 Live from Cambridge, Massachusetts. Extracting the signal from the noise. It's theCUBE, covering the MIT Chief Data Officer and Information Quality Symposium. Now your hosts, Dave Vellante and Paul Gillin. We're back in Cambridge, Massachusetts. Dave Vellante with Paul Gillin. Jamie Popkin is here. He is the Managing Vice President of Data Management Strategies and a distinguished analyst at Gartner. Jamie, thanks for coming on theCUBE. Great, thanks, glad to be here. I know you're super busy and we're sort of grabbing you as you're running out. So really thrilled to have you here. We had Doug Laney on earlier, talking about Gartner's infonomics. Have you attended this event before? Was this your first time here? This is my first time at the event. People that have been on my team have been supporting the event for a long time, but it was fortunately the first time I got to attend. Yeah, so it's evolved from sort of data quality and then of course the big data theme comes out and everybody's hopping on that and analytics and the chief data officer. What do you make of sort of the transition that we're seeing from this sort of old, stodgy data management topic to really, you know, sexy, exciting big data analytics? What's really happening there? I think that most enterprises recognize that data needs an advocate within the enterprise. And so the old ways of managing data, the old ways of building applications, cobbling together analytics on an as needed basis isn't cutting it anymore. And so I think people recognize they need to build a data governance strategy and they need to really treat information as an asset. You know, the promise of whether it was decision support or business intelligence, you know, for years, they ever literally lived up to that promise. And now we're sort of hearing a lot of those themes again, predictive 360 degree view of your business, you know, governance, you know, that actually is going to be achievable. Do you think we're at a stage now where we will realize that vision or is complexity higher than ever? It's going to be harder than ever. We think about it in terms of the analytic process. So there's descriptive, what happened? There's diagnostic, why did it happen? And the literally billions that have been spent in the last 15, 20 years were really good at doing descriptive and diagnostic. And I think most enterprises are there, but the future is predictive analytics and prescriptive analytics. And that's where the phase that we're in now is how do you take all the data that you have and manage it in a way that you can build predictive and prescriptive analytical applications that become embedded in the business operating model. And I think that's what we're seeing right now. What do you think the role of the CDO evolves to? It seems right now that a lot of CEOs are still dealing with basic data governance issues, data quality, a lot of plumbing work. Do you think that this role evolves to become more strategic or is there just so much plumbing work to be done that like CIOs, those who never got out of the infrastructure game, could they get mired down in this stuff for the long term? I guess I would argue that the data hygiene is strategic because if you don't have the proper data management and data governance in place, you'll never be able to do the advanced analytics. So I would say that, sure, some organizations are still going through some basic data hygiene. Others have figured it out and are operating well and now they need to prepare to the next phase. Start about your practice a little bit at Gartner. Your focus, how it's evolving, how you're helping clients. Sure, we're focusing on doing research for the chief data officer, for the chief analytics officer, helping them do what this conference is doing, understand what is the role? What am I supposed to be doing? What's the scope of my responsibility? What kind of impact I can have? So that's a lot of the research that we're doing. In fact, we just completed a survey, we got the results in and now we're starting to do the analysis which we'll publish in the next couple of months, where we're looking at that, understanding what the early adopters are doing and then figuring out how to talk to others about that. I'm also very much involved in big data analytics. To what extent, Jamie, do you think the job of the CDO should be to work him or herself out of a job? In other words, and you were on a panel with me earlier, we talked about this, to get data so imbued in the organization that essentially you don't need a data czar anymore. Well, I think for the near term, at least the next three to five years, the role is an important one to provide that advocate for data, as I said at the beginning. I think over time, the CDO role will go from one that's focused on risk management, making sure people don't go to jail because they reported improperly or violated compliance, to being much more strategic around bringing in the advanced analytics. So sure, five, 10 years, perhaps you don't need a CDO, but I think there's a lot of runway between now and then as organizations figure out how to use data more strategically. But now, so tactically to protect the organization from risk or getting sued or whatever, and strategic, but also potentially tactical to help grow the company to the extent that data can drive new sources of revenue, why wouldn't the, if in fact that occurs, the CDO role might evolve into an ongoing, you know. Sure. Not just an advocate and a cheerleader for data, but one that actually drives new growth initiatives. I think in the era of digital business, which is definitely the era that we're in, you have to use information more strategically. And for those organizations that are still doing some of that basic hygiene, they're falling behind a little bit, and those that are getting on top of it and creating CDOs and starting to examine how advanced analytics can grow the business strategically. That's absolutely where the game is right now. So I've heard some, yesterday, somebody said on theCUBE that the analytics role should be part of the CDOs organization. Now, of course, that's a, a lot of people would disagree with that. Maybe some would agree. I'm sure the CDOs might agree with that some anyway. A lot of CMOs might think it should be a CMO function. The Gartner came out with the prediction that CMOs would spend more than CIOs by whatever date you guys predicted. What are your thoughts on where analytics should fit? Is it analytics everywhere? What about that function? I think analytics needs an advocate as well. That's part of the research that we're doing right now, and hopefully this study that will be coming out in the next month or two will focus on some of those issues directly. There's probably a case that could be made that some CDOs should have the analytics function as part of what they're doing. There are other cases where you could make the argument that no, the CAO needs to be separate because analytics is focused on a different set of roles and responsibilities and functions than some of the infrastructure management and data hygiene. So I don't think it's been decided yet. I think you can make arguments both ways either way. It really becomes something that each organization has to look at their own needs and requirements. I mean, it really does feel like it's not a one size fits. I think we've determined that. I think that's been clear from these discussions. I wonder, I mean, doesn't analytics belong within the business units though? Do you need a CAO? Well, there may be some analytics, particularly as analytics gets embedded in the fabric of the business model that you might need some analytics shared services. Right now, a lot of the analytics absolutely happens within the business units and that's fine. Over time, organizations might find that some of the analytics as a service or analytics as a shared service might be useful for their business model. So the survey that you were talking about before, have you done a quick scan of the results? Can you share anything with us at this point or you have to go take a bath in the data first? Well, we certainly, I can't draw any conclusions yet from it. We're still going through the results. It was a qualitative survey based, so it's not just adding up the numbers on a quantitative survey, so we need to read them. But I think what we're finding and is certainly validated by this event is that there is a wide range of scope. There is a wide range of organizational structure and that people are really still in an early phase of adoption for it and what we're getting out of the survey is some of the richness of the experiences that people are having in different types of CDO roles. How does Gartner look at, you've mentioned before, you're involved in big data analytics. You know, when a hot topic like big data comes up, it cuts across virtually every part of the IT industry. From hardware, from software to services. How do you guys approach that topic? We've had Merv on before, Merv Adrian. Obviously, Articulate follows that. He's always at all the Hadoop conferences speaking. I think you're in a different part of the organization. Is that fair to say, right? Not so different. We cover similar topics. I think, you know, Merv is certainly covering a lot of things like the Hadoop technologies and some of the newer technologies. Frankly though, from a data strategy, analytic strategy, the issues, the way we characterize big data, volume, velocity, variety. Frankly, you don't need to have a whole lot of any of those to still have processing problems. And so there are lots of organizations that are still struggling with certainly velocity. How do you manage data from different sources coming in? How do you make sure that your current information is current for the customer experience? You don't necessarily need big data to get that kind of processing capability in line. So I think big data is causing all types of data processing to be looked at and say, are we really being responsive to what our internal customers need and what our external customers need? Is big data to some extent taking our eye off the ball, though? I mean, focus on it. It's kind of the shiny new bubble, which for a lot of companies have much more fundamental needs, about getting their own data at data house in order before they move on to something like that. I think it can be. So I'll get a call from a client saying, I have a big data problem. Should I use Hadoop? And then you start to talk to them and what you realize is, oh, you have a master data management problem because you don't have a single view of your customer, which has nothing to do with big data or using Hadoop. You have a problem with how you built your applications and how you were able to aggregate data across all of your different systems to see a single view of the customer. So worrying about Hadoop in that situation? No, that's a problem that's not solved by throwing on a tool. Okay, I'm glad you brought that up. So what are the common misconceptions about big data that you hear among the enterprise you work with? That somehow the new sets of tools are somehow going to magically solve a problem. We've heard that before. It's easy. It's so easy to do a vendor selection, a product selection, a technology evaluation, because we know how to do that really well. But if I have 37 ERP systems and I'm having trouble aggregating customer data, the new tool, right? Is it going to make a difference? Is it going to make a difference? Is that a tool? Exactly, exactly. It's a great conversation, Jimmy. So thinking about the evolution of this topic, you mentioned Hadoop as an example. People always talk about technology, process, and people. And data. People, process, technology, and data would be the four legs. So that's the fourth piece of it that is now injected into this equation. And technologies, they always say the easiest part of that as you just sort of alluded to. At the same time, technology has sort of created this problem. So in thinking about when you work with your clients, people, process, technology, and data, are we, when we started this business, we called it data processing. Are we sort of headed back to maybe a distributed but a data processing mindset as a discipline? I think the discipline is going to come into thinking about that data and applications are not just a way to remember that a transaction happened or report to management that those transactions happened. What digital business means is that we're embedding, processing, we're embedding data into the actual function of your business model, the way that you make money, the way that you interact with customers and suppliers is now being done digitally. And so having IT be something that happens later or helps us remember or helps us report, that's old style where the new is, how do we embed IT and data into our actual business processing? That's the exciting part. Well, that's right, that's exactly right. Just quick, are there any technologies you see right now that do really excite you in this area? Yeah, so I think many of the big data technologies, things that are happening with the Apache project, the Hadoop, the Sparkle, all of that's really exciting stuff that people need to learn and develop skills, learning Python, learning R as a way to build analytic applications. Person had been doing a lot with text analytics, sentiment analysis, natural language, question answering technology, all really fascinating stuff, and then smart machines, a whole new area that we're going to see coming on very quickly. So all of that's really exciting. Now Jay, I've seen you're linked in by also chairman of the symposium. I have served as, I'm no longer, but I have served a couple of times, yes. So I know it's, we're talking to Doug, it's coming up in the fall, it's your big event, and you participate in that, and that's to Gartner's big event, right? Yes, absolutely. So we'll have a lot on the chief data officer, we will be unveiling the research around the survey that I mentioned at symposium, certainly a lot on big data, analytics, smart machines, all really hot topics at symposium. Great, awesome having you on theCUBE, Jamie, thanks so much for your time. Appreciate it. Thank you very much. All right, we'll be back to wrap up day two, MIT IQ, the CVO symposium, this is theCUBE. Right back.