 From Cambridge, Massachusetts, it's The Cube, covering MIT Chief Data Officer and Information Quality Symposium 2019, brought to you by SiliconANGLE Media. Welcome back to Cambridge, Massachusetts, everybody. You're watching The Cube, the leader in live tech coverage. We go out to the events and we extract the signal from the noises. It's day two, we're sort of wrapping up the Chief Data Officer event, it's MIT CDO IQ. It started as an information quality event and with the ascendancy of big data, the CDO emerged and really took center stage here. And it's interesting to note, it's kind of come full circle back to information quality. People are realizing all this data we have, you know, the old saying, garbage in, garbage out. So the information quality worlds and this Chief Data Officer world have really come colliding together. Robert Abedis here, he's the vice president and CDO of Global IDS and also the co-chair of next year's, the 14th annual MIT CDO IQ. Robert, thanks for coming on. Oh, well thank you. Now you're a CDO by background. Give us a little history of your career. Sure, sure. Well, I started out with an electrical engineering degree and went into applications development. By 2000 I was leading the Ralph Lauren's IT and I realized when Ralph Lauren hired me, he was getting ready to go public. And his problem was he had hired eight different accounting firms to do eight different divisions. And each of those eight divisions were reporting a number but the big number didn't add up so he couldn't go public. So he searched the industry to find somebody who could figure out the problem. I was at the time working in applications and then built this system called service oriented architectures, a way of integrating applications. And I said, well, I don't know if I could solve the problem, but I'll give it a shot. And what I did was just by taking each silo as its own problem, which was what Edie County firm had done. I was able to figure out that one of Ralph Lauren's policies was if you buy a garment, you could return it anytime, anywhere, forever, or long you own it. And he didn't think about that but what that meant is somebody could go to a Bloomingdale's, buy a garment and then go to his outlet store and return it. Well, the cross channels were different systems. So the outlet stores were his own business. Retail was an emperor of business. There was a complete leader. Each one had their own AS for a hundred, their own data. So what I quickly learned was the problem wasn't the systems, the problem was the data. And it took me about two months to figure it out and he offered me a job. He said, well, I was a consultant at the time. He says, I'm offering you a job. You're going to run my IT. Great user experience, but hard to count. Hard to count. So that's what I, by probably 1999 was when that happened. I went into data and started researching. So how long did it take you to figure that out? You said a couple months? A couple months. I think it was about two months. Because it took Oracle about 10 years to build fusion with SOA. Well, this was a little bit of luck. When we started integrating the applications, we learned that the messages that were sending back and forth didn't match. And we said, well, that's impossible. It can't not match. But what didn't match was it was coming from one channel and being returned to another channel. And the return showed here didn't balance with the returns on this side. So it was data problem. Some of the forensics showed that. Some of the forensics. So what did you do after? After that, I went into ICICI Bank, which was a large bank in India who was trying to integrate their systems. And again, this was a data problem. But they heard me giving a talk at a conference on how SOA had solved the data challenge. And they said, we're a bank with a wholesale, a retail, and other divisions. And we can't integrate the systems. Can you? I said, well, yeah, I'd build a website. And make them web services. And now what'll happen is each of those will kind of communicate. And it was at ICICI Bank for about six months in Mumbai and finished that, which was a success, came back and started consulting. Because now a lot of companies were really interested in this concept of service-oriented architectures. Back then when we first published on it, myself, Peter Akin, and a gentleman named Joseph Burke published on it in 1996, the publisher didn't accept the book. It was a really interesting thing. We wrote the book called Services-Based Architectures, a way to integrate systems. And the way Wiley and Sons or most publishers work is they'll have three industry experts read your book. And if they don't think what you're saying has any value, they forget about it. So two of the, one guy said, this is brilliant. One guy says, these guys don't know what they're talking about. And the third guy says, I don't even think what they're talking about is feasible. So they decided not to publish. Four years later, it came back and said, we want to publish the book. And Peter said, you know what? They lost their chance. We were ahead of them by four years. They didn't understand the technology. So that was kind of cool. So from there, I went into consulting, eventually took a position as the leader, the head of enterprise and director of enterprise information architecture with Walmart. And Walmart, as you know, is a huge entity, almost the size of the federal government. So to build an architecture that integrates Walmart would have been a challenge, a behemoth challenge. And I took it on with a phenomenal team. And this was 2010. And by the end of 2010, we had presented an architecture to the CIO and the rest of the organization. And they came back to me about a week later and said, look, everybody agrees what you did was brilliant, but nobody knows how to implement it. So we're taking you away. You're no longer director of information architecture. You're now director of enterprise information management. Build it. Prove that what you say you could do, you could do. So we built something called the data cafe. And cafe was an acronym. Stood for collaborative analytics facility for the enterprise. What we did was we took data from one of the divisions and because you didn't want to take on the whole beast, boil the ocean, we picked Sam's Club and we worked with their CFO and because we had information about customers, we were able to build a room with seven, 80 inch monitors that surrounded anyone in the room. And in the center was a Cisco telecommunications so you could be a part of a meeting. The telepresence. A telepresence. And we built one room in one facility and one room in another facility. And we labeled the monitors, one red, one blue, one green. And we said, there's got to be a way where we can build data science so it's interactive so somebody, an executive could walk in the room, touch the screen and drill into features. And in another room, the features would be changing simultaneously. And that's what we built. The room was brought up on Black Friday of 2013 and we were able to see the trends of sales on the East Coast that we quickly, the executives in the room, and these are the CEO of Walmart and the heads of Sam's Club and the like, they were able to change the distribution in the mountain time zone and West time zones because of the sales on the East Coast gave them the idea, well these things are going to sell and these things aren't. And they saw a tremendous increase in productivity. We received the 2014, the to my team, received the 2014 Walmart innovation project. And that's no slouch because Walmart has always been heavily data oriented. I don't know if it's urban legend or not but the famous story in the 80s of the beer and the diapers that Walmart would position beer next to diapers, why would they do that? Well, the father goes in to buy the diapers for the baby, picks up a six pack, well he's on the way. So they just move those proximate to each other. In terms of data, Walmart really learned that there's an advantage to understanding how to place items in places that, a path that you might take in a store. And knowing that path, they actually have a term for it. I believe it's called, oh God, I'm sorry, I forgot the name. But it's- Selling more stuff. Yeah, it's selling more stuff. It's the way you position items on a shelf. And Walmart had the brilliance, or at least I thought it was brilliant, that they would make their vendors the data champion. So the vendor, let's say Procter & Gamble's a vendor and they sell this one product the most. They would then be the champion for that aisle. So the, oh, it's called planogramming. So the planogramming, the way the shelves were organized would be set up by Procter & Gamble for that entire area working with all their other vendors. And so Walmart would give the data to them and say, you do it. And what I was purporting was, well, we shouldn't just be giving the data away. We should be using that data. And that was the advent of that. From there I moved to Kimberly Clark. I became global director of enterprise data management analytics. Their challenge was they had different teams. There were four different instances of SAP around the globe. One for Latin America, one for North America called the Enterprise Edition. One for EMEA, Europe, Middle East and Africa. And one for Asia Pacific. Well, when you have four different instances of SAP, that means your master data doesn't exist because the same thing that happens in this facility is different here. And every company faces this challenge. If they implement more than one of a system, the specialty fields get used by different companies in different ways. And that's- The gold standard, the gold version. The gold version. So I built a team by bringing together all the different international teams and created one team that was able to integrate best practices and standards around data governance, data quality, built BI teams for each of the regions, and then a data science and advanced analytics team. Wow, so okay, so that makes you uniquely qualified to co-chair the conference. Oh, I don't know about that. There are some geniuses here. I say that because these are your peeps. Yes, they are, they are. So you're a practitioner. This conference is all about practitioners talking to practitioners. It's content heavy. There's not a lot of fluff. Lunches aren't sponsored. There's no lanyards sponsored. It's not like, there's very subtle sponsored desks. You have to have sponsors because otherwise the conference is not enabled and you've got costs associated with it. But it's a very intimate event. And I think you guys want to keep it that way. And I really believe you're dead on. When you go to most industry conferences, the industry conferences, the sponsors change the format or are heavily into the format. Here you have industry thought leaders from all over the globe, CDOs of major Fortune 500 companies who are working with their peers and exchanging ideas. I've had conversations with a number of CDOs and the thought leadership at this conference, I've never seen this type of thought leadership at any conference. Yeah, I mean the percentage of presentations by practitioners, even when there's a vendor name you have an internal practitioner presenting. So it's 99.9%, which is why people attend. We're moving venues next year, I understand. Just did a little tour of the new venue. So you're going to be able to accommodate more attendees. So that's great. So what are your objectives in thinking ahead a year from now? Well, I'm taking over from my current superior, Dr. Arkham Okerji, who just did a phenomenal job of finding speakers, people who are in the industry who are presenting challenges and allowing others to interact. So I hope I could do a similar thing, which is fine with my peers, people who have real world challenges, bring them to the forum so they get to be debated. On top of that, there are some amazing technology change. It's just so fast. One of the areas, like big data, I remember only five years ago, the chart of big data vendors maybe had 50 people on it. Now you would need the table to put all the vendors. Who's not a data vendor? Who's not a data vendor? But so I would think the best thing we could do is find, just get all the CDOs and CDO types into a room and let us debate and talk about these points and issues. I've seen just some tremendous interactions, great questions, people giving advice to others. I've learned a lot here. And how about long-term? What do you see as going, how many CDOs are there in the world? That's a really interesting point because only five years ago, there weren't that many CDOs to be called. And then Gartner, four years ago or so, put out an article saying every company really should have a CDO. Not just for the purpose of advancing your data, but, and to Doug Laney's point, that data is being monetized. There's a need to have someone responsible for information because we're in the information age. And a CIO really is focused on infrastructure, making sure I've got my PCs, making sure I've got a LAN, I've got websites. The focus on data has really, because of the information age, has turned data into an asset. So organizations realize, if you utilize that asset, if you don't, let me reverse this. If you don't use data as an asset, you'll be out of business. What, I heard a quote, I don't know if it's true. Only 10 years ago, 250 of the Fortune 10 no longer exists? Yeah, something like that, the turnover is amazing. Many of those companies were companies that decided not to make the change to be data-enabled, to make data decision processing. Companies still use data warehouses, they're always going to use them. And a warehouse is a rear-view mirror. It tells you what happened last week, last month, last year. But today's businesses work forward-looking, and just like driving a car, it'd be really hard to drive your car through a rear-view mirror. So what companies are doing today are saying, okay, let's start looking at this as forward-looking, a prescriptive and predictive analytics, rather than just what happened in the past, I'll give you an example. In a major company that is a supplier of consumer products, they were leading in the industry and their sales started to drop, and they didn't know why. Well, with the data science team, we were able to determine by pulling in data from the CDC, now these are sources that only 20 years ago, nobody ever used to bring in data in the enterprise. Now 60% of your data is external. So we brought in data from the CDC, we brought in data on maternal births from the national government, we brought in data from the Census Bureau, we brought in data from sources of advertising and targeted marketing towards mothers, pulled all that data together and said, why are diaper sales down? Well, they were targeting the large regions of the country and putting ads in TV stations in New York and California, big population centers. Birth rates in population centers have declined. Birth rates in certain other regions like the South and the Bible Belt, if I can call it that, have increased. So by changing the marketing, their product sales went up. Advertising in Texas. Well, and that brings to one of the points I heard a lecture today about ethics. I mean, we made it a point at Walmart that if you ran a query that reduced the result to less than five people, we wouldn't allow you to see the result. Because, think about it, I could say, what is my neighbor buying? What are you buying? So there's an ethical component to this as well. But that data is not political, data is not, oh, I don't know, chauvinistic. It doesn't discriminate, it just gives you facts. It's the interpretation of that that is hard for CDOs. Because we have to say to someone, look, this is the fact. And your 25 years of experience in the business, granted, is tremendous and it's needed. But the facts are saying this. And that would mean that the business would have to change its direction. And it's hard for people to do. So it requires that. Whether it's called the chief data officer, whatever the data czar rubric is, the head of analytics, there's obviously the data quality component there. Whatever that is, this is the conference or as they call them, your peeps. For that role in the organization, people often ask, will that role be around? I think it's clear, it's solidifying. Yes, you see the chief digital officer emerging and there's a lot of tailwinds there. But the information quality component, the data architecture component, it's here to stay. And this is the premier conference, premier event that I know of anyway. There are a couple others, perhaps, but it's great to see all the success. When I first came here in 2013, there were probably about 130 folks here today. I think there were 500 people registered almost. Next year, I think 600 is kind of the target. And I think it's very reasonable with the new space. So congratulations on all the success and thank you for stepping up to the co-chair role. Really appreciate it. Well, let me tell you, I thank you guys. You provide a voice at these IT conferences that we really need. And that is the ability to get the message out that people do think and care. You know, the industry is not thoughtless and heartless. With all the data breaches and things going on, there's a lot of fear and a fear of loathing and anticipation. But having your voice kind of like ESPN and a sports show gives the technology community, which is getting larger and larger by the day, a voice. And we need that. So thank you. I appreciate that. It's great to have you on. Great to be here. Appreciate the time. All right, and thank you for watching. We'll be right back with our next guest. As we wrap up day two of MIT CDO IQ, you're watching theCUBE.