 Hello everyone, my name is Megan Jacobs and I'm the Executive Editor for Data Diversity. We would like to thank you for joining today's Data Diversity Webinar, Show Me the Money, Monetizing Data Management. This year's October edition in a monthly series called Data Ed Online with Dr. Peter Akin brought to you in partnership with Data Blueprint. Next we have the floor to Megan Jacobs, the Webinar Organizer from Data Blueprint to introduce our speaker and today's webinar, Megan. Hello everyone and welcome. My name is Megan Jacobs and I'm the Webinar Coordinator here at Data Blueprint. We hope that you found the time to join us for today's Webinar on Show Me the Money, Monetizing Data Management. Thank you so much. A big thank you goes out to Shannon and Data Diversity for hosting us. We'll get started in a few moments after I let you know about some housekeeping and introduce your presenter. We'll now have our presentation followed by 30 minutes of Q&A. We'll talk about as many questions as time allows, but feel free to submit questions as they come up throughout the session. Now the answer is top two most commonly asked questions. Yes, you will receive an email, links to download today's materials, and any other information you request during the session with the next two business days. And on Twitter, Facebook, and LinkedIn. Instead of going back and forth on Twitter, so if you're logged on, feel free to use it in your tweets and submit your questions and comments that way. We will rely on the Twitter feed and we'll include answers to those questions in our post-session email. To introduce you to our presenter, Peter Meakin is an entrepreneur that I recognize that data management is always important. Many of you may know him or have seen him at conferences worldwide. He has more than 30 years of experience and has received many awards for his outstanding contributions to the profession. Peter Meakin is the founding director of Data Blueprint. He has written dozens of articles and eight books. He has written nine books now. The most recent one is Monetizing Data Management. He has experience with more than 500 data management practices in 20 countries and consistently as a top data management expert. Some of the most important and largest organizations in the world have sought out his and Data Blueprint's expertise. Peter has spent multi-year immersion with groups as diverse as the U.S. Department of Defense, George Bank, Nokia, Wells Fargo, the Commonwealth of Virginia, and Walmart. He has been working with conferences and is constantly trilling. Peter, where are you at this week? I'm in a closed location in a hotel room somewhere, Meakin. It's the first time I've actually had to say that. So we'll jump right in here. We're real excited today because this is a literally finale for this book. We weren't sure we were going to get this thing done. And I want to give a shout out to Shannon, of course, as well. We'll head up on the website very shortly so you guys can hopefully get a copy of this and give us some feedback on it. But this book was literally finished on Friday, and it was up on Amazon on Monday. So it's wonderful to be working with a team who's able to put this type of thing together, and I appreciate everybody who's been helpful in the process. I called it Monetizing Data Management. Luckily, I have some very smart marketing people who work for me. That's a nice topic, but you really need to call this Show Me the Money. In addition to that, you should have seen the cover I was going to put on the book, and this is the cover they came up with. Again, thanks also to them. Also thanks to Juanita and John Botega for their contributions to this as well. It's just been a real fun exercise working for us. I'll tell you a little bit about what I'm doing and what's going on on this topic, but more importantly, I also want to invite all of you listening out there to share stories of this type. And there's a forum for sharing those stories, and that's something called Enterprise Data World. It's a conference we run every year in conjunction with Data Diversity, and this year it's going to be in place with Austin, Texas. So it's one of the better parts of Texas, and we're just still looking forward to it at the end of April, and the reason I'm mentioning it right now is because the call for papers is out there right now. And this is how this has got started with people submitting stories, coming and sharing their experiences, and helping everybody to become smarter about this entire profession. So if you have trouble finding it, just go to cdw214 and you will find all kinds of hits on how to get there and make your contribution, or you can ask us some questions, and we'll come back to that after the end of the presentation. So with all of these things, we start them all out with a data management overview just to make sure that everybody's on the same page, literally, on this. And then we're going to talk about both motivations and survey results, and I'll look at some specific examples of how we've really got to change our talk about this, because if you ask people what you do and you can manage data, it sounds like you're an accountant. I don't mean that in any disrespect. I've been one for many, many years. But if you tell them to leverage data, they kind of get interested in that. And if you think about it, we don't manage data for the sake of managing data. We manage data for the sake of leveraging it. So if we start to think with the end in mind, or as Louis Broom, who's our CEO always says, give them a why first, because that will allow them to really understand the motivation of what we're trying to do. Then we're going to dive in, and I'll try to get in six quick cases on the return on investment and two cases for non-monetary ROI. In this case, it turns out saving lives is also important to some organizations. So we'll talk about how data management has helped out in that context as well. Finish up with a little bit on some legal considerations. Of course, as always, we look forward to the Q&A takeaway on this. So let's get started talking about data management as being five integrated practice areas. As shown on this chart here, it's a very difficult chart. We don't expect anybody to get it, particularly management, right off, just because it's a lot of material to absorb. But the real key to take away from this chart is that these are five interlocking chain links. And if one of the chain links is broken, non-existent, or weak, the entire practice area can only be as strong as the weakest link in the chain. The five chain links then are managing data coherently, making sure that we're all singing all the same sheet of music. They're all working to the same effort. And yet, when we go into organization after organization, we find really dedicated people that are working in work groups, not at an enterprise level. And oftentimes, these work groups are working at competing purposes for these areas. And that's kind of sad because it means you're not interacting with your efforts. You're taking valuable organizational resources and not using them effectively. And we do have instances where companies have gone out of business for bad practices in these areas. The second link chain is organization integration. This is the idea that in your organization, whether you are sharing data from a program to another program, from the marketing organization to another part of the organization, perhaps manufacturing, or whether you're sharing data between your organization and other organizations that are your business partners, that data sharing is typically not done in its most effective and efficient fashion. And consequently, there are savings that can be done in that area. And again, all the strategic leveraging that can occur as well. Our third area is data stewardship. We've learned a long time ago that if we didn't make this personal, if it doesn't say, Peter, your next raise depends on you improving the quality of this data by X, Y, and Z. A measurable amount within a certain period of time in a very discernible fashion. Then it becomes everybody's problem and nobody actually does anything to it. This is what started the entire data governance movement. And again, 10 years ago, we might have been talking about data governance. People would have kind of looked funny. Now we have entire tracks at Enterprise Data World devoted to governance and governance stories. In fact, an entire conference around that occurs twice a year. Our fourth function or fourth link in the chain is data development. And this is the idea that we need to add your specific data delivery systems in the past. And unfortunately, the colleges and universities of the world have taught people that the answer to this question is the relational database system. It is certainly part of the answer, but it leaves out a lot of other pictures including virtualization, portals, XML, and all sorts of things that we now call big data techniques. And their final area, our last area is data support operations. This is the idea that we do need to maintain data availability. It's not like a car, excuse me, it is like a car. You need to be able to drive it off a lot, but you need to replenish the fluids occasionally, change the tires. We can't simply let it run. It is something that requires active, proactive participation so that our data assets will be supportive of our organizational strategy. We look at the same time about Maslow's hierarchy of needs. And if your food, clothing, and shelter needs are unmet, it's very unlikely that you will be doing self-actualization activities, whether that includes musical instrument or something like providing a great American novel or whatever it is that you do for self-actualization. And the data management, it is no different. The thought management practice areas that I've given you up to this are necessary but insufficient prerequisites to all the really self-actualizing fun stuff that goes on in organizations. These are things like master data management, data mining, analytics, housing, et cetera, et cetera. The point is if you're going to do something in that green triangle, this actualizing triangle, you can do it without the benefit of the five data management practice areas. If you're willing for it to take longer, cost more, deliver less at greater risk to the organization than if you instead learn to crawl, walk, and run your way up to one point. So that's our first little overview just to make sure everybody's on the same page. Again, love the book on this. What we're going to do as part of the process is help us in the community better articulate what it is that we do and why it's important for management to pay attention to it. So one of the last two books that I've done have been short books, less than 100 pages each, in hopes that we can communicate this information to people who are very busy, who are smart, talented executives who don't know that they don't know something, and more importantly, they don't know that their IT people also don't know something either. What that means is that we have very smart IT people and very good executives making very poor decisions about data, and that is a problem for many, many organizations. The book is divided up into four parts. First of all, the first part is about this practice of leveraging data. Then I present 11 financial cases, five non-monetary cases, and then a couple of legal practices and IT practices in there in order to do this. Now, we've done this several times as we're leading up to the start of the seminar today, and Shannon was helpful and sent the link out a couple of times. We've done a lot of research at Data Blueprint trying to find out what's actually happening out there. And one of the questions we've asked in our survey is this one. These are preliminary results. If you haven't had a chance to take a survey, please do click on the link below, and go ahead and take it right now because we're trying to get to statistical seconds over the statistical significance point. The last question here says, how does the organization define the value? And you can see that in our results, customer satisfaction seems to be the edge with profit and followed closely by quality. Then we ask the question, what's the most important goal for your data management projects? And you can see that is a lot of organizational capabilities. 80% of the organizations have cited that as a very key piece of information. Then we have improving efficiencies and then aligning with strategy as our third piece in there. I've actually written an academic paper that addresses that in a little bit more detail if you'd like to learn more about it. Just ask for it at the end. We'll make sure it's included in the package that gets sent out after the webinar. Another question, what percentage of your data management projects are successful in there? The last question was, were 75% of them successful? 30% of the people said yes. 50% of them successful? 27% said yes. Were 25% of them successful? 20% said that. So we do have a challenge around our data management projects. We want to try and get better about them. There are unsuccessful data management initiatives. What were the reasons for failure? Lack organizational maturity seems to be a 70% category in there. Advocacy, which means you can't just start the project. You have to keep working on it. You have to keep providing something that my colleague John Lathley calls, sustainment, a real life cycle of the project. And 41% said lack of alignment with business goals. So again, these are good pieces of information. We'd love to get some more from them. So again, please do do this. We will be putting out a white paper. And if you're out of results, you can register at the survey and we'll be good to extend the copy of these results to you and hopefully be able to participate in future efforts. In question about funding, you can see that's just sort of weird because there's no value. There isn't a precedence. Or again, there doesn't seem to be much problem with saying these data projects need to be aligned with business objectives as well. And then we ask the question about data-centric thinking. And I'll explain what that is specifically. 70% of you said no. 30% of you said yes. So those are the survey results that we have for you in a preliminary fashion. We will publish the actual results when we get it all the way up. Let's take a minute and take a look at what we mean by data management first. And I always ask groups whether they're number 42. And people say that's the Jackie Robinson's Jerry Wright great movie on that. And certainly it's my icon in that. If you go back a little further in history, there was a great book called Doug Adam Seikir's Guide to the Galaxy. And 42 was the meaning of life, the universe, and everything. So if you've got nothing else out of this webinar, you now know what the meaning of life, the universe, and everything is. And what I've just done is I've combined a fact with a meaning. The fact that 42, the meaning is the meaning of life. 42 is my age 12 years ago. That may not be relevant to you, but it's a different fact and therefore a different piece of data that comes to play on that. Well, we'll start to look at what we're going to do from a delivery perspective. We'll compare data with our anticipated requests for information, and it now becomes a higher level of construct here. The improvement is a more refined concept than specifically the data. And if we get to intelligence, then we really get smart about it because then we've looked, we've observed the cycle, and we know how to strategically use the data. Now my point in describing this structure here to you is because it means that we have an architectural construct where the necessary requirement of putting data together for that information in order to get to strategic use of it is very, very insufficient condition. This also then leads us from an architectural construct into an industrial construct. And leverage, I've mentioned it several times so far, data leverage is an engineering concept. If we take a very large object on the left-hand side of your screen and make a very, very, very long lever, a very small human being can move a very large amount of bulk with a per-set of engineering concepts. That's what we're trying to do with data. What this means is that data leverage is specifically an engineering concept, which most people are not trained at. So we look at bulk and replace it with our organizational data, add people to the process, the organizational data managers that are working on this. And we then complete our triangle with technologies, the lever as well as the fulcrum that we have here, and add a process to this, which is that we shouldn't try and pull on a lever until it's properly in place. We can now start to see how data engineering is going to move mountains of data for organizations. One additional piece on this particular part as well is the concept of ROT. Money is redundant, obsolete, or trivial, and if we reduce the amount of ROT, our leverage increases. But data is critically important today because it leads directly to ineffective use of manpower, money, methods, as well as machines. And as I mentioned before, we have real problems in the educational area where we teach new students how to build new systems. Most of our work is not about building systems, and the last people we would put on new systems development are inexperienced young people, but we're teaching them incorrectly in the college university system. Business, things we're taking care of this because we call it IT. And one of the things we try to do with our customers is to convince them that unless they're really using an I component in there, most of what they call IT is really just T. Now, there's another very important point to this, too, and that is that data development activities are different from system development activities. I'm going to show you a specific chart on this here, which is to say that our system development activities are fundamentally oriented around creating capabilities where capabilities did not exist. However, our data is a different process and instead of being a creation-oriented activity, it is an evolution-oriented activity. So our rule here is that data evolution is separate from as we do and must precede system development activities. In fact, I have a rule that says, you can't tell me how many existing data elements your new development activities are going to reuse and how many new developments need to be created. You don't have a good idea of what your data requirements are on the project and therefore should not be spending any money on that project at all. The reason for this, again, is because we've built things and told people to build things incorrectly. We started out in, say, strategy and you heard about the alignment pieces earlier on in this hour. We talk about strategy, which leads to goals and objectives, and a very natural step is let's make some systems that come out of that. For systems, then we start to develop networks and data becomes an afterthought. I'll give you a very specific example here and say that the system that we're going to put in place is a software package. So we buy the software package, we buy a network to support the software package, and at the end of that discussion, we have a question, what data does that software package need, which is a very different question from what data do we need to support our strategy. Problems with this approach are absolutely and guarantees that the data is going to be formed around the applications and not around organization-wide information requirements, that the processes are narrowly formed around the applications and that the little data reuse is possible. We end up with a situation like this. If you follow the typical application-centric development path, each application ends up with their own stovepipe of data, which is how it is in your organizations, and when we try to connect them, we end up with a wonderful web. The guarantee is to keep somebody employed, but certainly not the best way for us to approach this. So we'll pause here and talk about the alternative for the development. In this model, let's see what the objectives occur. We have a strategy when we want some goals and objectives, but it's dead next of those two. We'll define the organizational information requirements. These organizational requirements then derive from the network infrastructure components, which finally draw system and application efforts. I can say that we have saved organizations millions of dollars in wrong software because they didn't understand their data requirements. Once you understand your data requirements, you can then determine whether the systems that you're being able to evaluate in fact meet and complement your organizational strategic objectives. Following this approach, the data-centric approach, means that the data and information assets are different from an organization-wide perspective. At the same time, support the organizational data needs and complement the existing process flows, and this gives us maximum data and information reuse. If you go and look on the web at the website, the primary motivation behind object orientation has been software reuse. And outside of a few narrow constructs, such as families of software and operating systems, software is not being reused. What is not being reused is because we don't have good ways of understanding it and understanding it. It turns out those are the fundamental attributes of data, so we should now put our attention into reusing data. Also, I hope that you can see on here as well that the pathway from data information to systems and applications that I'm showing on the right-hand side of this chart are the things that will leave clean systems design all the way around. If you're able to obtain cleaner system design, it means they're easier to maintain and easier to put together. And now we can finally achieve the process of some of the wonderful technologies in that green triangle I showed you before, such as service-oriented architectures, et cetera, et cetera. This leads us to our first polling question. Those of you that have been on these seminars before know that our ongoing quest for knowledge, we want to ask questions. And one of the questions we want to ask you all is, who in your organization makes the decision to invest in data management initiatives? Hey, guess what? I get to vote on this one. So we're going to take a minute here, actually, 50 seconds, if I recall correctly, from Chancellor's guidance, and ask you guys to take a quick look at this. Oops. Keep that up. I'm going to try to vote. I don't know whether I can vote or not. I'm going to vote. Uh-oh. I'm not going to let that pop up. We have a few seconds left and get everybody's responses in. All right, so we have all the results in. Let's see. Okay, so Office of the Chief Data Officer and Enterprise Data Office Equivalent, 31 percent. IT, and it's worded by the business team, the other is 7 percent. Wrong. Okay, never mind. Starting from the top. You guys can all see the results. IT, business data, 10 percent, both of them together, 30 percent's the way it should be. Not enough in there. Good. Well, again, thank you for participating in that. Again, we're using this as an ongoing process to try and figure out how we, everybody in the community, get better about what we're doing. Now, as I mentioned before, I'm going to briefly go through these cases. Each of these cases is in the book and in much more detail. And of course, when we get to the Q&A section, we'll be very happy to explain them in more detail. But here's a situation where an organization was looking at how much time was being spent by various units of the organization doing time and leave tracking. Now, in this, we took the employees, which is on the watch list here, and we took them by district in this case and looked at the specific pay grade. So one of the things you can see is that there are two grade 13s that are spending time, actually it's more than two, spending time and leave doing clerical type functions. And in the Lynchburg district, there's a grade five that has a lot of people doing it. We were able to come back and say authoritatively that at least 300 people in this organization were spending 15 minutes a week doing time and leave tracking. So what do I do with that information? Instead of going around and finding out their actual salaries, I can get the bable salary of a grade 15. How about the number of 15s that we're doing this? The bable salary of a grade five and figure it out. This means that I can put together some change like this. But just to say in district L, 73 people were doing time leave tracking, 50 people were doing time tracking. We can put in the cost of these timekeeping, we can come up with a semi-monthly cost of $21,000 and a monthly cost of $131,000. Which means when I went to add them all up across the entire organization, I came up with a total that it was very close to $2 million annually. That's a tremendous amount of money for an organization to spend. And once asked the question, is there actually a better way for the organization to do that? Now contextually, this occurred within a general organizational systems rethinking. And what they did is they used this information to make some cost-benefit trade-offs and said, how much time do we actually benefit from people doing that type of activity? And can we make this activity to the point where it doesn't cost the organization tens of millions of dollars? And so that, again, very quick sort of warm-up case on this just gets us to $10 million. Here's another one that takes us to $25 million. Now, our community and plus-dollar chemical company does fuel incentives to enhance engine performance. It helps the fuels burn clean, smoother, and last longer. And they run tens of thousand tests annually that cost up to $250,000 a piece. It's kind of fun because when the team went into this particular organization, we told them they might have to wear lab coats and they were immediately thinking, medical procedures or something like this. And we said, no, we don't want you to get oil on your clothes because when they talk about testing engines, they take the engines downstairs and they turn them on and let them run. And it comes up with a marvelous set of requirements in here. So when we look at these data management practices, I want you to picture a room full of chemical PhDs and each of these individuals are getting paid six dollars. These are really smart scientists who really do understand what they're doing and how they're adding value to the organization. And yet, in spite of the brain power in this room, we found instances where PhDs were taking digital data from computer system number one and turn it on and re-keying this information into data system number two. Now, many of us on this call could have absolutely figured that one out for them. But it didn't occur to them that there might be a better way in order to do this. We ran into many instances where we were doing medical file manipulation where they were duplicating data. They were taking FNMs back and forth. They were requiring tribal knowledge and most importantly, they were using what we call non-sustainable technology. So again, you can see we highlighted each of the instances that we were looking at here. Now, I have to tell you, right here, the organization said, wow, you can stop. You've already earned your money because we've never had a good window into what our people actually did as they were researching these chemical compounds and things like that. So they were grateful just for the chart. But when we looked at them also, again, that they were doing this and using a database that hadn't been around, that had actually not been used, that was not manufactured, was not supported, for a long time, they realized there was a significant amount of risk that they had in their organization. And then they went on to a conclusion that they said, you know, maybe this isn't the best way to go into doing this. So they did this process for them. We really came up with a measure that showed that their knowledge of workers, these chemical and GHDs, were spending approximately 80% of their time doing work which should have been done by somebody at a much lower set of K-grades and, importantly, by somebody who was a specialist in this area who could do it more effectively for the entire group. So instead of the entire group of 100 of them trying to do all this collectively, very poorly, we added one more person to the group who was expert at this who was able to increase the productivity of the group in a rather dramatic fashion. In fact, because we were 80% unproductive and 20% of their time was actually devoted to analyzing these chemical compounds that were the strategic needs of the organization, we only had to make them 60% unproductive in order to achieve an actual doubling of their productivity, which went from 20% productive time to 40% productive time. They told us that the year after this was the 25 million dollar improvement in their productivity. If they look at these results, they were able to reduce expenses and build their competitive edge and close the money on the table, in this case, because I guarantee we didn't charge them 20 million dollars in order to come up with this. Again, hopefully on this, but hopefully this is going to make sense for you all. Actually, I'm going to talk about the category discussion about the word tank. Now I think the tank is a critical word, but it turns out the tank is kind of important, and here's the scenario. There's a company who's into petroleum requirements where the tank will come from, and they bought an accounting package. And the accounting package that they purchased treated every transference of petroleum product from one tank to another tank as a retail sale. Now they're going, okay, that's really not the way it works, because we have a lot of things that occur in the petroleum manufacturing process before we get to the final sale. So the company had to make a decision of whether to change to alter to customize the accounting package that they were looking at, whether to work well for them, or to apply a form of data governance that controlled the vocabulary around the word tank. Now let me give you a brief analogy. Many of our organizations have a customer as a concept. We want to loop around, and absolutely avoid trying to make our short clients avoid using a concept as generic as customer. Because we want to get to the nuances of what a customer is. A customer can be a current customer or a former customer. And you want to make the same communications to your current customers that you would to your potential customers or your former customers. For example, you wouldn't want to tell your current customers that the thing they bought yesterday is now on sale. And yet that might be exactly the message that you want to send to a potential customer. So controlled vocabulary is a very important thing. So alternative one, this oil company could have changed customers that had changed customized and that had heavily in customized accounting package to work well for them. The last option, the one they chose, was that they said we will control the vocabulary around a tank using proper governance. In other words, this tank represents retail sales. In other words, this tank did not present a retail sale. This tank or this tank, oh wait, that's a completely different type of tank entirely, isn't it? Let's go back to that tank in just a minute. So you can see that the controlled vocabulary solution allowed this organization to save millions of dollars of work comprising the software that had to be applied to the software every time the manager upgraded, released a new version of the accounting software and that occurred nearly every single year. So again, controlling the vocabulary around a tank, save the organization from the horror of having to get into customizing an accounting package and kept them from confusing the various types of tanks so they could know what was a retail sale, what was a retail sale sale, and perhaps more importantly what was not the type of the type that they were looking at. One more, this fellow is tanked, we won't go any further than that, but let's move now to another story, which is a branch of the Armed Forces by heavy equipment in this morning, a tank, a tank also comes with 3 million percent of the resources of data for the organization. That's a lot of information. And just out of curiosity, how many of those pieces of information do you think will be the obsolescence or the useful lifetime of that heavy equipment? And if you don't know the answer to that, you tend to treat all 3 million values of equal importance. However, as you'll slide, if you are maintaining obsolete equipment, clearly some of those values are more important than other values that are in there. So the organization, by imposing good data governance around their tanks, in this case, a different type of tank entirely, we're able to look and think that they had in fact found building dollars worth of equipment that would in fact be obsolete, which meant that during the times of war, the 5 million dollars could be in place into a set of operations that were much more valuable than many things that they couldn't even use eventually. Large organizations have all kinds of problems like this, whether it is tanks, whether it's heavy equipment, anything that's a depreciable asset in many cases is subject to this type of analysis and understanding here. And just getting this process in place allowed the organization to see that not all of their data were equal. With the 3 million values, there were clearly some that were much more important than others. And understanding which ones were more important and which ones were not, we were now able to find in fact a better process for managing these things so that the armed forces in this case would no longer house, maintain oil, repair parts for, move them around, or count on these tanks during times when we had other needs for those particular resources and resources. This is an example here. We have a challenge in the organization where there are millions and millions of NSNs, national stock numbers, for SKUs, shopkeeper units that were maintained in a, quote, electronic catalog. The problem is that this electronic catalog, while it wasn't a database, just perfectly clear it was an oracle database that had been engineered to work as a hierarchical database. The reason we re-engineered the oracle database, which is a relational database, to work as a hierarchical database is because the programs accessed a hierarchical base and they didn't have to change those programs in order to swap the data out. So somebody told them they should upgrade their database. They really did, but it was kind of like, don't really have a good word word. From that perspective, as we move into a new organization package, a new software package for this, so this is the next one. People will look at this and, my goodness, we're going to have to manually extract this text. Structuring problem unresolved. You can hear the terms unstructured and structured used frequently, we prefer the terms tabular and non-tabular, because truly, if something was unstructured, that's the definition of unstructured data. You can't add structure to it. So we wouldn't be talking about transforming unstructured data. However, we can talk all day long about taking tabular and non-tabular data and converting them back and forth, right here. We think for them a proprietary, improvable text extension process, we'd call this text mining now in this to convert the non-tabular data into tabular and save them, in this case, $5 million off of the original estimate. Now, that's the thing that we're most proud of was the fact that this was the first time we had actually saved an organization a person's century of work. You for a person's year or a person's month or a person's day, here's a person's century. Let me show you how we got to these calculations. The first thing we did was look at how much we could actually do with our text mining process. Now, the key with this automation is knowing when to apply it and when to switch back to the manual process. Here's how we figured this out in conjunction with the customer. We found an 18-week process of which we held the hand side fixed. It was the cost of two of our engineers working on the text mining process, half time for each week for the total time that we went through. When we held that part firm, we were able to see what sort of results we got and we could identify the place at which the organization was experiencing to making returns. The point at which the returns were less than the cost of the investment and that's, of course, the place that you want to stop when you're trying to occur. Let's see how we did this. First of all, you can see the first week that we did this. We were kind of mad. We didn't actually match much of anything. And this is an important process. You have to make sure that people have an expectation. And we told them, it's going to take us several weeks before we fine-tune the algorithm and actually start matching things. That time, by week four, we had actually matched 55% of the things we were able to go in, read this clear text field with our algorithms and extract it, match it up against the goal and master what would be now called master data management. So in fact, I was able to solve more than half of the problem just after four weeks. We also had been able to determine that some of the items in these fields were ignorable, that they had no business value. The first week, we had 1% and that number grew to almost 12%. You're just a hair under 12%. So if you had the 12% but 55%, we actually had two thirds of the problem solved by the end of the fourth week. So the question became not, is this a good approach? We clearly had demonstrated that, but how much further could we take it before we turn it back over to the manual approach? Again, it was quite easy. With the unmatched items, you can see they went up and down a bit as we got better and worse at our algorithms. Now, the process involved interviewing the subject matter experts, helping them to help us code different algorithms, put them in place and apply them. I'm going to have time here to go to week 14. You can see that the number of unmatched items had dropped from 32% down to 9% of the items. As you watched, from 14 to 18 here, we went up a little bit. It went to 9.6, 9.5, and then back down to 7.6. It should have only been going down. And in this, we found errors that had been given to us and we had coded those errors incorrectly. But we did finally get at week 6 down to 7.5% of the old problem space was still unsolved and that we had by week 14 gotten to 22% of the fields containing absolutely nothing of value. So one fact of the data was in this case, rotten. We're not able to use it at all and our items matched increased gradually to 69, just under 70%. So when you reached the 69.22, leaving just 7% of the original problem, this was clearly the stopping point, the point at which we should stop working with this and back over to the manual process that was originally supposed to occur. Now, let's calculate the return on this investment here. Again, there were 2 million NSNs, two million stopkeepers unit. And if we put in there an average time to review and cleanse them in five minutes, give them 10 million minutes that we needed to do, we then took the time resources available of the year, number of weeks worked in, seven and a half hours per day, 450 minutes per day, minutes per year. You get the point. We did it all up and we came up with 92.6 person years of a person's century on this. Of course, we then apply a measure this organization valued their SME time at $60,000 a year. That's $50,000 here. Now, one of the other tricks that we do on this is a little bit of social engineering. The third measure that we are in fact underestimating the cost of this, and we would always point out during one of these meetings, can you clean these things up in five minutes a piece? Again, the end of the course, most of you know, is absolutely not. Five minutes is an absurd underestimate for this. So if we just double that to 10 minutes, we'd have two person centuries and $10 million, 15 minutes, three person years, and $15 million. Once again, hopefully you can see that there's probably a value to the organization applying good data management technologies in this case to this problem to help the organization save millions and millions of dollars and years and years and years of value-added person effort. One more quick question on this. One more example here real quick. This is a very small investment. I'll tell you what's going on in this. We're going to roll out a new master data management strategy. And in order to do this, we wanted to communicate the sense of this project to everybody in a clear and understandable fashion. So here's a little bit that they invested approximately 500 British pounds, not a lot of money, and came up with a little flash animation that gave a bit of information to all of their associates. It was sent to them in an e-mail. They were able to attract the number of people who were in the e-mail, played it more than once, and came up with it. So here's the whole thing. Oops, I hit the wrong button there. Sorry about that. Let's try it again. This is actually in stereo as well. It's quite good. It's a brutal piece advertising what they're going to do. And they double-checked across the organization. They were able to find out that a lot of people remembered which of the seven stacks they remember the number of the seven stacks that they were doing on that. It was a great investment in a small amount of coding and animation to convey a very complex subject to a vast number of business, as well as technology people. And again, I just think that's a marvelous piece. We have that on our website at Data Blueprint, and we'll certainly post the link to that later on afterwards. I encourage you, they encourage us to encourage you to download it and take a look at it and perhaps play it for your own organization and see if an investment like that might also pay off similarly in your environment. So we get to polling question number two. Again, we want to know how hard it is and what sorts of problems are you running into obtaining funding for your de-management project? We'll help us out by voting on this. It'll be perfect. We have about 20 seconds left on the poll. Let me get your guys' answers in and then we'll see what comes out. Some of you are not answering it, because we're not giving you the right choices. Most of you say yes because it's hard to show value. Again, that's what we need to do here. So please let us know whether these examples are giving you some inspiration or not. 8% of you said yes because we've not aligned with business objectives. There isn't no present investment set. And by the way, you have the ability to clearly demonstrate value. And that is phenomenal. Again, I would encourage you to take your questions to Enterprise Data World coming up so that we can get some of that experience and share it back and forth across organizations, because if you have something out there that we don't know, it's something we'd love to have shared amongst us in the data management community. So thank you again for participating there. Let's talk about a really sad episode that occurred in the early part of the Ask your questions. The answer here is that a lot of people are using a device to point out a target. It's called lighting up the target. So I have a small room in a nondisclosed location. And if I took my laser pointer and pointed it at a building across the street here, then the air force that was bringing in the bomb would take that building and drop the bomb on that. Hopefully it's a place where bad guys are and that sort of thing is on it. But in the process of doing that, they also double check themselves. And they have multiple people live there. There's lots of safety checks and procedures in place to make sure that nothing bad happens as a result of this one. In this instance, in a couple of weeks, they had to change batteries on one of the units that they were doing. And when they did the unit, lit their own position up instead of the position they had been formally dropping in the bomb, dropping on their own troops. They're not well known. It didn't happen a lot. But it's just in our mind, absolutely terrible that something like this would happen. That data management system was set up where we lock in on a position and transmit it, and then check the coordinates, not alerting the operator that something had happened. Our smartphones are better than this. We have the right to expect in the field that this also will be the case, as well. So another part of this is that we at Data Blueprint were very fortunate to play a small supportive role in the Suicide Mitigation Project. You all probably know the terrible statistics now that more of our soldiers are losing their lives with their own hands and they are to bad people doing this. And in the past long story short of doing this work, we ended up doing a lot of data mapping and trying to find out how different data was used for certain things. This is just a notional map again, these are actual artifacts from the project that we're on here. And we would end up with these situations where we have a group that we call the Council of Kernels where we have a lot of people come into a room and say, you can use my data for this purpose and you can use my data for this purpose and my data can be used to help do this, et cetera, et cetera, et cetera. But they were mostly presented in a non-helpful manner. In other words, my data can only be used for this purpose. So we were very fortunate in that we had a senior Army official who came into the room and said, ladies and gentlemen, I've heard my data enough out of the area of the Army and it's all my data from this point forward. So if you have any questions about that, you are welcome to make an appointment and come talk to me about why your data shouldn't be used to support our troops. Justice to the emphasis with which this message was delivered. But you get the picture here, this empowered the team and the conversation turned from can this be done to how are we going to accomplish? It certainly said that mistakes would be tolerated along the way, but it allowed us to quickly put up a prototype where we can start to analyze the various communication patterns from our forces so that we could then try to make early intervention possible that we would again save lives under these circumstances. Then there's an instance of an Army official taking charge and saying, you know what? I know you all think that your data use is important, but there is a strategic use here that we need to do, which is save the lives of our war fighters. And that is going to be our mission and we're going to do it in a way in which I prescribe. And it made a difference. I have talked to dozens of corporations and told this story and said to the CEOs and the boards of these organizations that if you were willing to make a similarly strong statement, I'm willing to work on a percentage for you because I can guarantee you if you would take this effort just as we did here in the military and save lives, I can save your organization tens or hundreds of millions of dollars. And all I want is one percent. And we would be doing much better off all the way around. But for some reason, many organizations are reluctant to make this investment in us. And so the question now, again, we asked you all similar way off the survey here, what percentage of your data projects are in fact successful here? Because if you know these statistics internally, you can go into it with your eyes wide open the next time people try to do this. So again, just very quick because our last one here is we're headed up to the top of the hour. And I'll tell you one more little story and then we'll get to the Q&A portion of our session. Again, what percentage of your projects are successful here? Ten seconds, less than the clock. And it looks like most of you guys get answers in. So we'll go ahead and get those answers up in just a few moments. Yes. Yes. Seconds. We're going to keep you guys online here for 10 minutes while we compile the results. I'm just kidding. We're not that good. All in. You should be able to see the results. Some of you are just not able to give us these information, but 10 percent successful is the largest number there. Boy, that's to get better than that. And then again, we hope that these pieces of information will help you to move onward on this. We include these survey results as the final communication here, too. So I'm just going to move on to the next part, which is just very briefly. We do a fair amount of expert support here. And I'm just going to tell you a little story particularly, but Company X says we're going to implement. I actually got a director from Hickord who said, you need to implement the software package. And they said, well, we'll go to the software package and ask them for a preferred specialist who's expert in software in this particular business. And with the preferred specialist they began implementation for a six month effort that was supposed to be finished by December 31 so they didn't have to use two systems in the next year. They realized in January that they had missed the milestone and then in this case said, well, your data was bad. Now, if you don't know what your data looks like before you let somebody convert, how are you going to defend against the charts that your data is bad? We had arguments. It went back and forth but they worked actually another six months on the project and did not solve the project. So when these things go to arbitration in this case, because that's how most of these clauses are governed as an arbitration clause, they go through a process of producing expert reports. Questions over who owned the risks, who was the project manager, was the data interactive for quality? Did the contractor exercise due diligence? Did they have an adequate methodology and were the record standard of care followed in order to do that? Our expert reports that we produce for them show that the conversion code that was used introduced errors into the data in two substantive ways. The data that converted was a measurably lower quality than the quality of the data before the conversion. We had a bullet on that one. That is why I had caused harm by not performing an analysis of the legacy systems were there and they had what held specific information on here. Now, now that we're with programming, there are objective programming standards that you should use that are published out there on the web by organizations like the IEEE and the Association for Computing Engineering. For example, if I'm reading in source and this is pseudocode so don't try and figure out which language this is. It says if column one is an M, then set value of the target to L, L set it to L, that's wrong because it doesn't make into account the fact that other values such as O, Z, and P might be coming up in which case they would all get set to C. That would be incorrect. And we're able to show that we're able to show that all of these values that they should have been testing for were not in fact being tested and this was governed in this case by Canadian law and the organization was very, very easy to show that they had introduced errors into the code. Similarly, this is some PeopleSoft code so it was a PeopleSoft implementation and when they couldn't get the data to go in correctly to the data set PeopleSoft provides you from adding duplicate numbers into the data set. Now that sounds a little bit normal. You wouldn't want duplicate records in the database and PeopleSoft correctly keeps you from doing that because in this case when they couldn't get the job to run the first time, dummy code that allowed these things to go in. So they were sending records to the data set. So when we looked and said how many records are in there, there were 63,000 records in a place where there should have only been 6,000 records and 100,000 records in a place where there should have only been 10,000 records. These are unconscionably stupid mistakes for somebody to make and yet they continued to make them and that happens in lots and lots of cases that we look at in the systematic it was marked as a high risk at the beginning of the project and the response, those of you that are familiar with Project Management Institute says you should develop a response to the risk and monitor for it and their only risk response was if it doesn't go in right we're going to charge you more money for it. So if it didn't work they actually got paid more money. How's that evidence that shows why was acting like the project manager and so there was no question as to who was in fact producing those results. Another piece again these are all pieces of evidence from these instances here. There were a bunch of instances of project plans and we looked here if you look to the next to the last line you'll see the largest one it had almost 500 instances of tasks with 15 predecessors out of 500 tasks. Again about project management you know that that is simply not good project management technology or not good data management about the tasks that go on with the project. Or here finally is somebody who is charging 2000 hours to the task of project management and only over allocate in themselves by a factor of five. So they were working instead of 40 hour weeks they were working 200 hour weeks. All of these are crazy but you'll see what happened in a lot of cases where if you look at your contract that are signed by the IT shop they are warranting that the service is provided will be performed in a professional and workman manner in accordance with the industry standards. If you ask the question where are the industry standards I don't know there's nothing written down I say if there's nothing written down then where are the standards so I wouldn't know where to go because for precise standards we're going to apply good data management practices to show in fact that the organization should be doing better. So very briefly legal situation here what I've done is I've walked you through time and agency leave tracking with an example it gave you one million dollars in data management savings and in chemical company for 25 million dollars and ERP implementation for there to come up with it and then some non-monetary examples that are really problematic as well as our legal example and we are at the top of the hour and time for questions so let's see what sort of questions we've done a good job of enticing you into this and again hopefully add you all to submit to the upcoming Enterprise Day World on this so that we can now start to come up with additional examples so that we can all share from the type of piece and there were a couple of contributed pieces in the book here as well as you as well so hopefully that will be good. Anyway we moved to our Q&A session Megan back over to you. Thank you Peter that was a great presentation now it's time for a Q&A time for you all to ask your questions so just click on the Q&A window at the top of your screen you should submit your questions through that Q&A window just a minute here so you guys can get your questions in and we'll get started this is going to be available at the DataVersity Bookstore right current book so they can go to the DataVersity Bookstore and get a copy of this while you were talking so we are good. Wow Shannon that's faster than Amazon we'll call you this from now on you will be faster than Amazon that's a little title I think it is we had one somebody said your Indian name but something you shouldn't do and there's one here called runs with scissors so Shannon is going to be faster than Amazon it looks like we got our first question then it is how do you retain organizational attention when other issues rise as being more costly and if you could make the cost case initially that is one of the things that we have to recognize is that while what we do is important we do have to keep in mind strategy as far as the organization goes let's give you an example of a company that some of you are probably familiar with it's a large credit card company that has the sort of slogan what's your wallet to give you a little context on that what's your tax and there are valid reasons for saying yes I understand that this is important and that you should do it this way however do you want to get paid on Friday which is a valuable question to ask from some organizations now don't worry about suggesting that this organization is going to go out of business but there was a pressing business need it's important but sometimes other things are more important however from a governance perspective what this organization did is they said you can proceed with the project absolutely go right ahead but we're going to get the financial results that you're going to get on this project I'll give you a very concrete number the business in this case needed to raise an additional $50 million annually a temporal business opportunity but that it would raise an additional $50 million and they said that's terrific we'll be very glad to do that we'll give you a waiver on the architectural component here for every that you don't go back and implement this inquiry two reasons for that one you'll get benefits but you won't get the benefits of the implementation and the other two projects will ensure that there's money available in this organization to rebuild the system correctly when the time becomes appropriate so a very good question from the question of perspective some of this stuff is important but it's not as important as other business concerns the question that you asked was that monthly tax was what kept it as a important criteria within the organization a good component though and I'm going to go back to the BT example here not that one is the only thing that they did I remember we did a little animation that they had there but we have seen organizations use this type of communication to talk to people about projects and one of the most important things to remember about projects of any sort is that you're not talking about it some people are just speculating about it they're making stuff up and if they're making stuff up it's probably not the stuff you want to have said about your project start off with a little bit of cute animation like this just to give that type of thing but giving a regular progress report in the same format there will actually cause people to look forward to these reports and if you tell them the truth projects going great but we had to stop it for a few minutes because something else came up then to what you say people do trust the things that you come up with from an office on this now IT hasn't always been as good about articulating these things as it would be and again I would encourage you to learn from our marketing group at Data Blueprint they do a phenomenal job of keeping everybody focused on track one of the things I do is I come up with lots of ideas and they go nope but they also are very very good at that communication so don't just internal to your own group to do it go out to your corporate communications group to your marketing group and ask them for some help tip encourage all groups to do is to keep a website up and running because that website is where people will go to try and get information particularly if you provide them with a good source of information but if I didn't please come back and give us another crack at it the next question is how do you provide this conference and not opportunities for example what is the best way to ROI for attending conferences for the networking benefits if you've already gone through many of the basic courses I'm sorry I'm laughing on this one but I do about 30 conferences and it knows me it's put me on the spot with that one one of the things I provide is back into business value and I'm not suggesting at all that you guys should go to 30 conferences a year it's heart and mind believe me but from a training perspective one of the things that you will see is that there's a marvelous opportunity to look at what good companies do I mean you look at good actually excellent companies go ahead and be excellent companies invest in training and it turns out that most IT people need a couple of weeks training or a couple of events such as what the questioner is asking just to stay current so you can go out and benchmark your company against other companies again you can look at training specific and if the average company producing average results is getting two training sessions a year then you can make an argument and say that going through should help now the other part it is to look at specific objectives and to say within our organizational strategy what pieces of strategy are going to be met by training objectives for example people that talk about this in here we want to learn how to be for example good at the things that are the self actualizing data practices that we talked about early on right and we want to learn more about master data management or mining just an example from here that what is good to learn about those text mining pieces the question is what business value will text mining deliver to us in your organization a legacy challenge and in order to meet the big legacy challenge we wanted to reduce the burden of the legacy conversion cost from $100,000 down to something less than $100,000 then in how to do text mining you can look at an investment in text mining might be on the order of $10,000 instead of millions of times I usually mention every session that we have fully trained and verified data engineering staff of any firm out there and I very much stand by that number of accomplishments that we have on that we brought in my colleague Dan Linstat in the summer who trained everybody in a data vault methodology and when Dan is training he comes up with specific training objectives and in a larger context a lot of the work that we're doing requires warehousing techniques so the data vault methodology was something that we needed to learn that we needed to learn to help our customers learn it and that Dan provides excellent training in that area so it gives specific measurable objectives and Dan controls the certification he grades every one of the essay exams that they take at the end of their training so it's not a matter of taking a multiple choice test and coming away to print it on the end of it he does his work and does an excellent job so it didn't mean to get too much of a Dan commercial there but Dan is a wonderful guy and his data vault training is very, very good and we were able to turn that into a very significant investment monetizing our own consulting practices in the area I hope those are good examples for you I hope that helps and if not again please ask other questions thanks the next question is how do you handle pushback improvements and soft savings don't always translate to the bottom line good question and I'll give you this example here when how do people spend their time six figures to do a job then you really want to make sure that they're doing jobs that are worth six figures and then digital information off of a computing system and then manually re keying it into another digital computing system by typing click click click look at these click click click back and forth is clearly not worth six figures so it's very easy to come up with these soft skill measurements in the area I'll draw your attention to a book that I saw in my book as major inspiration it's called how to measure anything it's a wonderful book that describes that virtually nothing is unmeasurable and that's the problem that's why we call them soft skills because they've gotten this sort of wrap that they are hard to measure again what is the value of training on HR professional can tell you precisely what the training is for organizations in your industry it can provide more dollars for share earnings for investment again whatever the measures are that you're using in your specific area and we call these things soft skills because in the past we haven't paid as much time and attention to them as we should but that we absolutely can drive value specific tangible value from these soft savings remember in this example here this was not just doing this and saving those people effort but that this translated into improved productivity on the parts of these individuals to the tune of their estimate $25 million gain in productivity a soft skills investment that resulted in a very tangible gain in this case they were able to measure it in specific terms of new products entered into the market improved testing reduced costs for the organizational teamwork because one of the things that happened as we were doing this was that we would actually had this happen in the drug oil industry as well researchers who literally worked next to each other in cubes who were doing research and they would launch and talk about things and perhaps even commute together but they never talked about their work and until they went to a conference and one fellow saw another colleague present a paper realized oh my I'm putting you've got peanut butter and we've been working next to each other these two things together again a wonderful wonderful set of examples in that area we call them skills not because they are soft but because we haven't been as good at practicing our measurements around them so again how to measure anything by the way Douglas Hubbard I don't know whether that's on your website but knowing Shannon is faster than Amazon she's probably got it up on her website by the time we've finished this particular webinar just a terrific terrific book and monetizing those particular skills and I do call it out in the book great thank you my question is can you say a little more about how to shift from an application centric development approach data centric or center development approach what are the ingredients our objectives to accomplish this shift great question and I always should add a caveat here than I did today which is to say that while this is an absolutely important thing to do it can't happen instantly and you do have to in fact take it greatly in order to do that so let's just review very very briefly here the application centric mix hopefully is on your screen there shows that most organizations start off a strategy to come up with goals and then go to what seems to be a very natural step of developing systems and applications and if you think about it systems are generally defined as hardware, software people, processes and data five ingredients in there so people tend to say yes let's go do that but the minute you start talking about again we said people soft several times in this talk so we'll say people soft at this step the conversation becomes about how people soft is going to do this and people soft is going to do that and that we talk about how people soft is going to help the organization instead of this system achieving its objectives and trick approaches which is lipping the systems and the data pieces of that and talking about the data objectives now you can go to your organization and say stop doing all these projects you're doing them all wrong Peter says you're doing it wrong by the way you're welcome to say this because I know when it gets you know shot for those things and that's perfectly valuable that way you won't get shot so you can tell them that Peter says they're wrong what we have to do is to look at it and the bigger your organization is the easier it is to do the best thing to do for this situation to implement this in several companies is that you say look you continue to do these projects the way you're doing them now but let's take a small portfolio of projects and in a different manner do them in this data centric fashion because if we take a significant portion of your portfolio and do them differently we can now compare A to B and whether this is in fact the better way to do this the goal is to let the organization see gradually over time that developing their data assets separate from external to and preceding the same development lifecycle activities produces better outcomes and a very good method for improving the productivity of software development however in the middle of an agile session and they say oh well we need another database to get lots and lots of different piles of data in the organization how does it function so what you should do is step back from that process and the only thing that will speed up agile in most organizations is taking the data out of agile when they get out of agile and say that an agile project should not start until they can specify precisely which data elements they're going to be reusing and creating and if they don't have those two pieces developed they shouldn't start on the agile project then we have a project to go forward once it's met this data governance threshold the data that they're going to use in there as a part of data governance to do the only way we can speed up agile processing is to take the data out to recognize that data is separate from this is absolutely different from anything we teach you in college and university in fact it's just the opposite of what we teach you we teach you that the life cycle approaches go off and develop is the life we've been getting and that those pieces and the sanity is doing the same thing over and over again and affecting different results I hope this gives you some ideas again it's not everything that you're doing in your portfolio can stop but take an experiment and let's try this for a couple of projects and see how it goes and once you're able to show the organization that projects done in this dyscentric method are producing better outcomes in projects that are going from the traditional way you can decide what projects meet the criteria from moving into this particular mode of operation and as we do this the organization will practice it get better at it and eventually you'll have a much better application development outcome by the way our estimates show it's 40% cheaper to do it this way and then the program or project are being implemented when your analysis shows that the systems are involved would critically impact CG data I guess because of pretty data and the business wants to move forward because it is a CGIC effort and budgeted for wow that was a mouthful I've got the sentiment on it absolutely this year we celebrated in the springtime the case for the chief data officer and is to understand that because these data elements in fact coming out the data evolves and it's not created data asset for organizations an asset it should have a manager at asset as strong as your GFO is an asset manager again imagining your organization if anybody could write a check for any amount of money for anything at any time that would be considered very poor fiscal asset governance so that's an idea for most organizations and in fact it's illegal in some parts of the organizations our data as an asset is ungoverned so the first we have to do is put somebody in charge of it because if everybody is in charge which means governance will then create any development activities there's a chart in the other book that shows that governance then controls as a gateway function the technology development that occurs so if you have your data requirements down the technology development can proceed but if you don't have that and it requires strong leadership again I could name the corporations that I have stood on their corporate boards and said if you step it would reduce your IT spend by 40% and a lot of them like saving 40% on their IT spend but not they have been willing a few of them have but most of them have not been willing to say okay we'll subordinate our programming projects to our data governance function so that you subordinate all of your IT projects to your finance function so why shouldn't we support these other functions to that I gave a very clear answer there but did it make sense if not please do ask a clarifying question on that for one more question what if data-centric development approach still influence leadership to step back and stop the execution of the program or project does the approach in other words we still manage expectations in this process as well now one in three IT projects actually succeeds on time with full functionality with engine with access to the organization so better than those numbers is going to be considered a success if we're doing this and we're not doing any better we're not doing something right and the idea is that we should be careful about our measurements and our expectations so just the same as we can't say what stop the cart we've got to change everything and we can't do any more development until we have our enterprise data model that's crazy the organizations aren't going to wait for that so we manage expectations and say we're going to introduce this project gradually in a series of controlled exercises so that we can isolate the incidents are you going to provide data-centric development on your most important strategic asset probably not something that is of less importance but still doesn't have a good track record and do a little test to try it out see how good you get see if you're in fact getting better about it and then move on to what you're trying to really do which is to be in place for all of your projects again carefully manage expectations don't go out and say Peter said stop and start it doesn't work but gradually transform and look for a way of comparing because when you do this the results speak for themselves and management won't need to be convinced of this because it will be obvious to them that this is a better way of approaching the management of our sole non-depletable non-grating so we actually have one more question and I think we have time to answer it quickly should data have an asset take and in fact any of you that are going to an argument in the book about whether data is an asset or not so there's actually a couple of I've got a couple upcoming events where I'm going to be arguing with some thought leaders about whether data is an asset or not and I'd welcome your participation in those events so if you want to know about those just look on the web you'll find some things out there that we're going to have some debates about whether data is an asset or not I think it's you know if somebody's email address that's only valid for about 90 days really a data asset for the organization depends on what you define as a data asset if information about your buying habits of your customers that's going to be an asset so again I think the answer is yes the data is an asset and it should be managed in a manner comparable with your other organization assets including your assets including your capital equipment et cetera et cetera those are all the questions we have for today thank you everyone for participating in today's event we hope you have enjoyed it thanks to the University and Shannon for hosting us once again you will receive today's materials within the next two business days next month will be business value through reference and MDM hope you can join us for that as well as always feel free to contact us if you have any questions thanks everyone and have a great day we've got some ideas about how to put topics together for this for Enterprise Data World absolutely we'd love to see it in Austin in April so we welcome your submissions we're actively soliciting them thank you and Peter just to emphasize and get us those questions in the cutoff day is the 14th of this month so not much longer left or almost through October I can't believe it everyone again as always thank you so much for your active participation I love all the questions coming in and everything and I hope everyone has a great day and thanks Peter and Megan for another great presentation get that other book up on the website it is it's up there yeah we've got a new nickname for Megan we need a new t-shirt for Shannon it is a special on the floor bye everybody have a great day you too