 Hi, and welcome to this week's Wikibon Weekly, presented by theCUBE. Every week, we are combining with theCUBE to present research insights that are relevant and important to our research communities that we serve. And today, we're very, very lucky, because we have Bill Schmarzo, who's at EMC Global Services, CTO of the Big Data Practice, who has willingly taken time away from playing Pokemon Go and being chased around by rat attach. I was told that I'd get a rat attach if I came here, so. Yeah, we'll find one for you some day ago. But to take crucial time to talk about some of the big trends in the computing industry that are being affected by big data, big data analytics, and the evolving practices and experiences and associated with building or deploying big data in business, so Bill, welcome. Thank you, thanks for having me. I'm looking forward to the conversation. It's a hot topic. So Bill, we're going to talk about three separate segments here. The first one that we really want to emphasize is that for 50 years, the information technology industry has operated on what I'll call stylized data. These are data models that are constructed specifically to support accounting applications or HR applications. Nothing wrong with that, very, very successful, but they have allowed us to satisfy legal requirements or other requirements from an operational standpoint. But big data is taking us somewhere different. While we're using analytics to look at that better and look at those operational elements better, big data is taking us into the real world where we don't have access to these stylized data models. We have to find ways to translate analog data into digital data that can, in fact, then be applied to business insights and models to run the business differently. Correct. What do you think is, I mean, how hard is that? As you work with your customers, what are some of the things you're doing to help them envision analog data in digital forms to build these better business insights? So, first off, it's really hard. And the reason why it's really hard has nothing to do with technology. Has everything to do with the culture. And, Peter, you summarized it really well. Stylized data. When organizations stylize data, they, by very virtue of doing that, take value out of the data. They make assumptions about what's important and what's not important, and they leave a lot of the valuable aspects of the data laying on the floor. They cut out sometimes the best parts of it. So, it's a hard cultural challenge for organizations who, for a long time, have thought about data as a cost to be minimized instead of an asset to be utilized. So, there's this cultural aspect of it. And the way we help organizations to start realizing the value of the data is we do something I think is really simple. We help them to figure out what decisions are you trying to make? So, we go through a whole envisioning process about what are your key business initiatives, who are your key business stakeholders, but really boil down to decisions. Because really, at the end of the day, it's all about making better decisions. Decisions about which customers to make, what offers to, which wind turbines, they need, what kind of servicing, what patients need, what kind of treatment. So, if we can focus on the decisions, it's a very simple or simpler way to get the business people to understand the value of the fact that there's value in the data and that if I can apply that data to the analytics to make better decisions, then I've got a way to drive value in the organization. So, in many respects, what we're really doing, Bill, is we're using the decision as a way of circumnavigating or capturing the data elements that will be required to generate the insight that then leads to the activity. Bingo, and the beauty of decisions is that every business user you talk to understands that they're trying to make decisions. It's very natural to them. And by the way, that's very different than questions. People always say, well, what question are you trying to ask and answer? Questions don't imply actions, right? You have all kinds of questions. Decisions imply actions and actions are important. So, we find that when we have a focus around the decisions, then there's two things that fall out of that. There's, number one, is the data itself. What data do I need? What data might be useful? Opens up this whole concept that there's all kinds of different data sources out there that I might want to contemplate that might be more, that might be valuable and help me make better decisions. The other aspect, there's all these analytic models, right? We call analytic profiles, right? The models that help me to support the decisions I'm making that combine with that data to help optimize the decisions and drive analytic lift. So, when we think about this notion of capturing data, ensuring that it is appropriate to the decisions, we're actually successfully, hopefully, putting some bounds and constraints on it. Now, this is somewhat antithetical to what a lot of data scientists think, who just want everything and then they'll figure out what's valuable. By focusing on the decision, we are actually not stylizing it, presuming that we're going to generate greater success in these data than we might in other data. Now, if I focus on the decision, we do two things that organizations do really poor. We focus and we prioritize, right? Because, yeah, you could throw any amount of data you want into a data lake. You could throw all kinds of data, but you'll end up with one of our clients called a Dead Sea. You end up with all this data out there that no one's using, no one knows it's out there, and it becomes this anchor on your creativity. And so, while we want to embrace the creative thinking process that data scientists normally go through, we also want to focus and prioritize, because not all data's of equal value, especially with respect to trying to drive decisions, and not all data's of equal implementation feasibility. Some data's hard to get, some you have to go out and buy, and so you've got to weigh the value and the cost aspects of each of those data sources vis-a-vis the decision you're trying to drive. So the decision informs the type of data that you need, which then also gives you some visibility into the value of that data in the context of the decision. So you've hit on a really interesting topic here, and organizations are really struggling with this idea of how do I determine value? How do I figure out how valuable my data is? And organizations, when they've used data to report on what happened, really didn't care because the value was minimal, right? You talked about compliance, regulatory reporting, things like that, but now I'm trying to drive decisions. If I can improve my ability to reduce churn by 10%, there's a hard value associated to that. If you're Chipotle and you can increase your same store sales by 7.1% year over year, which by the way is a big struggle for them right now, that's worth $200 million a year. So there are value in the business initiatives and the supporting data, and consequently there's value then in trying to figure out which data sources are most appropriate for helping me to achieve that particular objective. So as you know, we both are business people. We've been in large and small companies, couple of startups, and as executives we always asked ourselves, okay, what financial resources are gonna have to go here? What people am I gonna have to put here? What capital assets am I gonna have to put here? But now you're saying that we need to add to that simple rubric the idea of what data do I need to put here to be successful in this business endeavor. Amen, amen, data's got value, right? You've summarized it many, many times for your talk about data in use has value. And so data is a key aspect of that. And that's by the way the cultural challenge we face with organizations. It isn't that they don't know what decision they're trying to make, it's that they come from an environment where data has always been treated as a cost we minimized. You sell to the infrastructure people, right? You approach the storage people and you're talking about data and they're trying to figure out how to drive down costs when the very top of the organization they're trying to solve some hard business problem regarding customer acquisition or predictive maintenance or quality of care, whatever it might be that relies very, very heavily on having the right kind of data at the right level of granularity and the right analytics to help drive that. It's a huge cultural mismatch that we've got to help organizations to sort of address. Now I'm going to ask you a question. We actually think that it's possible to even think in terms of data capital or digital capital, where you have stocks of data that could be in the warehouse but hopefully or in the data lake. Thank you, yes. The Dead Sea. Yeah, not the Dead Sea. The Dead Sea. That could in fact, that if it's properly catalogued, you know what it is, you know how to get it, you know where to find it. You have other sources of data from partners and maybe that you're buying periodically but that ultimately we can actually think about how these stocks of data can be applied to decisions in the form of capital. Is that a concept that your customers are starting to use or at least think about even if they're not using the same term? Yeah, we're actually seeing more and more customers starting to realize and make that flip to realize that data is an asset. I can't put it on the books yet, yet being the key point but I know it's an asset in my ability to drive better decisions and drive monetization efforts. And so we're starting to see more and more organizations starting to realize that the data that I have has got value to it but there's also ability to go out and get third-party data you can buy or even publicly available data. You know tons of it out there that I can start bringing it in and that data when I lay it next to my operational data that I've already have greatly enriches the data I currently have. It takes and adds a whole new dimensions to what I got from a perspective of leveraging the data to help drive those decisions. I want to come back to kind of the original question. The, you know, you and I are having a great conversation, analog. Now we're capturing it on video so we're translating it into a digital format that then can be broadcast out to Wikibon clients and prospects and others. As we think about some of these disciplines for going and gathering that data, how are your clients starting to think about making investments to increasingly take more of that analog data that's so crucial to how business actually gets done and turning it into digital representations so that we can bring it underneath this umbrella of digital capital, big data analytics and some of these new actions. Some of these new insights and decisions. You've hit on a lot of really interesting leading topics there. Obviously, the data has value but what you started talking about there, Peter, was is that the representation of the data in a manner that helps me to make better decisions is a really key point where only a few of our customers are starting to get their heads around the fact that yes, data, so there's, to me there's kind of a multi-step process versus a step that you realize, wow, data's got value because it helps me make better decisions but here's an analogy for you. Data is to analytics as money is to wealth. Having one doesn't guarantee the other and so people are starting to realize that I've got data, I'm starting to collect it, I'm starting to have association value, but if I can't- But where wealth accretes as a consequence of having money, is that what you mean? Yeah, okay. Right, I mean you can have money but not be wealthy. That's right. Right? You can have data not be analytic. The point is to apply your money to activities that then generate a superior return. Exactly, and that's the point I think we're seeing some of our leading customers and realizes that I've got the data and I've got to start building analytics but not one-off analytics. We're starting to see a lot of pretty mature organizations starting to go step back and go, I got a whole floor full of orphaned analytics, one-off analytic projects that I did to solve a particular problem but it's not being reused again. So in the same way that a data lake is a concept that's helping organizations to capture and enrich and mine that data as an asset, organizations start to realize I need to have a framework around my analytics so that I'm not one-off building analytics, I'm building analytics that have reuse, to have the same capital value, right? Digital value that data has. And so we're sort of at the very early stage of the organization realizing that what I'm doing for the data lake regarding data, I need to do something similar around my analytics to avoid this problem of these orphaned analytics. So we mentioned earlier, you mentioned earlier that we know that data has value and increasingly that's being defined in terms of the context of the decision of the work that needs to be performed and that we're not in a position yet to put data on the books or put the value of data on the books. But that doesn't mean we shouldn't stop thinking or conceiving of how we can actually construct an overall framework for capturing, measuring, and ultimately continuously investing in data so that the value of that data goes up. Are you starting to see customers and are you yourself advocating? Because we're starting to. This notion of I don't wanna call it the shadow books but let's call it the data books where data capital gets applied and generates returns that then can be subsequently recaptured and applied to new activities. Yeah, so smart companies don't let the accounting rules dictate how they create value. And so what we're seeing organizations starting to realize that in the big data discussion there are three sources of intellectual capital. Things that have sustainability long-term and by the way, it has an interesting aspect on the technology. Remind me to come back to that. So the three areas of we see organizations and realize there's intellectual capital within the big data conversation are the data. The decisions are use cases which are closer to the decision I'm trying to make. And then the analytic models that I'm building to support that. And we're advocating for our customers that the analytic models, you wanna build the analytic models around your key business entities. We call them sometimes strategic nouns. So think about that might be, if you're a public school, it would be students, teachers, tutors, administrators, classes, courses. There are a finite number of entities that comprise your classrooms would be one, right? That comprise what is your organization. And I wanna learn as much as possible. I wanna gather analytic insights when I understand propensities and tendencies and inclinations and passions and interests and associations and affiliations. And success rates. And success rates, right? For each one of those things. Let me give you a great example. So we had this project for a public school system we did, right? And we created what their teachers called Netflix for teachers, right? Which basically made a bunch of decisions for them regarding how to group students, what students, tutoring, what students need interventions. We even created customized curriculums for every student. In order to do that, you have to know details about every student. You have to build up analytic profile on every student and every teacher. What we found out, by the way, we had some teachers were not doing very good work, right? Their performance levels were really low. And what we found is if you look at the teachers, the problem wasn't the teachers, the problem was the mesh of the teachers with the students with the classes in the subject areas. And when we moved them around, their performance jumped. And so we're starting, very long with an answer to your question, is that organizations are starting to realize there's three sources of intellectual capital long-term, the high-term value. The data, the models. And the decisions. That's it. We call it, by the way, we call this a Rubik's Cube. Because the challenge organizations have is, how do you take this and put it into a Rubik's Cube where all the different sides match up? It's three dimensions. It's three dimensions. It's clever though. Yeah, that's very clever. Don't find a fourth one, please. So let's close this first segment here on this notion of how data, analytic models, or the three assets, data, models, and decisions. Because we're going to pick that theme up in a future Wikibon Weekly. Oh, super. So thank you very much for joining us. Bill Schmarzo, CTO of the Big Data Practices and Global Services at EMC. I'm Peter Burris, broadcasting here from Silicon Valley and Wikibon Weekly, The Cube.