms of building solutions like these. And then keep the end customer in mind, meaning who is going to be the consumer of this data? Who is going to benefit the most, which is the business user? Meaning people who are making those business decisions, let's keep them their ease of use in mind while designing the solutions. We can bring in a lot of technology because technology is cool. You know, it gives you a certain type of, you know, a pleasure in designing, you know, a very sophisticated solution. But then at the end of the day, if business needs are not satisfied, then that's not the right way of designing solutions. So we truly believe that technology is an enabler. But then let's keep in mind the business goals, why we are building and how we are building this solution. And number six is connect all the data, you know, because as I said, you know, this data cuts across many different business lines. So the whole goal was that let's not, you know, create another yet another silo yet try to build the integration point so that when we talk about customer, we should know what that particular customer used what services. Those services were consumed in which park and watch fashion and everything. So basically, they should be a fabric, you know, let's create a fabric which stitches together all those data points so that the business users are able to, you know, make some sense out of that data itself. So that was the underlying approach that we brought to the table. These are some core architectural tenants that we believe in while designing a solution like this. So to cut through the chase, this is basically a high level view of what that solution actually looked like. And yes, I know the picture is, you know, this could be an I chart, but let me walk you through what this platform actually does and how it kind of evolved over a period of time, and what basic functionality that was trying to provide. So on the far left, if you see the lower half of the block here. These are basically the systems from which Universal was supplying us the data from. So, you know, their email campaigns that go out, the responses of the end customers, how they are actually responding to those email campaigns, what sort of entitlements meaning the tickets and the guest passes, you know, that those customers are buying, you know, what does the picture from that, you know, sort of a view look like. You know, what are the profiles and the account information of those guests are when they log into the universal dot com site or their mobile app. When the guests are actually in the park, you know, how they are, you know, consuming different type of services like Wi-Fi. And when they are actually checking into the results in the park itself, you know, what that experience looks like. So as you can see, wide variety of transactional data, you know, these are the touch points, these are the interactions that those guests are having with the brand, you know, with Universal Orlando or Universal Hollywood. So the whole idea was that how do we convert these transactions into a guest centric view. You know, Leah mentioned early on the 360 view was the biggest challenge was the biggest objective that they wanted to build. Because yes, there is a lot of data which is sitting in different systems where all these, you know, transactions are being captured. But how do we bring them all together and give them a guest centric view. So that was the whole premise of building this architecture. So as you can see, Universal is basically supplying us these files, you know, through NiFi jobs and other technologies that they use. And simply, you know, pushing them into a data lake environment or a collection layer as we call it on AWS S3. S3 is a technology which is basically think of it as a simplified storage fabric which is, you know, offered on top of AWS platform. So these raw files on a nightly basis are being pushed to S3. And from there, Merkel brought in accelerator which we call as data loading framework or DLF in short. Basically, the whole point of that framework is to pick the files which are being dropped in S3 location and then route them for identity resolution. And when I say identity resolution, you know, that basically means the identity of the guest or the customer across all these different systems, right? So if let's say Uncle Jay is going and buying a ticket and then engaging in different activities within the park, the different rides that I take. How do we know it's the same person who is coming in from all those different transactions? So that's the solution what we mean by identity resolution, which Merkel brought in a solution called connected recognition, which is part of our mercury offering. So that connected recognition is integrated with this data to resolve the identity of the guests across all these different transactions. So we perform, you know, standardization of data, we perform data quality checks. We also enrich the data by adding in, you know, certain demographic attributes. So for example, you know, additional attributes about the, you know, the hobbies or the buying powers or what kind of cars would they like to drive? What sort of, you know, vacation packages they may be interested in what's their affluence power, those type of additional attributes which allows marketing to segment the customers in much more informed way. You know, that's the enrichment part that Merkel also bolted on in this solution. So identity resolution and data enhancement, you know, those are two additional, you know, functionalities that we bolted on on top of cloud platform and created what we call as a conformed layer. Think of this as a data in its harmonized in its standardized format with data quality checks being applied. Any enhancements that we needed to do in terms of data attributes, we applied those and then the identity being resolved. Then finally, what we call as the consumption layer is built on top of it, which allows us to create those different type of views that business was looking for for marketing purposes. Right. So it could mean aggregating the data, it could mean creating facts and dimension type of views. It could mean, you know, that whole customer 360 different type of view is what we created here. So this layer consumption layer is pretty much optimized for different type of consumption pattern. So as you can see here, orchestration tool, which is essentially sending out the email campaigns or data and analytics team, you know, the data science team. They want to develop some sort of advanced analytic model to figure out what sort of next best action or next best offering that they can recommend to the guest based on their previous and past actions. So that type of modeling was developed by their data science team, Universal's data science team. And then on the top piece, which I haven't talked about yet is the digital aspect. So the interaction of the guests on the website itself or on the mobile app itself, how do we understand that behavior? Because the bottom portion here is focused on the transactions which are happening in the park or in the resort itself. But the digital interactions, you know, were happening on the site or on the mobile app. That's where we brought in a CDP customer data platform technology, which was ingesting in ingesting data in real time from all these different mobile apps and different websites. And we created a bi-directional integration of that, meaning those digital IDs which were being resolved digital behavior that was being learned, you know, from those CDP platforms. We were bringing that in into the cloud platform, the snowflake, you know, the database where all this data is being consolidated. And then whatever intelligence we are deriving through the data science, we are pushing it back into the CDP as well. So we created this bi-directional integration between the CDP platform and the AWS and snowflake environment as well. So I know I spoke a lot here. Hope it all made sense, but open to questions. But to move on here, I'll kind of quickly give you a view on what the timeline actually looks like. As you can see, it was pretty sophisticated solution bringing in data from a lot of different aspects. So we started with what we call as minimum viable project. Our focus was that what is the quickest way that we can demonstrate here to kind of, you know, have business adopt this type of thinking and create some wins out of it. So the use case here was that let's start profiling the guests and create some sort of segmentation here, give the business the ability to segment the customers in much more intelligent way. And it was a three month MVP minimum viable product. And like I said, we started with Hollywood. And the whole focus was that how can we create a guest view, a 360 view from all the transactional data that we are bringing in. So the results or the business value that we developed here was that we gave the business the high propensity annual past prospects that they can actually target via media as well. So this was a foundational build, meaning this was the first foray into building the cloud platform for them. And essentially going back to that architectural tenet is that start small yet scalable in nature. So the scalability kind of came in when we voted on in the second wave of the solution development here to bring on the next best action NBA as we call it, which is how do we recommend the next best action to the guest based on their previous or past behavior, purchase history and, you know, engagement that they have demonstrated in the park itself. So this particular round of development, this particular round of project was a little bit bigger in scope much more complex than the MVP itself. And this is where we onboarded Universal Orlando, you know, on top of the foundation that we built in the previous phase. And if you remember from the previous slide, we had this whole data science type of tools, you know, like Jupiter notebooks, Python and other advanced analytical tools, which were then, you know, integrated with this cloud platform, so that the data science can data science team can develop those sophisticated models of predicting what the next best action should be and orchestrating that through the orchestration and CDP tool. So both the digital and the terrestrial by terrestrial I mean the in park activity, you know, and digital meaning the activity happening on the websites or on the mobile sites. All those were kind of correlated consolidated, and we gave them a unified identity across all those activities. And definitely it gave business the ability to optimize, you know, their audience selection in a much more advanced way. And then finally the third phase, you know, would be to widen the scope to completely retire the on-prem presence. You know, so there are still some solutions, still some footprint of on-prem databases which are out there. So the stretch goal here is that let's retire all that on-prem footprint, migrate everything to this cloud platform that we built and pretty much run all the campaigns from the cloud perspective. So overall, as you can see how this was kind of distributed in chunks to demonstrate wins and basically going in a very methodical way to demonstrate quick wins and then further keep on building on those. I would, you know, ask Leah to kind of come in and then describe how the results were, what successes they saw, and anything that you want to chime in from the timeline or the methodology perspective as well. Sure, thanks. So, you know, a big theme that hopefully you've seen emerge is the, you know, building for now and scaling. And, you know, that's kind of stated here as well with the foundational data platform with global scale. We were able to see targeting wins right away out of our MVP that USA Universal Studios Hollywood data platform was used to immediately inform our email campaigns. We also had planned on using that to inform media campaigns, so digital ads and so on, but the part closed. So, but we got some good benchmarks right away with our engagement rates being so significantly higher when we were able to use data about our guests to inform our email targeting strategy. We also have, you know, the broader solution, we're building out a city graph, which is the solution that helps us connect all the different identities associated with an individual. So, you know, again, the email address, their device IDs, their, you know, their name and address phone number, all of that coming together. So, we can get that 360 degree view of the guest. Not only is that critical for now, but is becoming increasingly more important due to privacy and regulatory changes and the need to really, you know, have the ability if somebody wants to opt out of marketing, opt them out of all the marketing that we can identify and all those touch points. I see a question here on the data science. That's a great, great question and actually into the results. So, our legacy platform models that would take hours and hours, sometimes days to score to apply the model scores are now taking minutes and hours instead. So, and you know, that includes the transformations, applying the algorithms, we have a much, much better ability now to also integrate our cloud services into a data science platform. One of part of the beauty of all of this, too, is that because it's more pay as you go type of platform, even while we were closed for during the COVID closure, you know, it was such a huge, huge savings for us that we hadn't invested all that money in the marketing in the technology infrastructure up front. Because as our data science, you know, we were pulling in, realizing the savings right away, the huge infrastructure. We're also, we are looking at a number of data science platforms. Just to kind of round out this question here, and many third party platforms are available, but AWS also offers some, some functionality. So from that end user perspective, anchor had mentioned the consumption layer. And that's what my team, the data analysts, the campaign managers really want access to is they want, you know, they want that end user business access. They want the ability to define the different transformations and even play around with them themselves. So that data democratization, like we're already, you know, have a place to do this improvement in the data lake ahead of us. And finally, the last bullet point here, I kind of integration with data science platforms as well as media platforms is, is a tremendous need for us. And just much easier because there's, because those, those integrations, those APIs and native integrations already exist within Snowflake within AWS. So as marketers, we're very, very happy with, with what we have been able to realize out of the solution. Yeah, yeah. One last slide. Before we open this up for further Q&A. Leah, this one is, you know, lessons learned, essentially, you know, like what were the lessons, if we were to do this all over again, what would we change or what would we keep? You know, so one point from my side is that, you know, that from technology perspective, you know, there was a certain set of architecture that we went with, right? It was the foundation for how the solution is going to be designed, how the solution is going to be built. And when I say no surprises on that architecture, meaning we didn't have to go back to the drawing board and, and change that architecture or change a particular set of technology. I think those technologies performed the architecture scale as we, as we anticipated. So definitely a check mark on in terms of technology choices in terms of platform choices in terms of architecture itself. Leah, you want to describe a little bit on the sponsorship and governance. Yeah, so one of the things we did early on before we started the project really is myself and my IT. We really got to know more goals, you know, cloud space, their strategic perspective to marketing technology. And we engaged our leaders because it was really important for our CIO and our chief digital officer to be aligned. And, you know, that we have support and sponsorship from both of them. Markle was really critical in that because translating marketing needs to technology speak and technology speak to, you know, our marketing functional needs is, you know, it's a rare skill and it's difficult. And we really needed that strategic advisor that both the business and technology really trust with, you know, with our future. So it's really critical to seek help from the outside way to subject matter expertise and service that kind of neutral ground to bring people. And that really allowed collaboration, true team collaboration from, you know, the top to, you know, to the, you know, day to day individuals working together because we know that, you know, everybody knows that that escalation path. We all have the same objectives and goals in mind so cannot emphasize that aspect enough. I think one of the number one reasons that systems implementations and transformations fail is that lack of sponsorship at governance and so, you know, we're just very purposeful about how we went about that. And the next point here is the collaboration. And guys, I mean, you know, if you remember the second phase which was the MVP was delivered sometime late 2019 or 2020. And then we kicked off the second phase which was the NBA piece next best action and bringing on Orlando data into this platform. So we kicked off this project around first week of March, I would say in 2020, which was right about the time when the news of COVID started coming in, you know, from all possible directions within states, you know, New York was becoming a hotbed, California, Seattle and everything. I would say had it not been cloud, we didn't apply breaks I mean the project, you know, was was going full speed. Even through the pandemic and if it were not the cloud platform on which this whole solution would have been built. It would have been a totally different story. So, meaning cloud made it so seamless, you know, for us to collaborate with universal steam sitting in Orlando, Merkel team sitting some people in Denver, some people in Austin, some people in New York. I mean, it was truly a, in fact, we have a global team, you know, and some members of the development portion are actually situated in India. So you can imagine how the global team, you know, came together from both the sides from Merkel and NBC worked on a platform, which is cloud based with all these different tools and technologies bringing together and making it a success. So I would say a big check mark there as well is that we did, we did right by choosing cloud as the interpinning as the underlying platform and all these cloud optimized technologies to technically build a solution. The next point here is around the data analysis. And of course, you know, we went through a very thorough exercise on understanding the data itself, what kind of files will we bring in, what kind of structure it would be, but then again, data analysis is never, you know, complete right so looking in retrospect, I think we could have done a better job in doing a much more upfront analysis in terms of understanding the quality in terms of understanding the completeness. As we started resolving the identities we came across anomalies where we thought data should be structured in one way versus the other, which caused us to go back and bring in some additional data sets or perform some additional quality checks. So again, a lesson learned there, although not a big surprise, but again, always, you know, data is the key on how you want to structure the data. Last point here, Lea, you want to talk about iterative versus waterfall project management. Sure, yes, absolutely. So as we move to this more kind of agile approach, this was another critical area where the business and the technology teams had to really be tied at the hip. I think, you know, often we can get stuck in the circle of what do you want to build? Well, what can you do? You want to start with the uses and that technical background. Again, can I emphasize now how important it is for the business to rely on to be supported? What type of guest? Who are you targeting? What is your marketing objective? Is it email engagement? Purchase of a ticket. So we got down to the level with the Universal Studios Hollywood use case of saying we want to first profile and segment the audience, the customers, past customers. Then we evaluated probably 20 different marketing ideas and landed on, let's let's do an annual pass holder acquisition campaign. So look at our past annual pass holders, learn what we can about them, and then go try to find others in the market that are like them who are likely to purchase an annual pass and serve advertisements. To see or if it's targeted approach to that is in fact higher return on investment than kind of acquire those individuals. And so again, that was, you know, the marketing side that business has. And then working with Merkel and our IT counterparts define what does that translate into in terms of the first data elements we need available, the first connections we need available, so that we had a very achievable, very measurable and very concrete use case to build our agile sprints around. Awesome. Well, that pretty much concludes, you know, what Lea and I had to talk about this particular case study. Why don't we turn the tables and have you guys ask questions to both of us. So let's see if there are any questions. So I think you're not able to see the questions. There are, there are two related questions. One is controls sensitive data. And so like what a challenge, and then there's another question. Cancel PII and other sensitive data in S3 and snowflake. Yes. So definitely yes, there was PII data at play and let me start from the S3 side and then we'll work our way into the snowflake side. So, you know, first of all encryption. Anytime that we have PII data at play, we encrypt the files. You know, so that this sensitive information doesn't, you know, fall in wrong hands and even if it does, you know, there is no way that they can make any use out of it. So definitely everything is kind of fully encrypted. The S3 buckets. Any exchange that we do of that PII information with connected recognition or mercury on the on the Merkel side to generate the IDs, you know, against those PII, that is also done in a very secured and encrypted fashion as well. And then finally on the snowflake side, I think, again, I'm not a big snowflake expert here, but I believe snowflake has a version, you know, which allows you to encrypt and provides much more robust security controls than their basic version or basic software provides. So from all those aspects, you know, wherever PII was at play, you know, encryption, tighter security controls, identity and access management, you know, who has the access to what data. Those are all the kind of checks and, you know, balances that we had to bring to the table to make sure that security is well rounded. There's also a question here about whether we purchased a CDP or created one. So we are licensing a third party CDP. We consider it to be part of our consumption layer because it's more, you know, what the business uses a lot. Okay, so we can start with these cases and we've evaluated several CDPs from LITICs to end others before we made our selection. And, you know, anchor can probably talk more about, you know, the benefits of licensing versus building, but for us, it was really it's about, you know, the CDPs are constantly evolving, both in terms of their functionality, real time, you know, as they move towards more of their real time decisioning support and then the integration partners, the hundreds of integration partners that exist. We just wouldn't be able to keep up with that if it was something that we built ourselves. Yeah. And then just to kind of add to layer the response on the CDP, I mean, CDP is such a loosely used term, it's like big data, right. So yes, to tackle the whole terrestrial aspect, you can consider this whole AWS plus Snowflake platform as the CDP for terrestrial data. And then the third party tool that Leah mentioned that, you know, they bought off the shelf is essentially the CDP for the digital side and the key of this, the key part of the solution was how to integrate the two together so that they can get the full view on the digital and terrestrial side. So that's where the focus was, instead of building a CDP ground up from digital side, we decided to go with a commercial of the shelf tool, which is optimized to work alongside, you know, a cloud ecosystem, a cloud footprint of this sort. And then yet, you know, be able to bring those integration points that how can we pump data in and how can we pull data out, you know, from that CDP platform, and then provide more visibility into the analytical models that were being built. So that was the main focus. So we didn't build that, you know, the third party CDP tool from ground up instead we went with a commercial of the shelf tool. All right, thank you so much, anchor and Leah. Can we give them some claps everyone. All right, thanks again, Leah and anchor for this great presentation and thanks to our attendees for tuning in. Please complete your conference session survey on the page for this session. The next sessions will start in about 10 minutes. Thanks again. Have a great day.