 From the campus of MIT in Cambridge, Massachusetts, it's theCUBE, covering the MIT Chief Data Officer and the Information Quality Symposium. Now, here are your hosts, Stu Miniman and Paul Gillan. We are back at the MIT CDO IQ, CDO and Information Quality Symposium in Cambridge, winding down our two days of live coverage here. I'm Paul Gillan with Stu Miniman and we've been talking a lot about the CDO role today and there were two panels this afternoon at which that was the topic. Where is the CDO role going? How is it evolving? Will it be necessary in the future? Is in fact the need for a CDO an indication that the company has some deficits in its use of data that has to be filled by a dedicated role? Might that role not be needed? Might CDOs work themselves out of a job at some point if they're truly successful in what they do? Jean Leganza of Forrester Research was a panelist on one of those sessions and the guy who's followed these trends for a long time, over 30 years in IT. Jean, can you describe your role and the big data and how CDO relates to your role at Forrester? Yeah, the CDO role kind of emerged as being I think a key pivotal point of research in my research. I started researching specifically on information strategies around the 2007-2008 time frame. Big data was just becoming a thing. It was still kind of a kick to tire sort of thing, a lot of curiosity about whether it was industry hype or something very real. Although there were people with the Chief Data Officer title at the time, there weren't very many of them and it wasn't something to getting a lot of attention back then. I did some research that was relevant just looking into who's heading up information strategy and information architecture activities at large organizations and started getting some interesting surprises around the fact that in most cases, there was no unified strategy. There was no cohesive effort around advancing data capabilities and that if there were people who were in charge of data strategies, they were often very business oriented, not data oriented. So very early on, I got the inkling that people weren't looking for data technicians to solve the gap between data capabilities and being advanced in data. But they were looking for people who really understood business and could relate to business people. So that was interesting. That continued as a theme if you look into the actual formal CDO role and who's filling it and who's doing a good job at it. Do you believe, we were talking about before we started broadcasting about the data-driven companies and you think of companies, obviously Google is one linked in, but companies like CarMax or Uber where data is really the main asset they have tend not to have a Chief Data Officer. And is that an indication that that role is actually making up for a deficit of some time? Is this a creation of the CDO role, the result of a crisis which now has to be dealt with? Mostly yes. It's always dangerous to make broad generalizations like that, but I think that the majority pattern is that people in some form have shot themselves in the foot relating to data and often can be pointed to an explicit event. There was a gentleman from SAP who did a session yesterday who gave his kind of four year journey around data governance and getting business value out of analytics. And they started with Chief Executive getting visibility around some data quality problem. They said this is not going to happen anymore. And that's the kind of story that's happened a lot. Something bad happened and somebody on high said this won't happen anymore. Or simultaneously there's been realization in business people's part that there's gold in them on our hills. There's a lot of people that start to wake up to what some of us nerds on the IT side have been saying for the longest time that the business should care more about data. Getting the business folks to come to the table and argue about customer data for two hours has been a lost cause. The hype from big data helped people think there's actually something going on here. There's something in it for me. And let's get good at it tomorrow. And the fact is there's a big gap there, right? So when people realize they really want to get better at data and there's a whole lot of work to be done, you can't just buy a package and be good at it tomorrow. They decide they need somebody to head the charge. And the experience is that if someone doesn't own a problem, it doesn't get solved. You can't really do it by a committee. You might need a committee in there somewhere, but someone needs to have resources. Someone needs to be driving the bus. And especially on a multi-year journey, it needs a lot of continuity. And someone who's had a lot of awareness of the problems can communicate well about it. So it's largely filling the gap of folks who see that there's a lot of value in data, know that there's a lot of work to be done to get to where they want to be and need to rally the organization. A lot of it relates to culture change. So you're characterizing CarMax, Uber, folks like that. A lot of the dot-com natives is not necessarily needing chief data officers. That's really true. Basically, the understanding of the value of the data and how it drives the decision-making around strategy and the operating model for the company is really well ingrained. You can't say that about most organizations, and that's the problem. So, Gene, I'm wondering in your research, are there certain mandates that have a high chance that they're going to succeed in certain areas that you might warn people, hey, if this is the problem you're trying to solve, maybe this isn't something that you're going to have an easy time with? Yeah, definitely. The common mistakes are acting as if you have a mandate when you don't have one. I'd love to tell people that data's really, really important. You need to start moving on it now, but I've seen plenty of people just die of gooey deaths trying to get going and getting a lot of resistance in the organization, and the time was just not right. So, in Tom Davenport's characterization of four stages of data and analytics, you might still be stuck in 1.0, and if you're a data architect, enterprise architect, VP of data, you're not the CEO, and you may not be able to actually get the entire organization rallying behind your agenda. Having a mandate from someone at a very high level is actually a key starting point. The second biggest mistake is a phrase you hear a lot at conferences like this is boiling the ocean. Getting your arms around data and improving your data capabilities is really an impossible task. I mean, it's huge, it's really, really huge, and you have to start really small, and there's two things going on here. Fixing data problems in the enterprise is huge, and getting people to rally around data having value and data capabilities having value requires showing results. So, what has to happen in a fairly short period of time, if you're trying to build an agenda and motivate people to follow it, is that you have to do more than say, data's good for you, data hygiene is good for you, good data practice is a good for you, no one's going to buy that, no one's going to change their behavior, unless they really see what's in it for them. Take your medicine. Right, so you have to do pilots, and you have to find areas where you're going to deliver real value while people, in a small way, and then build from there, and use that to market your abilities, and eventually go enterprise-wide, but start fairly small. So, one of the truisms about, you have to have the enterprise in mind, and you can't do siloed kind of work, that the whole problem is everything is siloed as well, you do have to start with very specific use cases to show value, and call that siloed, but you have to kind of dig into the detail of the business problem, which is usually buried within the operation of the business unit to make that work, and then use that to eventually go enterprise-wide and have an enterprise approach, but you have to find real concrete value and show people that there is real gold there. Well, one thing that struck me talking to several CDOs we've been interviewing here over the last couple of days, as well as at the conferences, this is a superhero type of job, because the CDO, in some companies, the CDO is a political manager, is a relationship broker, is trying to get people to give up data that they don't want to give up. So, it's a people skills job, but in other companies, it's a technical skills job. Now you've got that data, you've got to put it into something, you've got to figure out how to make sense of it and rationalize it, and then in other cases, it's an analytics job, it's a forward-looking, a strategic type of role. Can one person do it all, or are we going to have sort of a herd of data CDO migrants who are good at one or two of these processes and move from company to company? This came up in my panel discussion as well, in terms of if you had to choose between skills and roles, which one should it be? My take on it is that you need a team. It's not a one-person job, and the key aspect of the chief data officer role is to be the evangelist, the diplomat, the cheerleader, the program manager, you have to have all those other skills on your team. You have to have the data science skills and the data management and data governance skills and the person who shepherds the data stewards and identifies the appropriate business matter expertise. You need an awful lot of stuff happening and you not only are not only gonna get it in one person, getting it in that person who would be appropriately skilled to be that evangelist and bring the whole program forward is very, very unlike, I'm talking about unicorns, being data scientists, unicorn being signed with all the appropriate technical skills and the communication skills and leadership skills to be a chief data officer is really out of the question. So what you really need to do if you really wanna get somewhere is get that CDO who is a strong leader, a strong front person and can collaborate with a multidisciplinary team to lead the various efforts that you need to do because there's gonna be a lot of things going on at once and in a way the chief data officer is a program manager of program managers. All the folks on their staff should be running various programs to advance processes and organizational structures and skill development and best practices around consumption and use of data and those sorts of things. Great point. So Gene, the CDOs we've talked to have tended to center around industries that have a lot of governance. So a lot of government, we've seen healthcare, we've talked to some university CDOs. Are those the kind of the primary verticals that you see? Where are some of the areas that we're gonna see more growth of the CDO? Well, actually that's two questions. There's a lot of, right now our survey data showed that there's a lot of differences between industries in terms of where there are chief data officers and it's true that the more highly regulated industries are more likely to have chief data officers. And also countries with a lot of government regulation. So for example, North America, this year United States came in with 35% of US companies have chief data officers. UK last year was 40s, now it's like up to 50%. Last couple years, Germany and India and China were in the 60 plus percent companies have chief data officers. So regulation has a lot to do with it. The part I'm saying is different is so those are the organizations that are very likely to have already moved towards the chief data officer to get their data act together to get data quality, data governance together where the growth can be almost anywhere. And I think where we're seeing a lot of growth and chief data officer role in manufacturing in oil and utilities, oil and gas and utilities. Internet of Things is a big player right there, right? So Internet of Things introduces sensors and lots and lots of data and being smart about what to do with that data not only in terms of being able to provide good products and services but also thinking about new products and services and changing operating models requires understanding data and a data agenda. So the growth may be independent of that remedial work, right? So the places that are highly regulated need to do the remedial work, need to do a lot of the work. So they most likely had a chief data officer a couple years ago or are getting one now. I think where the growth is gonna come from are the organizations that are waking up to the fact that there's competitive advantage in doing what Tom Davenport this morning called taking on the offense of data. They're doing all the glitzy stuff around targeting a segment of one, not my population segment but me and my behavior based on that I come to your website yesterday and today and why didn't I buy anything yet and giving me an offer or things from the Internet of Things or products and services around making customers happier are where a lot of the rally growth is gonna come from. So yes, a lot of that requires getting your act together about data as a prerequisite to doing cool things with data but it's not necessarily a case that the folks who have to have had a chief data officer could be highly regulated will necessarily be able to grow the fastest based on that. They also might have some handcuffs in terms of how they can use the data. So if an organization who's not highly regulated but gets their act together about data can maybe start making leapfrogging others by doing cool and different things with data because perhaps they don't have the handcuffs on. So, Gene, I'm curious. You've done some survey work. You've been looking at the space from your research. Is there anything that you've kind of unearthed that doesn't match kind of the common viewpoint that we've been hearing, you know? We had somebody came on and said, you know, saying, you know, CIO and CDO, you know, should be in two different groups and fight, you know, makes a good magazine headline but a lot of times isn't the truth. What are you finding out there that doesn't meet fit with what we've been hearing the last few years? There's actually there's two things. The first one is when you just mentioned, as soon as we started publishing research on the Chief Data Officer role, every reporter who called and wanted an interview tried to make a big deal about the fight between the Chief Data Officer and the CIO and we're just not seeing it. You know, the success stories are all around collaboration. The people who present in panels and in sessions like at the MIT conference are all saying, if you're not in the IT department and that's probably a good idea that you report to the business side, you better have a really good relationship with the CIO and the IT department. There's an awful lot of technology involved. So all the good stories are around collaboration and there's not that much around competition for who's got the boardroom's attention. So I haven't really seen that. The other thing that I think I've learned that actually my thinking has changed very recently on based on a recent study I did was that I kind of assumed that because a lot of Chief Data Officers started in remedial mode, get our data quality at together, they get our data governance together and once we get that down we can kind of move on to doing analytics. And so I thought it'd be interesting to see if several years from now people go from being Chief Data Officers to Chief Analytics Officers or if the title doesn't change but the responsibilities change, all the Chief Data Officers you talk to aspire to do cool things in analytics. So I'm thinking, okay, so what's gonna happen is you'll bake the ability to do data management well into the organization. You'll create the structures, the processes, the data stewards, the data governance processes, the appropriate committees. Eventually in a couple of years that'll be running on its own. You'll have people managing the process. That becomes a background. The Chief Data Officer can move on to be the Chief Analytics Officer and then focus on the cool stuff. My last recent interviews of a bunch of Chief Data Officers turned out the folks who have been the Chief Data Officers have long said absolutely not. And this is because, yes, the analytics and what Tom Davenport referred to as the offense is the glitzy stuff. It's where people clearly see value. Because they don't see value in what he called defense, the processes and the structures that do data management well because it's not glitzy people will lose interest in it. And so if you say, okay, we need to have a champion for analytics and data management takes care of itself. You might have started there because you just took care of data management issues. Two years down the line, you'll have data management issues. It'll decay. You do an acquisition. You got another data management problem. Exactly. And you take your eye off the ball and nobody owns it anymore. You'll just recreate the same problem. So the idea is let the interest, the overriding interest in analytics and the cool stuff subsidize the lack of interest in the uncool stuff of data management and keep those together. So I'm no longer telling people, oh, the Chief Data Officer with a temporary role that will eventually go away. Saying, no, there's really strong evidence because I really believe data will have value to enterprises for the, for the, we got it right. It sounds like a good offense. Your best offense will be a good defense. Yes, yes. No one wants to play defense, but it's critical. We're just getting started here, but we're at time, so we have to cut it off. Obviously a lot to talk about. We'd love to have you back next year on theCUBE and see what are the issues we're going to be talking about a year from now. I'm sure they're very different than the issues we're talking about now. Gene Laganza from Forest Research. Thanks so much for joining us. Thank you. Thank you for having me. Wrapping up our coverage here. Two days at the MIT CDOIQ, we'll be right back.