 Hey everyone, welcome to theCUBE's presentation of the AWS Startup Showcase, Analytics and Cost Optimization. This is season three, episode two of the ongoing series covering exciting startups from the AWS ecosystem. I'm your host, Lisa Martin. Today I'm joined by Cube alum, Eric Amuega, VP of Marketing and Operations at Rutter Stack. Eric's here to talk about Rutter Stack, getting the most out of your warehouse data lake investment. Eric, great to see you. Thanks so much for joining us today. Great to see you as well, Lisa. Tell the audience a little bit about Rutter Stack. I see 30,000 plus sites and apps run Rutter Stack. What's the solution all about? Absolutely. Rutter Stack is a warehouse native customer data platform that helps companies collect, unify and activate data. We're going to each of those pieces in a bit, but essentially it is a product that's been built over the last three years to enable companies to make the most out of their data to deliver delightful customer experiences. Delightful customer experiences are all the rage. We all expect them. But talk to me about the impetus or the catalyst to launch Rutter Stack. What were some of the gaps that you guys saw with legacy customer data platforms that you thought, we got to fill these? Absolutely. So when you look at the history of customer data platforms, you'd have to go back to the early 2010s. And essentially in early 2010s, you had these products called Tag Managers. So Google Tag Manager is one of them. You have Tilium. And essentially what they did is they enabled companies to collect behavioral data from their website. This is a very simple example. Behavioral data from their website and send it to a destination to do something with that data. That was the first iteration of that. That was a one-to-one connection where you're collecting data once and sending it to one destination. And then you'd have to have multiple tags to do that. Say, for example, you needed to send data to an email service provider. And then you also needed to send data to a product analytics tool. Those would be two different tags, right? Now that was an innovation at the time. But over time, you had traditional CDPs that were more complete and to end work flows for customer data. You had, you know, segments you had in particle around the 2015 timeframe. And what that enabled you to do, what that enabled companies to do, is collect data once and send it to multiple destinations downstream. So instead of the one-to-one tag connection, you now have one-to-many tag collection. Now, when Radistack, when our founder, Sumida Mitra, was at a different company, he was essentially building a customer data platform internally and realized that solutions that existed at the time did not solve the use case that he was looking for. And essentially he saw a market opportunity to re-envision what a customer data platform was, taking advantage of the major developments that had happened since 2015. One important one is the rise of data leaks and data warehouses, right? That decouples storage and compute. What that enabled companies to do is to have a way to store the operational data and to essentially turn their warehouse of data into the single source of truth. Now, when building Radistack, the question was, how do you build a product or customer data platform in this new age when companies are using their warehouse as a central source of truth? And essentially the difference between the traditional ones is that while the traditional ones stored data in their own ecosystem, right? That created a data cycle separate from the warehouse where companies are already moving to, Radistack was built to work with your warehouse of data leak. So essentially instead of storing data, it collects data, sends it to your warehouse for your teams to model that data and then activate it out of your warehouse. So that's where the warehouse-related concept comes in, where instead of having a traditional CDP storing data, hiding it behind a UI and not providing access to the data engineers to that data, we're doing that with the warehouse at the center of the customer data platform. At the center, got it. Can you explain or compare contrast, a packaged customer data platform or CDP against composable CDPs, DIY builds? Help us understand the differences there and really the unique value drivers for Radistack. Absolutely. I would look at kind of the framework I would use is how hard is it to get value from a specific CDP? And on the left-hand side, the hardest to do is DIY. It's building that your entire infrastructure in-house, right? That's the hardest to get off the ground. You need, you know, potentially dozens of engineers over a long period of time. On the other hand, you have packaged CDPs. Packaged CDPs, they essentially have all of the different components to deliver on our customer data platform. The different components that I mentioned earlier is data collection, data unification, and data activation. That's a packaged CDP. So one solution that handles all of those three. And then in the middle, you have this composable CDP, which is essentially having point solutions for each of those steps, right? You have one vendor that does data collection. You have another vendor that does data unification. And then another vendor that does activation or reverse CPL, right? So especially in this micro environment, what you're seeing is point solutions don't work anymore, right? Companies are looking to optimize their data investments and looking to reduce cost. And so when they look at the difference between a composable CDP where you have to stitch together three or four different tools to accomplish what a CDP does, that becomes cost prohibitive, both from a hard cost perspective and a soft cost perspective. The hard cost is you have to pay three or four different vendors, right? Each of them is making a margin on whatever they're charging. Two, you have soft costs around the engineering cost that it takes to stitch all of those tools together to make sure that they can speak to each other. That takes engineering time, right? And then three, you just have the opportunity cost, the time that you're spending integrating all of these tools is time that's spent on non-value added tasks, right? If you want your data engineers focusing on delivering delightful customer experiences, whether it's running personalized marketing campaigns, sending highly personalized product recommendations, right? Now enter the package CDP, right? And there's kind of two flavors to it. There's a traditional CDP that we talked about earlier, which in this case, you do have the downside of store, they store your data. You're also storing your data and paying for storage in your own data link or warehouse. So from a cost perspective, the traditional CDPs, you have the, you know, you're essentially paying for storage, right? You're paying them to store your data, the data is hidden behind a UI and you're essentially creating a data silo separate from your warehouse. Now, the warehouse-related CDP is a different flavor of the package CDP. And what it does is that, and that's essentially what Run of Stack does. It has all of the three components that make up a CDP, right? You've got the data collection and then you also have the data unification. So it's actually building identity resolution in your warehouse where, for example, you can tie together the entire customer journey and then you have the activation, right? Once you've done the modeling in your warehouse, you need to do something with that data. The way you get value from your data, the way you get value from your warehouse investment is unlocking use cases that drive the revenue, right? And that's the activation piece. So having all of these three components form part of the package CDP. And what we're seeing is that companies are moving more towards that package CDP model and away from stitching together many different tools to accomplish the same task. You talked about customer experience and as I made a joke a few minutes ago, it's all the rage. We all expect to have a great customer experience whether we're dealing with another business, a consumer application, we're getting a ride share. Talk about the customer experience that Ridershark enables and how that really gives competitive advantages to your customers. Absolutely. So the way I think about customers have elevated their expectations in terms of what they mean by a delightful customer experience. One simple example that's actually fairly complicated depending on your tech stack to unlock is having personalized experience, right? So the example is you have a personalized offer based off of your next product to buy, right? Or you have a coupon that encourages you to make a second purchase, right? Now, to deliver that experience there's actually a lot that goes into it, right? The first step is you need to ensure that you have all of the customer data touch points. You're collecting data from all of those, right? Whether somebody purchases a product from your e-commerce website and then a few days later goes to a store and purchases a different product from the store, right? It's a natural exercise to stitch those two together to identify this customer purchase a product online and then went to the store a few weeks and purchased another product, right? Without stitching together so that the first piece is collecting all of that data and sending it to your warehouse. Once that data is in your warehouse you need to stitch that together. You need to perform identity resolution to be able to resolve these two transactions under one customer, right? Once you have that full customer profile in your warehouse, that's a unification piece you can now, you know you can now build ML or AI models to, for example, determine what's the third product what's the next product to buy given this customer purchase two products one online, one at retail, right? What's the third product that they would need based off of, you know, some deterministic or some probabilistic modeling? Now, once you calculate that, right you need to now have a touch point with the customer where you're sending an email with, you know 20% of coupon to buy this third product or product recommendation that you're able to deliver when that customer comes back, you know to the website or logs into the e-commerce platform on the app, right? That is an end-to-end customer experience that's very difficult to deliver with if you have incomplete data if you have inconsistent data, you know across your tech stack. Well, what you described is a challenging problem to solve it's also what we expect. It has to be personalized experience has to be relevant offers don't offer me something I've already bought I want you to know what I'm coming back for and be able to predict that and that's kind of the auto magic behind optimized customer data platforms. I understand that there are three costs that companies should need to consider when they're selecting a customer data platform. Can you walk me through those three costs that people really need to be considering? Absolutely, and I touched on this earlier I would say there's the cost that you see upfront, which is the vendor cost how much are you paying in your contract with your sales vendor that is the, you know what's most transparent that's why I call it a hard cost upfront cost you're paying so many thousand dollars to get a tool, right? That's the first one. Now, that's only a small piece of the entire cost, right? Once you've purchased a tool you need to implement it, right? You need to spend engineering hours to make sure that it's implemented that it's getting the data that you needed to get in the format that you need that data in, right? That's a natural access requires, you know engineers that, you know make hundreds of thousands of dollars and spending their time teaching together, you know different tooling across your entire stack and then the third component is when they're doing this implementation work, right? That's taking away from taking away from value added activities, right? For example, building, you know you know, highly, you know performance ML models to determine, you know tolerance and ensuring that we can save customers that are at risk of turning, right? So any hour that, you know your data team is spending teaching tools together reconciling different data schemas it's time spent away from driving revenue growth, right? And which is which goes back to the distinction between a warehouse native package CDP like Redostat versus a composable approach where essentially during, you know three or four different implementations before your team can focus on unlocking value from those integration. So you talked about vendor costs kind of the obvious engineering costs but there also sounds like there's opportunity costs talk a little bit about those because that's that's somewhat hidden that organizations need to really be aware of. Yeah, and this this goes down to, you know 10 value of money, right? A dollar today is worth more than a dollar, you know, three months from now, right? So when you think about, you know why are companies investing in customer data or in the technology around customer data? The reason they're doing that is really to drive revenue growth or increase marketing efficiency or, you know those are really the main two use cases that we see we can drill down into each of those but essentially what companies are looking to do then looking to unlock value from their data investments and the way to do that is drive growth or increase spend efficiency. Doing that today is significantly more valuable than doing that in six months, right? So if it's going to take six months to implement and stick together three or four different tools you're losing out on six months worth of revenue. And in this environment you know where it's a challenging market environment is a hyper focus on reducing cost, hyper focus on revenue growth. You know those are trade-offs that are difficult so if you have to wait a long time to implement and get value from your data investments. So the opportunity costs are massive and you've been talking about impacting bottom line, impacting top line. So then summarized all this Eric, how does the cloud data warehouse or data lake help to optimize those costs that you talked about that are really clearly there? Yeah, so I would say there's a number of different ways of thinking about it. So one is a lot of companies that we work with already have a cloud data warehouse or a data lake as part of their infrastructure, right? Now they already have it and the options are they could either purchase tools that are parallel to that cloud data warehouse which means that you're purchasing for example a customer data platform that's storing data in their own infrastructure that's essentially a separate data site from all of the investments that you've made in your data warehouse or data lake. So it enables an investment in your cloud data warehouse if done right and if the center of gravity of all of the data switches to the cloud data warehouse or the data lake it reduces the storage costs right? You're storing data once instead of multiple different times of the different tools that you have that's number one. Number two is there's a there's a data quality argument to be had as well, right? If you have data, duplicate data stored in many different tools in relation to your warehouse you oftentimes you could ask the same question off of each of the different data sources that you have or data storage that you have and you get a different answer, right? Which means that your data team has to go back and try and reconcile why is, why am I getting different answers from my data, right? Having a well implemented cloud data warehouse or data lake as a center of your data strategy enables you to essentially not deal with data quality problems that create a lot of turn within the team to figure out why data is inconsistent, right? So by ensuring your data quality, data governance reduces the amount of time to spend, you know, running around trying to figure out what is the actual source of truth when it should be in your warehouse or data lake. Data governance, data quality, data privacy is so important. So Eric, take us home here you must have a handful or more customer stories that you think really shines the light on the value that Reuters Jack is delivering. Share an example with me. Absolutely. So I'll start off by saying we have, you know hundreds of customers who have grown significantly over the last, over the last two years and we're seeing an acceleration on bad growth, you know, due to the micro environment folks focused on cost and efficiency and but if I have to pick one we do have one of my favorite stories, a large furniture retailer that's publicly traded they were essentially working with one of the traditional CDPs and they were unable to get value from that investment right that spent, you know sticks or so months trying to implement the tool and they're not getting it fully implemented for a number of different reasons and they were looking for an alternative, right? Because the integration cost, the engineering cost time spent on trying to integrate and building out their customer data stack with the traditional CDP just didn't work given that they wanted to reduce the storage cost they wanted to make sure that there was data quality and data governance across the board it did not work. That's the first piece. Second piece is they were looking to get a full customer 360 view of their customer. So they did want to be limited to the handful of data sources that this traditional CDP provided, right? So they looked at, they came to Rada Stack and they realized, one, there's a faster time to implementation, right? Instead of you know, six months and not making progress they were able to get up and running in a matter of weeks and then number two, they could actually get customer, customer integrations built, right? You know, Rada Stack is an accessible platform it's fully flexible and works around your cloud data warehouse which means if there's a new data source right you can build a customer integration either using Rada Stack or once the data is in your warehouse you can actually use our identity resolution features to build that customer 360 so that you have rich customer data that you can then enrich and unlock value from that data. So they were able to reduce cost from a vendor standpoint, reduce cost from a data storage standpoint by migrating all of their workloads to the data warehouse that they had already and then also accelerate time to value right instead of spending six months on integration they were able to get up and running in a matter of weeks two to three weeks, right? So that's an example of where they were able to reduce cost and get accelerate time to value by using a warehouse native CDP that Rada Stack is. Yeah and getting that customer 360 which is really the pot of gold at the end of the rainbow for any kind of business. I know all of the vendors that have that 360 on me works every time. Eric thank you so much for coming on theCUBE and talking to us as part of the AWS startup showcase analytics and cost optimization. We really appreciate your insights and learning more about Rada Stack. Thank you. Absolutely. Thank you so much Lisa. My pleasure. We want to thank you for watching and say keep it right here for more action on the theCUBE your leader in live tech coverage.