 Live from Cambridge, Massachusetts, it's theCUBE. At the MIT Chief Data Officer and Information Quality Symposium. With hosts, Dave Vellante and Paul Gillin. Welcome back to Cambridge Mass, everybody. This is Dave Vellante with Paul Gillin. We're here at the MIT Information Quality Symposium. This is theCUBE. theCUBE is SiliconANGLE's live mobile studio. We go out to the events, we extract the signal from the noise. The Chief Data Officer is a role that's emerged within regulated industries, in particular, financial services, government and healthcare. And we're going to drill down into the CDO role and maybe even talk a little bit about the healthcare segment. Professor Yang Li is here. She's with Northeastern University. Yang, welcome to theCUBE. My pleasure. Yeah, so you're doing a lot of work now, particularly in healthcare. You've got a cross-discipline, you've done work in supply chain, many, many things. Right now, you're focusing on the Chief Data Officer and the role of the CDO, and in particular, you're focused on healthcare, right? Well, currently, yes, yes. Because I thought that the healthcare is one of the major areas that many CDOs are working in the healthcare and financial industry as well. And I thought that many things are done, but maybe data managers or CDOs can contribute something that they haven't really figured out. For example, medical error, is it all done? Many people are working on it, but is it something that not medical people, data managers or CDOs, can they contribute something in that area? Because data flows all change. Not just database, but how data is collected, stored, managed, and then they're consumed and used, and then it's also reused. So I tell my students that data has more than nine lives. Cat has nine lives, and data has more than nine lives, and somebody's information is somebody else's data. So it gets very complicated, and also any process that you see, you'll see some silos. So left-hand doesn't know what the right-hand is doing. So once we open up that sort of structural problems, so we examine business process, and we examine how people work, and how routine work is handled, and how data flows. I think there's something that we can figure out and contribute in addition to what people are working on. Yeah, I read a stat one time, I think the average data set that's created has, the number 13 is my head, 13 and 25 may be copies, just copies. So you're talking about the full data life cycle, the pipeline of data, a friend of mine in the financial services calls it the data pipeline or the data factory, and you're studying that particularly now within healthcare. So when you look at a, when you initiate a research project, how long do you spend? Like for instance, in the healthcare, how long do you spend on that segment? Is it months, I presume? Or is it years sometimes? Or how long do you take studying a segment? I think depending on the purpose, the main thing is that you really talk with people who have been working on that. You know, I usually talk with the practitioners in the hospital or insurance and all that sort of space, healthcare space. So they have been working on 10 years, 15 years, 20 years. And we might have just studied one or two or three years, but we can get all the wisdom from the practitioners. And then we bring in the theoretical background. And so we collaborate together and come up with better insights. So do you formulate a premise going in or do you go in with an open mind? So what are you after right now in the healthcare industry specifically as it relates to information around the chief data officer and data? What are you trying to extract? One of the things that I realize is that the data that is collected and used for the clinical side and the data that collected and used for the management side, they may not be consistent. We could speed up many decision making by making it consistent. So that is one area. And another area is there are so many routines developed in hospitals, emergency rooms, and even in IT shops. And are there any routines that we need to review? Is there anything that we could compare business process with how data flows? So that are the areas that I think we're just opening up now. You published a paper in March, the cubic framework for the chief data officer, really very comprehensive. Study of, you've interviewed great many CDOs for this. The profile that you define here looks like kind of a Superman. I mean, this person has a marketing role, development role, an internal evangelism, external selling concept externally, they're involved in all kinds of different projects. I mean, did you find that is there any one profile of a CDO that you think is emerging or are these people, is this job going to be very specific to the environment and the needs of that company? Well, first, thank you. I think I agree with your assessment of their Superman and Superwoman. The CDOs are, yes, indeed. The reason why the roles are all over, for example, I identified sort of seven roles based on the three dimensions, whether they're focusing on internal aspect or external aspect, or they're focusing on different data space, or CDOs are focusing on just doing the sort of operational, keeping the lights on so to speak, or they're focusing on strategic aspect of company. But I think the key insight here is collaboration. So CDOs collaborate with other officers. So in that case, depending on the organization's goal, and depending on which CDO that they're collaborating with, their role will be quite different. So most common goal is I recognize the first one, that they are internally focusing on it and they're trying to make the service done very well. That's probably most of the CDOs that start with that role. But other CDOs move on to external roles that they work with other supply chain partners. And later on they can really work with the CDOs to really set up innovative process to come up with a new product and new process service, or they could change even the way they do business. So it's more of a transformational role. So it varies. Did you find that many of these CDOs were building organizations underneath them, or were most of them operating solo, alone? They are working solo. I would really urge them to really work together because I found that CDOs who are successful and have longevity are the ones that will collaborate with other officers. I think in corporations it's often thought that if you don't have a staff, if you don't have a budget, then you don't have a job or you don't have a job in the future. Since most of these CDOs are working on their own collaboratively, as you say, do you see that as a threat to the role? Oh, I see what you're saying. I think what they work on their own but that they should have a CDO office. In other words, they should have their staff and they have their budget, just like any other C-STEP. And the staff might work on things like standards for data in a particular application. Right, so there will be some of their staffs will work on data quality, some others will work on data governance, data strategy, architect, so depending on their agenda, they will work on different things, like new things like data analytics. So one person cannot do all. I think we can see the value, inward value of the CDO but in what context did you find they had external value? Many CDOs, after they have some sort of house kind of in tech, so to speak, and they go out and talk with other CDOs in the industry. And for example, they will develop some kind of consortium and develop to work on standardization of their business process or standardization of their technology or standardization of their data. So let's decide on certain items so that in this industry we say these are standards and those are not. So that is very important because now companies exchange share data. So making sure that their item in their own internal data are all set, that doesn't necessarily mean that they're going to have a great data use because data is going to go over different companies and different industries. You mentioned collaboration being important. So what is the ideal personality profile of a CDO? What kind of people are they? I would say people who can reflect on what they do. So I think you can go and sort of, you know, you can be go and gather and communicate very well. That could be one aspect that, you know, regardless of whether you're a CDO or other officers, they will be welcome. However, I think particularly CDOs, you have to know how to reflect on what you're doing because there's no sort of set rules or set answers. These are new area. So you should really, you know, I think this symposium is a great opportunity that they can come and talk with other CDOs in different industries and then they reflect on what they do and then they go home and maybe they will tweak some of the things or they might initiate something different to do. Professor Lee, what do you say to an executive in the corner office who says, you know, I'm thinking about maybe having a CDO. Everybody tells me I need to have a CDO. Why do I need a CDO? Why do you need a CDO? Traditionally, many data related work has been done without CDO. The trouble was CDO, the data management's work is really closely related to business process. So people who are at the middle level and then they try to do something and then they found that somebody else's business process is either flawed or not handled very well and they go in there and then I hear this again and saying, you are digging in my backyard. So the managers, they may be all good leaders but they don't have organizationally sanctioned authority to do the work in other business process. So if you have the C officer and the company says, right, your work is to go and dig all over gardens. So the entire organization becomes their own beautiful garden instead. So when I hear Paul describe in your paper the Superman role, there's a wide spectrum of responsibilities. Mario Faria last year was on and I'm not sure what was left for anybody else to do because the CDO sort of had that purview from security and governance and on and on and on. Basically the only thing was left is, I guess, IT infrastructure. They usually focus in one item at a time though, yeah. The CDO is there. Right, right, right. They focus in one item at a time. They move around and find a new focus. And do you feel that's best practice or should they take a more systems view and have sort of ongoing attention on each of those domains? Well, in terms of domain, I think it will be totally depending on organization. In other words, if that company, in other words, you need the company that sort of everything is new and you have to pay attention to everything and so be it. But I found that CDOs tend to focus in one item at a time, although they may review other things. So what's going on at the conference for you? Some of the learnings that you've encountered, what's exciting you here? The dean yesterday said, do something different. Do something different, yes, this is terrific. Did that inspire you? Have you heard of the things, other ideas that you're going to take away from us? This was terrific. I probably have too many ideas. But it was great to talk with practitioners who are, as I talk to Paul again about the reflective practitioners. The way they sort of ask questions, I love to talk with practitioners who ask questions or who challenges some of the researcher's papers. So yesterday, for example, somebody said, can that be done by accountant, not CDO? So those are very good questions. Right, so we have a very interesting comment, but what of our guest yesterday, who was a CIO and he said that he's foresees in 10 years from now, we will have many fewer CIOs and many more CDOs. It makes sense when you think of infrastructure now becoming increasingly part of the cloud. Information is the asset. Do you see one position superseding the other? Will the CDO become more common than the CIO? I would say if you ask me, more CDOs will be established, my answer would be yes. And more CDOs established, that answer would be yes as well. But if you ask me, which either CDO or CIO, which would grow faster, I would say CDO. Because I see the need, you can see from this symposium and other forums that I think CIO will stay and their focus might be different, but I think the growth of the CDO will be there. And I hope they collaborate. Well, I think obviously the sentiment in this crowd at the MIT Information Quality Symposium is that the CDO role is emerging. By the way, that's the fact that is, we know that. And should be independent of the CIO. And that's very clear in those industries that I mentioned, the regulated industries, that seems that's a consensus that's well accepted. Outside of those industries, it's not as well accepted. Yes, yes. Do you agree with that? And there seems to be a prevailing, maybe it's not prevailing mostly, but if you even ask, if people have even tuned into the question, if you're like, oh, well, that's the role of the IT group. But the IT group really has never had the authority to implement data governance. That's always been sort of separate, but it's kind of been a back office function, records management. Right, right. And so what's making it emerge now? Is it the big data meme? Is it technology? Is it awareness of data as a competitive advantage? What's, why now? I also think nature of data. Now more people are using data, and the quality of data more exposed or more visible, and everybody is using data. More people will be using that, and data will be used globally, going back and forth. So CIOs role, I mean good CIOs, I'm sure they take care of technology and data. But I've heard that they focus more on systems, so making sure that all the systems and systems infrastructure is done very well. So when I went into hospitals, when there's a data quality problem, and you ask people working in technology division or CIO what's going on with this data, and they will say that is actually the people who collected data, that's the people who made a mistake. And so if you ask collectors and say no, no, no, it's because of the custodians, it's the IT people. Or most often they claim the data users, the data consumers that that's wrong. I guess the traditionally CIOs are focusing on the middle piece, the systems, making sure the system is running right. But there are things that beyond the system that data can have errors and there are many processes, that business process and data process. So that has to be handled more business oriented sort of approach. So Paul, I feel like the infrastructure is relatively straightforward, right? The CDO doesn't want to run the infrastructure, but the application group, a lot of CIOs I know, you can confirm this, it's been a lot of time worried about the application, the business, and that's where you need data to have an effective application, and that's where it gets a little gray. That's right, well that's where the collaboration is important because the application exists to manipulate the data. Did you find in all the CDOs that you interviewed, did you find any common characteristics about their background, where they came from, did they come from a technical background, accounting perhaps, or more of a quantitative background? Were there any patterns there? I think more and more CDOs coming from business background. So it could be they're working in the marketing area, and then they become marketing area, they become CDO, and then they're working in the finance area, and then they become CDO. CIOs typically come from IT background. However, there are some CDOs reporting directly to CIO, they tend to be from IT background. But people reporting directly to CEO, or CO, or they tend to be from business background, more domain background. That is interesting because I think we talk of the CDO in a technology context, and this conference has a heavy technology overtone too, but it sounds like you're saying technology really is not part of the profile. I think people savvy about how we use data, there is community of sort of the people who are data consumers, they know more about data than people who handle system alone. So, I mean, we need a knowledge about system as well as other processes. So somebody who has entire holistic knowledge about all processes, probably that those are the good candidates for CDO. That person's probably already running the company too. Yeah, yeah. Very good. All right, Professor Lee, we have to leave it there. Thank you so much for coming on theCUBE. Pleasure. Thank you. Always a pleasure. All right, keep it right there, everybody will be right back after this word, this is theCUBE.