 Welcome, my name is Shannon Kemp and I'm the Chief Digital Manager for Data Diversity. We'd like to thank you for joining today's Data Diversity webinar, Data Strategy. Plans are useless, but planning is invaluable. It is the latest installment in the monthly series called Data Ed Online with Dr. Peter Aiken, brought to you in partnership with Data Blueprint. Just a couple of points to get us started. Due to the large number of people that attend these sessions, you will be muted during the webinar. For questions, we will be collecting them via the Q&A in the bottom right hand corner of your screen, or if you'd like to tweet, we encourage you to share our highlights or questions via Twitter using hashtag dataed. And if you'd like to chat with us or with each other, we certainly encourage you to do so. Just click the chat icon in the bottom middle for that feature. And to continue the conversation and networking after the webinar, just go to community.datarcity.net. To answer the most commonly asked questions, as always, we will send a follow-up email to all registrants within two business days containing links to the slides. And yes, we are recording and likewise, we'll send a link to the recording of this session, as well as any additional information requested throughout the webinar. Now let me introduce to you our speaker for today, Dr. Peter Akin. Peter is an internationally recognized data management thought leader. Many of you already know him or have seen him at conferences worldwide. He has more than 30 years of experience, has received many awards for his outstanding contributions to the profession. Peter is also the founding director of Data Blueprint. He has written dozens of articles and 11 books that the most recent is Your Data Strategy. Peter has experienced with more than 500 data management practices in 20 countries and consistently named as a top data management expert. Some of the most important and largest organizations in the world have thought out his and Data Blueprint's expertise. Peter has spent multi-year immersions with groups as diverse as the U.S. Department of Defense, Deutsche Bank, Nokia, Wells Fargo, the Commonwealth of Virginia, and Walmart. And with that, let me turn everything over to Peter to get today's webinar started. Hello and welcome. Welcome and thank you, Shannon, for being patient with us. We had some little technical difficulties there. I did it correct, but we are actually all ready to go with this now. So let's jump right in. The real key for this, of course, is that most people focus on the end product in their data strategy. And there's a quote I'll use a little bit later on from our previous President and General Dwight David Eisenhower that says, plans are useless, but planning itself is invaluable. And the reason for that is because most people, when they dive into data, don't really understand what it is they're going to be experiencing here. So we have a fair amount of stuff that's got to work in order to make all this work. And unfortunately, most companies are failing in their efforts to become data-driven. Peter? Yes. Your computer just booted you up. We just lost you. Entirely? Yeah. Log back in. We've got the app still, so just need the computer login. I'm afraid if it leaves, I'm still in the event. If I leave it, it's probably going to hang up, Shannon, so I'll have to do the whole thing. So your phone is separate because you called in. So you're not logged in. You're about to be booted out. Yep. Coming back. Sorry, folks. Tech is great when it works. Mm-hmm. We're having some money problems with here. I can at least talk while we're doing that, so let's see if we can get... We can hear you. Exactly. The real key is, these first two slides are just talking specifically about the fact that companies are actually themselves admitting that they're having trouble doing data. And a couple of the findings that I'll carry in the... The notes there where you get them is that the vast majority of companies are realizing that data is a lot more complicated than they originally thought it was going to be and that they're having all kinds of really significant problems. Shannon, I think the Internet is dead here at VCU. Okay, some of your slides. Oh, well... 400 megabyte presentation. That's not going to happen. So what else can we do? Everybody's on here. Have your hotspot. Ooh, good idea. Let's try. See how panicking and she's keeping a little head. That's my job. Uh-huh. Todd wants to know if we know how to tap dance. Yes, but I don't know that you'll be able to see it. Hear it. Fun time. Okay, I think we've got it now. Can you hear me? Yes. Oh, and then you're visible. Amazing. Hey, we may be able to do a fundraiser to see us tap dance. Data governance jeopardy. I love it. What's the data dictionary? You're muted. And let me give you controls. Okay, there we go. And the presenter. You are a presenter. All right, you can see. Yes. Take it away. I pointed these slides. I'll just very quickly go through them and go through them in more detail. But people are not paying attention and doing the right kinds of things in order to be successful with data. And of course, part of that is a strategy on this as well. We're getting some fairly good objective data, though, in terms of people that are doing big data and all sorts of things. So this is Randy Bean's annual insight that he does every year. And he's just released his other one, this year's one, so we've got to go back and look at those. But you'll notice, again, some of these numbers in here where how many organizations are taking or thinking about taking steps in that direction in order to get going. And this is one that I find tremendously useful. What types of differentiation do you have in terms of job title, data job titles within your organizations? That's a really good leading indicator on that. Again, 80%, managing data challenges, no problem with that. 70% are managing data within departments instead of trying to do it across the entire enterprise. And finally, we've still got an awful lot of companies, 60% that are reporting to IT on there. Now, I'm going to put something else up here for just a quick second. Let's have a little bit of audio on it. So see if we can get it to work. If I was doing this live, I would ask you guys to raise your hands when you recognize this song. Now, what you see in there is, of course, Bruce Springsteen doing a show in Melbourne, Australia. And what I want you to think about in terms of the data strategy is that Bruce Springsteen doesn't appear on stage with a set list and do the same set every night. In fact, it's one of the things he's known for not doing his improvising. He does a tremendous job of this. And he told the band two hours before they went on stage, hey, I think it'd be really nice to honor the Australians since we're going to be playing in Melbourne, Australia with a song that is done by Australians, in this case, the Bee Gees. And I'll tell you, in 1977, from high school, if there was a song I hated more than anything in the world, it was Stayin' Alive by the Bee Gees. But if you go to YouTube, you can see this video in its whole. It's a wonderful, wonderful thing because he does a great job with the music. And this is the key. You've got to have some good music to do this. And in data, you've got to have a good plan. But there's another part of it that most people miss. And when I look at how people are doing data strategies worldwide at this point, I see an awful lot of emphasis on trying to write the world's best pop song the very first time. And it's just not possible because only through practice, as well as good material, can you actually come up with something that does equate to really good music. And so that's what we're going to be talking about today, is specifically looking at strategy and saying, hey, it's an inherently repetitive process that can be easily improved with practice. It is dependent because data strategy exists only to support the organizational strategy. That in the early phases, we're going to be able to see what's going to happen. That in the early phases of maturity, process is much more important than the product on this. The output, the plans are of limited value and you always discount the obstacles that you're going to run into on this. Randy Bean's findings show this year after year, people in process problems are 95% of the problems in data and not the technology piece. And finally, how do we get to Carnegie Hall? Practice, practice, practice. So we're going to dive through this pretty quickly just to make sure I cover all the material, but we're going to talk about a data strategy and how a data strategy specifies the data assets that are to be used to support the organizational strategy. We're going to look at what is strategy, what is the data strategy, and how do the two of them work together. Then we're going to take a look at why data strategy and data governance are interdependent. You need to have this because both data strategy and data governance are focused on improving your organization's data, how to use their data, and once you've got better data and better people skills, then you can use and improve the way your people use data to support your organizational strategy. We'll talk about some specific prerequisites. Most organizations have a definitive set of things that they need to get through in order to do this and compensate for the lack of data deficiencies. And we talked last webinar, which was back in December about the seven deadly sins. Then we'll get into the iterations and really once you learn the process of doing this, it becomes a matter of as simple as it is on the shampoo bottles. Although be careful, if you follow the directions on the shampoo bottles, rinse and repeat, you would never actually leave the shower. Let's get started on this and dive in and talk about what is the word strategy. Now strategy is currently derived from the military of these things. About 1950, though, you'll notice on the Google graph there over the use of word strategy over time that businesses got a hold of it. Because there's a lot more business people than there are military people. They went, wow, strategy is great. We can do this great thing and do all sorts of strategy work. A better definition of strategy is a pattern in a stream of decisions. Anything else beyond that is really complicated. And to prove that to you, I'm going to give you three examples of strategy. Two, very quickly, and then a third one that is more complex. One is Walmart's previous business strategy. Many of you actually think you know it already. Every day low price. Well, okay, that's good. That means when somebody is making a decision at Walmart, they err on the side of making the customer have the lowest price. They will never get beaten up for anything along those lines. It's got to be a strategy simple enough that all of the million-plus people that work at Walmart know what to do. And believe me, they have done a great job of doing this. Everybody understands it. Another definition of, excuse me, another instance of strategy is Wayne Gretzky as a wonderful sports enthusiast. He puts down his strategy for winning hockey games. He skates to where he thinks the puck is going to be. After all, the puck is a piece of plastic. It's hard plastic. It runs around on ice. It moves much faster than people. Good luck. You will never beat any of Wayne Gretzky's records. And finally, our third one is a sort of a little bit more complicated one. It's an example right out of history. Napoleon in the blue is facing two armies that are bigger than he. The red is the British and the black is the Prussians. Now, how does one beat the enemy when their forces are bigger? The answer is divide and conquer. So let's take a look at what divide and conquer actually means in the context of this strategy. First of all, Napoleon noticed that there were two supply lines supplying these two armies. The British were being supplied out of the English Channel in Ostend in Belgium. And the Prussians were being supplied out of Liege and they are two separate places. Now, Napoleon know very well that if you hit the army at exactly the right position by dividing them, they may back up. If you hit them really hard, they may fall backwards. And if they fall back, they are more likely to fall back towards their food than away from their food. So the first step is divide and that was this attempt here. And the second part was conquer. So conquer means, okay, I've got this large army and I need everybody in the army to turn to the right and defeat first the Prussians and then everybody turn to the left and everybody beat the British. If I get half of the people doing it one way and half doing it the other way, it is not going to work. And of course it didn't in this case. This was Waterloo and we know that that did not work out well for him. But the complex strategy is taught still in the military environments here as an example of good strategy. So let's review this complex strategy for just a minute. Hit both armies at just the right spot, divide, and then everybody in my army turn to the left. Excuse me turn to the right and defeat the British and then defeat the Prussians. But if I don't do that then we are going to lose. And of course I want you to do all of this while somebody is shooting at you which doesn't make it easy. It wasn't easy. He didn't actually win this battle and that is where strategy comes in. So I mentioned at the top our previous president and general like David Eisenhower has a wonderful quote that I use all the time. In preparing for battle I have always found that plans are useless but that the planning is indispensable and in my professional experience I have looked at hundreds and hundreds of organizations data strategies and most of them end up on a shelf and are simply un-useful. And the reason they aren't useful is one they are too complicated but two Mike Tyson has a wonderful quote everybody has a plan until they get punched in so let's go back to our strategy a pattern in a stream of decisions and now let's look at it in the context of a data strategy. Our data strategy in this case is the highest guidance available to organizations focusing on data related activities focusing data related activities on business goal achievement and providing guidance for folks when faced with a stream of decisions or uncertainties. Now let's talk specifically about data governance for a minute and I like the definition data governance is managing data with guidance. Any definition is more complicated and that is likely to be misunderstood by people who are outside of your domain although I do like to add one additional word into that which is managing data decisions with guidance and this is the reason we have to be in a position to impact the results of those decisions because there have been too many poor data decisions over the years. So guidance says well how is the data used what business processes consume it where does it share can I move Brazilian data into Ecuador whatever but most importantly about everything else in what order should I approach the list that I'm trying to figure out because it is a non trivial problem in order to do that. So if we look at these two in context data strategy says what are the data assets doing to supply to support data strategy and governance then it's about how well is that data strategy working. Of course this has to be done in the context of making absolutely certain that our data strategy only supports the organizational strategy there is no other purpose for it and we've got some other related things that IP and other pieces are working on. I generally don't like to show this picture a whole lot because it just makes it really really complicated I think it does present a good few of the world but I wouldn't do that in the public. Keep it at this level here which is to say that let's just talk about how well is the data strategy working and oh by the way the data strategy should be expressed in terms of business goals and the language of data governance if it's not metadata becomes even more problematic because it's hard to keep things focused especially for people who really do not understand that aspect of what's going on. The most people will think we have an organizational strategy but you have a IT strategy and then you have a data strategy and I say this is so incorrectly wrong it is unbelievable so pull the buzzer on it do not let people do this instead the correct way to do it is to look at your IT strategy as complimentary to your data strategy and vice versa. In fact if I had to I would say your data strategy should drive more of your IT needs then your IT needs drive your data strategy but I have seen strategies over the years literally where they say well we are going to rely on big data now we know that big data is succeeding at the same rate of most IT projects which is to say about 30% of them achieve a positive return on investment data science of course is so ill-defined we do not have any good measurements on that and it gets worse people will say well we are just going to rely on analytics or SAP or Microsoft or Google or Amazon all these fine companies are going to take care of all of that well again a data strategy has to be very simple and describe how it is going to work in supporting the organization the strategy also of course has to do with better understanding that data is an asset that is typically less well recognized than most organizational assets that are in there again cash no problem people understand this just go with that we also have to understand too that a good important part of our data is that not all of our data is equal now data that is better organized does increase in value if I can put my hands on it quickly that's knowledge worker productivity and time to market and other types of time related decisions on that and therefore poor data management practices are costing your organization's time money and effort which is unfortunate and the really sad part about it is that 80% of your data in your organization is wrought you may say to yourself goodness that's a pretty strong statement well the only argument I get over it is that it's not 80% at 85 I had one organization tell me 90% wrought of course stands for data that is redundant obsolete or trivial and if it is redundant obsolete or trivial why are you spending any time managing it at all you should be paring it down calling it if you will in order to get it out of there my wife added the attribute incomplete as well and so we could actually call it riot but I think wrought out a little better than that when you look at data assets compared to other types of assets data is the only asset that you have it isn't depletable it isn't degradable it is durable in nature at the strategic level so data assets really do compare very favorably to other types of assets and yet most people still go back and say well data is the new oil and again I say this is the wrong way to think about data we never think about what happens to the gasoline when we take it out of the gas pump and put it in our car and consume it to change that slightly when somebody says oil and say instead let's call it the new soil there's two reasons this is a better metaphor first of all the oil is a production function where data is most valuable when it's reused rather than when it's used and so the key here is that data is not the new oil but the new soil and when you plant things you don't just talk about the yard or the forest or whatever and scatter seeds everywhere and hope good things happens that's called evolution we're doing something a little bit more purposeful than that secondly we never plant something on Monday and expect to consume it on Friday if they take much more time in order to do this now got a lot of groups that say it's the new bacon and that's fine too if you need to sell it that way but really data does deserve its own strategy as a strategic asset it deserves attention that is on par with similar organizational assets and it requires professional administration to make up for past neglect so if we look at what's actually happening in the data world here what we will start to see is that organizations are able to jump in and do a lot with data when they put their time and attention into it let's get to a motivational slide on this sorry I'm going to skip a slide on this because I had one that didn't work there we go and the key to this of course is that data has basically three motivations around its strategy everybody would like to go straight to the end improving the way your data and your people support your organizational strategy but you generally can't do that because your data is not in good shape data points to where valuable things are it has intrinsic value in and of itself and it has some wonderful combinatorial value that we use but nobody's data in the shape they'd like it to be in and with 80% of the data being wrought it's not even clear that you have a good return on investment of cleaning it up until you have first culled the data and gotten rid of the parts that are not able to do that similarly however we also then need to use data to measure, manage and motivate change and your people really don't have much education around how to use data finally once you have improved your data and improve the way people know how to use your data you can now start to use it in a way that supports your organizational strategy and I love to tell this story of Rolls-Royce because it's how to create a competitive advantage using your data now Rolls-Royce had an old way of selling things which was simply to say that they had jet engines that were extremely good when I get on an airplane I do not worry that the engines are going to flame out in the journey Rolls-Royce jet engines are among the most reliable engine parts ever created and by the way it's not just is it a Rolls-Royce the other jet engine manufacturers do exactly the same thing the key though was that Rolls-Royce wasn't able to have some conversations with their customers because they were selling high quality products that worked a lot and everybody considered to be a good value the vendors wouldn't treat them to a conversation so Rolls-Royce said aside from the fact that the Southwest plane crashed didn't crash actually landed in Philadelphia but did kill one person one U.S. aviation death in the last 10 years wonderful safety record the new model that they put in place was not to sell jet engines but instead to sell hours of powered thrust which meant that their position in the bargaining was very different as a vendor they were across the table and it was a us versus them situation so with the new model the new business model selling hours of powered thrust they called it by the way power by the hour pretty cool stuff they could now go and have some other conversations where things they had learned from NASCAR or Formula One or other types of organizations and again here this is a little bit of a bit it's an old show of the Indianapolis 500 the way they used to change tires a little more himself just making a comment about changing the tires here I love the way they change the tires by hitting them with a hammer so I'm going to shorten that a little bit because it goes a little bit long it changes the tire and we'll skip and he'll change the other tire there he goes on the other side putting the tire back on get a little bit more so if you couldn't hear the audio on that 67 seconds to change two tires and our new measure 4 seconds for tires now that type of business result can happen but that entire conversation couldn't take place until Rolls Royce changed their business model and the new wing to wing process helps all of us flying consumers because when they can change an engine on an airplane faster it means we will make our connections faster now the key thing at the end of this little vignette on Rolls Royce is when was this new business model invented the answer is 1962 oh my goodness in 1962 Rolls Royce had decided that this was an important part of their business and that they needed to move forward on this and the only way they were going to be able to do a good job of all of this was by having different conversations with their customers if they didn't have those conversations with them they could never get to this model and never be able to use this that's the thing that governance and strategy working hand in hand in can help your organizations also achieve these sustained competitive advantages which is what all organizations want to do unfortunately the sustained competitive advantage is only good for about three years so it does give you a little bit of a timeline here and it questions whether that sustained competitive advantage is good but I'll take three years as opposed to no years in order to do this our data strategy is critical for data governance because if you don't use your data strategy to improve your data as well as improve the way your people use their data you will not be able to help your people use your data to support your strategy better move on now a little bit further where we get into some prerequisites for this because most of the time people like to try to do this but unfortunately they are not quite ready so let's talk about the lack of organizational readiness that they have the failure to compensate for the lack of data competencies in here and again the seven data strategies that we seven data deadly sins that we talked about before data strategy has some prerequisites to do it well you need to get good at it it's not quite the ten thousand hours that Malcolm Gladwell wants you to have to become an expert but it does require some prerequisites before you can get to the point of ladder, rinse and repeat and as prerequisites fall into a couple of categories the first one is an organizational change category if we don't get the organization set up to do some change and understand that things are going to be different there is no way we will be able to just do it from a technology play remember 95% of our data problems are people and process problems not technology problems there's another component that goes into it too which is that you've got to do the right type of talent in order to pick it up and eliminate the things that are going to cause you problems within there let's take a look at the change management aspects of it real quick first of all people like me run around the world saying that while CIOs are good and do a great job on the most part CIOs are not really doing a great job with data if they were then we would not be on this call so we've got this wonderful scenario where people like me are going around the world and saying CIOs aren't that was actually the title of the book that I wanted to put out but they said if you want to alienate an entire class of individuals go right ahead for it so instead I put out a book called the case for the top data job on this and they even made me change that title to the case for the TDO because they figured nobody was going to understand what a top data job was now interestingly this book got translated into Chinese it was translated as chief data officer combat so I thought there they were at least able to get a good idea of what was going on in this because when you do introduce something like this into an organization you get to fear doubt and uncertainty that comes to mind Mario Faria who's a great analyst for Gartner did a really good article on that and I'm giving you the link there that describes that so luckily we are faced with a situation where things need to change specifically in our organizations and there's a class of professionals who do an absolutely awesome job with all of this they are called change management and leadership consultants you probably have them in your organization I'm talking to you right now at Virginia Commonwealth University and we've got them here at the university as well and what we're really trying to do is to make it harder to do it the old way than it is to do it the new way and that the organizational change doesn't happen easily in fact it's kind of like a lock a key lock the physical lock not the electronic ones that we do so if I look at a company and I see anxiety but I see that they've got a vision strategic vision they've got some incentive and some resources and an action plan I know that what they're missing is some skills and if I see frustration I know they've got the vision that they want to go strategically they've got incentive to make people move it they've got the skills they've got an action plan but there's no resources and this little diagram here Mary Lippert has pulled a wonderful set of things that show that only when you have the vision the skills the incentive resources and the action plan do you actually end up with change in the organization and these are components that are going to go into the process of learning to use data strategically because culture is the biggest barrier biggest impediment to our shifting our way of thinking about organizational data I don't have time on this webinar to do this but I did create a case study on this and you're welcome to go out and download it so when Shannon sends you the slides just click on that link it's live and it'll get you right to that particular case study again dramatic change in organizations if we don't get that dramatic change then we have an issue similarly we need to find talented people to do this sometimes you will be able to find them internally and sometimes you will not let's take a look at how this works most of the time when organizations attempt to do this they have some business needs and they take those business needs and translate them into some sort of a solution or roadmap again makes perfect sense but it's the wrong way to do it and the reason is we're leaving out an important part of the equation what is the current level of preparedness in the existing organization what step are they to do the kind of work that needs to be done and only when there is a match between the business needs and the current state of the organization are we likely to find something that is going to be successful around that why is that the case well the answer is pretty straightforward we teach our knowledge workers nothing about data and yet my definition of a knowledge worker is somebody who works with data now that's bad enough for our workforce but our IT professionals it's much much worse we tend as a rule for the last 30 years we have taught our IT workers exactly one thing about data so all the classes we teach them and the one thing that talks about data we give them a course in how to build a new database if there is a skill we do not need anymore of on the planet earth it is how to build a new database there's another problem with this though in addition to lacking information with people who are building these things and not knowing it our leadership has also gone through these programs and when you look at a program that says there's 10 things you need to learn about in IT and the only thing we're going to teach you about how to do data is how to build a new database it's no wonder that that is the solution to every proposed problem yeah well what they taught me in school was that when I have a data problem I build a new database now don't get me wrong I'm not complaining this is what is keeping us all employed and will keep us all employed for a great number of years but it does speak to a rather significant disservice that we've done here we can actually go back and look and we've got these scientific papers if you want to read about them which shows that IT and particular data within IT used to report directly into a CIO and has now been pushed down almost three levels in a matter of short 20 years we've not had changes like that occur in most major organizations on that degree of rapidity and what this means is that our leaders have gone through this and said I need data people only when I need to build a new database so if I'm migrating databases I'm not creating a new one and I don't need organizational data management if I'm implementing a new software package I'm not creating a new database oh there's a script that rolls out of database but I'm not designing a database so therefore I also don't need data management or data skills if I'm installing an ERP the list goes on and on and what it really comes down to is that most organizations have little idea what data that they have they don't know where it is and they don't know what their knowledge workers are doing with it and that unfortunately leads to a series of bad data decisions it's no surprise that a large number of companies report these but when you have business decision makers that are not daily knowledgeable and you have technology decision makers that are also not data knowledgeable then we have a problem where we end up with a lot of bad data decisions and those bad data decisions result in poor treatment of organizational data assets and of course from there poor organizational outcomes and we'd like to get off this treadmill again this is the leather rinse and repeat stuff that we have is a big problem but I want you to imagine your organization going out and sitting down to interview somebody for a chief data officer position how can the people at leadership know what they're trying to hire when we don't even have a body of knowledge around that ourselves and we're seeing an awful lot of data leaders turn out on the very first one we tend to say that most organizations most data leaders and organizations trying to do data strategically will not succeed the first time that they will need several attempts in order to do it and think about this from another perspective too if I'm going to go in and change a bunch of things in the organization I just talked about the need for organizational change management in these areas and the knowledge and skills the same ones we need to have as a data scientist to fully exploit the data absolutely not so you may want to look at phasing your data strategy in over a spot that says first let's hire somebody that can get something changed and second let's hire somebody for running things on a more smooth operational type basis so this then is the how do you recruit the knowledgeable staff and talent in there talk to other data professionals and find out what they know you're not going to find your people who are above you know a whole lot about the characteristics of what it needs to be to be a good data leader and to do data from a strategic perspective so now we're going to briefly go through the seven deadly sins that are problematic in this area too first one we already talked about failing to address the culture issues there are some sequencing issues in data and expectations they spend an awful lot more on IT because of bad data decisions and if you can help that organization by saving time and money right up front and then using that money to invest in things later on it's a whole lot easier than asking for thousands or millions of dollars in order to do it later on there's a number of sequencing components that are important you have to manage expectations again nobody looks at you planting tomato seeds and eat those tomatoes on Friday but we've got tremendous unrealistic expectations that are occurring in this field the data program has to be aligned with IT projects and I'm using the words deliberately there so IT projects are the way IT is rolling and should roll out no problem at all with that but data cannot exist successfully in organizations if it only is constituted on a project by project on a basis a program is a much more important way of dealing with this because it ties things in a lot better we also need a robust programmatic means of sharing the data which means that you shouldn't start any sort of development activities if you don't have the data components ready to go now one of the questions that comes up on these is that I'm trying to do my data strategy but those agile stuff keeps getting in the way because everybody wants to move faster in the best way of developing higher quality software faster but if you want agile to do work in this and you are working on an agile sprint and you discover you have a data problem you need to pull the ripcord on that and stop working on that sprint because the only possible outcome in an agile format is more small piles of data and that's probably not the outcome that everybody is attempting to get second one leadership we addressed that earlier but we do need to have it in place and understand what data thinking is data centric thinking now dataversy has done a number of events that we've done over the years where we've talked about what it means to do this we've got some ideas around this in terms of the data doctrine and there's a number of things to get there there's a little website out there if you want to go jump into that discussion or come to some of the events that we've got upcoming on all of these things so these prerequisites are really going to hurt and trying to go through the three major classes of things I've just described organizational readiness lack of data competencies on the part of individuals and those seven deadly data cents that we have and then picking up a 100 page data strategy document that is typically of what I see in organizations and trying to make it work it is a guaranteed failure and I can name outright on my 10 fingers a number of these things that I've seen that have just been absolutely disastrous in terms of getting people ready to do this so let's take it to the next step here which is the next phase if you will once we've eliminated the prerequisites how do we actually start to get this to work and the answer is we need to go into something where we talk about lathering, rinsing and repeating the process which is much more important than the actual outcome some of you will have read a book called the goal I know I did and was actually required to read it by a very dear friend of mine who said I'm not going to talk to you about this stuff unless you go through and read this book the goal talks about these five steps we're going to come back to those five steps but it evolves around something called a theory of constraints and we're applying that same theory here so the goal again wonderful book you can pick up a copy of it for five bucks on Amazon I just looked because my students are going to read it this semester the theory of constraints says that in any system it's being limited by a small number of constraints kind of like a Pareto analysis 80% of your problems are caused by 20% of whatever it is you're working on in those areas there's always at least one constraint and the theory of constraints uses a focused process to identify the constraint and restructure it for the rest of the organization to address it the chain is no stronger than the weakest link and the data links in most organizations are absolutely weaker when you compare them to some of the other aspects of IT that's going on in that area Golderath died in 2011 but his book has provided an awful lot of things out there most importantly the whole DevOps movement is focused around this theory of constraints so when somebody says well DevOps will solve all of our problems DevOps is a definite step in the right direction but DevOps still does not really address data it's more about code in terms of how that works out so let's see how these five steps actually work five step process at the generic level outside of data is to identify the constraint it may be a machine on a machine floor it may be an executive in your organization that doesn't understand data it may be your systems but something is keeping your data from being more helpful to the organization achieve its competitive advantage and the goals of the strategy identify that constraint and then exploit it try to get some quick improvements around it sometimes it's just the visibility that will help out more often there's a little bit more work to do it that way though which means you need to subordinate everything else that isn't the constraint we probably have too many IT projects too many things on our data plate that we'd like to do that we need to take a step back from and say let's eliminate everything except the essential that we're doing here subordinate all of other things to other constraints if the constraint still persists then we need to restructure a repeat until that constraint is eliminated so that is the theory of constraints and if we move it into the data world using your data what that does is what is the area that our organization is most being blocked from from achieving success strategically and how data can support it exploit that constraint try to correct it operationally fix it without restructuring if that doesn't work improve your data evolution activities to ensure a singular focus on that objective don't let it become another thing when you discover you have a flat tire in your car you stop driving there's many of these IT projects that have bad data flat tires and that they still keep driving in here you may need to restructure to address that constraint and again keep repeating until we actually address the strategy all the way around now each of these components in here allows the organization to get better at what they do just in exactly the same way that Bruce Springsteen's band has been playing with him for 40 years so they can look at him and tell what song he's going to play next even though he won't write down a set list for them I'm not saying your organizations need to be as empathetic as Bruce Springsteen's organization but it won't hurt to get better at teamwork all the way around in these areas because once you've got that part of things set up the way you want you can go back to your business needs again remember we don't like that initial solution because most organizations forget to consider the state of the existing organization in there in order to pull these two pieces together only when we have that match do we then start to execute and now the new part of this is we make a roadmap and our roadmap for our data strategy is a very important aspect we have to do a balance between delivering some business value showing some progress in attaining our strategic objectives from a data perspective and we also have to get everybody better at what they're doing practicing now if you do too much on the left hand side of this diagram where you're delivering business value and delivering business value we tend to see that organizations become dependent on a method or a person much more so than is healthy and if that person gets promoted or the method proves inadequate to some form then we have problems there on the other hand if I weigh the outputs of my strategy too much on the right hand side there with new capabilities management says well okay when are you going to get them to me and we say five years remember the average CIO is only in their job from two to four years these days so if you're talking about a five year payoff the CIO is not going to think you can be part of their success in addition to that just to give you a comparable measure organizations with CIOs and CFOs those organizations that have them in the same organization organizations that have them both the CIOs tend to stay two to four years but the CFOs tend to stay twelve years and that is a tremendous difference three to four times longer than a CIO in place so you've got to understand that you need to tie your successes in the data area around something that will be lasting and notable that will make a difference to the organization failure to do that means you will end up with a situation where you have a nice data strategy document that sits there on the shelf but you're never giving the resources and even when you try to do it the business isn't ready to absorb the changes that you're making in the data resource area so this data strategy framework is an absolutely key document to make sure that you have out there but also make sure that people understand you're creating a data strategy as a small series of data steps that you can take small little bite-sized chunks away from it one step at a time by focusing on things now again I'll go back to the idea around projects IT projects most of the time when I work with companies one of the things I like to understand is how many people they have in IT so a company I'm working with right at the moment has about 300 people in IT and they've got 500 projects going that's a crazy number I'm sorry but that's nuts that means every person in your IT organization gets to work on some part of a project or some part of the day that's really putting time and attention into it and the organization is getting bled to death by a thousand cuts or in this case 500 cuts and what needs to happen there is that somebody needs to go back and say which of these projects are really essential to the organization and more importantly which of those projects are dependent on data in order to get us from where we are to where we want to be practicing that process is exactly what I want all of you all to do as you're out there trying to help your organizations do more with your data so let's give a quick recap on some of this strategy is an inherently repetitive process that can be improved most people don't think of strategy that way but if you have somebody in your organization that has worked with the military who invented the term strategy they will tell you exactly that is why the military does what it does in the way it does because the military does not know what they're going to encounter when they are in a battle and so by drilling by repeating exercises by doing things over and over again data strategy is a much better way to approach it by saying let's fix a little thing and let's try it and we get good at this process of fixing things we will then be better able to take on some of the more challenging projects that we need to face in data such as for example implementing a master data management solution just to pick one off the top of our heads where we've got lots of possibilities so if we're going to go out and implement an MDM let's not spend a lot of money with huge consultants let's instead figure out how to use master data strategically in our organization and you can do that not by spending 60 million dollars and hiring 60 people is literally three plain loads of people that I thought this one organization that I was working with who would come in and work and put this beautiful solution the technology worked really well but then I'm walking the halls of the organization after they spent 60 million dollars on this thing and I hear well I couldn't figure out where to stick the data so I stuck it in the MDM well that is not using data strategically and it's certainly not practicing and more importantly recognizing that the process of using data strategically in your organization is an inherently repetitive process that can be easily improved I see too many organizations where they have a data strategy that's trying to drive the business strategy it may be that your data strategy is in fact a better way of going than the organization is going but if you don't make your data strategy so that it clearly supports the organizational strategy nobody is going to pay attention to you we only exist to support the organizational strategy and what we do is we support it with the most dynamic resources the non durable non depletable resources that we have in our organization use resources as types of resources that have much different value and are much different characteristics are much more important to the organization strategy all the way around that the process of doing this particularly at the early stages of maturity is much more important than the output again we send our students out of here and say you're not going to get everything right when you go out same thing is true for data your first part of your data by the way a very important piece of data strategy always number them if you hand somebody a data strategy with version 1 on it they're more likely to expect version 2 to come along presently whereas if you hand them your data strategy and then another year later somebody else says well I didn't like that type of strategy here's the next one wouldn't it be a better story to say we achieved all the goals that we were trying to on data strategy number one and that's why we have data strategy number two because we're now going to start working on different things in another way to help the organization at the early phases of this process is much more important than product the outputs that you get are going to be of limited value and you're always going to have obstacles particularly in the early phases on this technology is a small part of what's going on again the people and the process aspects of this are much more important and getting to nirvana how does one get to Carnegie Hall the answer is of course practice practice practice so we're approaching the top of the hour and I'll do a quick little summary here and then we'll get to the question and answers again a data strategy specifies how the data assets of your organization are used to help support the organizational strategy but of course if you don't know what a strategy is and not knowing what a data strategy is or how they work together you're going to have difficulty implementing it and that's what we see most organizations struggling with data strategy is also the thing to keep your data governance organization focused on objectives I've seen so many data strategy meetings where people are going wow we're talking about this we're doing that and one of the more popular talks I give these days is called rekindling your data governance because people kind of get bogged down and if you have a strategy that says we're going to help improve our data we're going to help improve the way people use our data so that we can take that improved data and those improved capabilities and use them in support of the organizational strategy governance becomes much more focused and much more practical a lot of people will tell me that's not what data governance is about I'm sorry I don't care what data governance is about I care what is the organization achieving and how can data be used to support that we talked about a number of prerequisites I went through them fairly briefly here lots and lots of more detail on some of these others and some older websites sorry older webinars that you can go back to and finally when you do get to the point where you've got the ability to go in and start to practice this like sitting down with a piece of music and trying to play that song every single day as you go through the process of focusing in on the theory of constraints and finding out where a lack of data capabilities is hurting your organization's ability to achieve its strategic objectives and then taking small careful steps to move those things from the negative column to the positive column that is where you will succeed as you're starting to do data strategy and again you can see this is about not the plans but about the planning process and that's what will help your organization in the long run achieve success that it's likely to do. Again I apologize for the technical difficulties at the beginning of this thing we didn't really miss a lot you'll get all the slides and it's now time to turn it back over to Shannon for our Q&A session Peter thank you so much for this great presentation especially after the rough start as you mentioned I appreciate all the patience there and if you have questions for Peter feel free to submit them in the bottom right hand corner of your screen and just to answer the most commonly asked questions again I will send a follow up email to all registrants by end of Thursday for this webinar with links to the slides links to the recording and anything else requested throughout Peter so please discuss the role that metadata plays in data strategy Sure let me jump around for a couple of different slides here to emphasize a couple of points here really good question metadata is generally referred to as data about data and so the role of metadata within here also speaks to the general way in which our organizations are not approaching data and I glossed over this point a little bit earlier but I think it's worth going back in repeating in 2017 Randy B. from Davenport survey that they do and again they do this every year so we'll get the new one sometime coming up 37% of the firms that were out there self identified as being data driven and in 2018 that number had dropped to 32% and for the 2019 survey it was 31% so I think this is actually kind of good news and the good news is that people are understanding things like metadata and realizing they don't have a good handle on it one of the documents I'm trying to create and maybe this is where our wonderful data community can help us out is I'm trying to draw a picture of all of the things data it's kind of like the blind man and the elephant sort of scenario so we've got data scientists that think I've got everything to do with data and there's nothing else important in here except for me on the other hand I went to a data engineering meeting in a city recently where their definition of data engineering was ETL and they said if it's not ETL we don't want to talk about it well each of these are correct but each of them are also the blind people coming up to the elephant one's picking up the trunk and saying well you know this thing with elephant is kind of like a snake right it's got sort of long been and moves about and wiggles and all this sort of thing and somebody else has picked up the tail and said no it's more like a broom it's got this wispy thing on it right all of their perspectives are correct but none of them have a full perspective on it metadata about data needs to be just recognized as data so the main thing that most organizations have trouble with from a strategy is that they will say something like our goal at the strategic level is to have the metadata for the entire people-sauce system by the end of the third quarter and while that may be a good goal if you don't say and that data will help us then to reduce the amount of time that we spend messing with the highly customized modules of pay and personnel in there then there's no point in it so metadata is the idea that we are going to be managing some data with extra care using it out of perhaps a higher level of abstraction as we do it and being able to work within those particular constructs if we don't have good management of our metadata there is no way that we will have good management of our data great question I hope that answered it definitely so this is more of a statement Peter but I'm going to read it here for you for you to comment on in real life most of data knowledgeable people such as people who join this seminar don't have too much influence on the to the organization leadership decisions data strategy is made by high management if they don't really know the correct concept how and what do we do that is a great question and a wonderful way of expressing something that's been frustrating us collectively the community for a long long time the only thing we can do is to show that we can achieve results if we achieve results people will pay attention to it and as we pay it if they pay more attention to us we will start to get invited to meetings that perhaps we weren't invited to before because you're exactly right as I said my data decisions slide that's back there that is a huge huge problem for organizations where they just don't know the decisions that are made to prioritize some things over other things is an issue I go back to my slide of defining data governance because it's exactly right and and by the way part of your strategy is got to be focused in on helping the people in the data governance organization also understand this so while most people appreciate a discussion around managing data with guidance it's really the decisions and those decisions are so important and if we are not at the table for those decisions then that's a problem for us I can relate a specific story and I can feel the pain for the questioner in here I was with the Defense Department for a number of years and I had a wonderful boss who would say to my boss my boss is boss I guess and say hey can you give me a one half page decision memo that I can get generals to read in the middle of a decision so they'll make the right decision and I need that by the way on Friday morning at 10 o'clock when the meeting is and my boss blesses Hart was good but he would come back and give her a 20 page thing in fact can't condense it any more than this it needs to be 20 pages well I'm sorry a general officer is not going to read a 20 page document any more than your CEO or your CIO is going to read a 20 page document so we've got to get people into those meetings that understand what they're doing part of it's a maturation process but the other part of it is the more you demonstrate value the easier it will be to have people invited to various meetings I agree 100% with the sentiment and it is a tough tough problem we have and you mentioned you know that one one of your books had the top data job in the title you know what and so what is the top data job is it really a CDO or is it sometimes the CIO CTO etc a real good question I have seen data people work at all levels of the organization my definition for the top data job is the highest person with the highest responsibility for data in the organization I prefer to call it that and there's a couple of reasons the CDO even though we kind of you know accept that as a title the first thing that happens in most organizations is that they say do we need another chief around here and when you're having that discussion you've completely derailed the purpose of it which is that all you were trying to do was do more with data so that you could help the organization achieve more one of my slides a couple back it said that 60% of data people still report in to the IT organizations out there now CDO is fine I like the title enterprise data executive because that can exist at any level of the organization and so I would say to all of you out there who are data leaders go ahead and make up a sign and put it on your desk and say I am the data leader for this organization because sometimes sooner or later somebody's going to come by and say hey who told you you could be in charge of that in which case you can pick up the sign and say oh well if it doesn't belong on my desk whose desk should it be on after all there is somebody in charge of IT there is somebody who's in charge of HR there's somebody who's in charge of finance if we're going to say that data and IT are managed together and expect to achieve better results than we have been which I think we'll all admit are not optimal then we're really shooting ourselves in the foot so make that sign put it on your desk and that sign will help people to crystallize the nature of the argument while IT is important and IT leadership has done a lot of good data has been less mature than those areas so a chief data officer is a fine way to say it it's now part of federal law most I think we have 26 states now that have chief data officers that are out there lots and lots of things that are happening but it's not happening fast and it's not happening in a way that results are out there I mean think about this for a minute every state is having to relearn the lessons why aren't all 50 states getting together and saying how can we use our data more efficiently to support our citizenry and they insist on learning it themselves because they are afraid to share information back and forth if we describe a class of data leaders in here as chief data officers then at least we could have a chief data officers conference for the states by the way I know they did have a conference they just didn't accomplish much but then getting together on that because they had sort of the wrong motivations around that event but we're working on it working on it with all of them anyway data leadership is something we're going to be working on and continue to work on and is not again a fully baked idea just yet so you touched on this a little bit in your answer there could the existing organization structure hinder the implementation of good strategy well the thing that is hindering implementation of data strategy in most organizations and again I do have extensive experience around this and some good scientific data that we can look at as well if you could see the numbers on it data cannot exist as a project and so even though we've taught our students for the past 50 years that you can with one systems development lifecycle approach create all of your data and all of your IT needs out of a single lifecycle we've seen over the years that that is simply incorrect and that data in order to exist as an asset have to have a programmatic management around it and managing data programatically is completely the opposite of the way most IT shops should function so we have an inherent disconnect between the structure of these two things IT projects are perfectly appropriate should be done do a great job we're getting better at it data is not a project and until we recognize that data to be most valuable has to exist at the programmatic level and that's what data governance programs are designed to do is to support data as a organizational asset if we don't have that in place then we can't count on much else happening good it was terrible English wasn't it Shana well forgive it you couldn't hear me cursing I was trying to get the machine up and running no worries so you discussed the process is more important than the outcome and agree with that but what do you think are the most important deliverables of data strategy study fantastic question so let's go back a little bit that one hey this is pretty cool I'm using my new MacBook and I can touch buttons on it will take me the right slides that's really neat anyway the outputs from this should occur in two places I should change something in the business now here's where we have trouble as a profession I'm certainly not speaking directly to anybody because many of you have done some really really good things in your own areas but if I just produce a data result we have a thing that happens data things happen in the organization and what we need to do is to show how data things happening in the organization because we tend to stop there and say hey let's do the metadata for PeopleSoft or whatever it is we're trying to do we need to take it the next step which is more work is more difficult takes longer but will result in better value and say that the value of that PeopleSoft metadata is this so there's two things that have to come out of each of these cycles that we do one something that is a tangible business value if you're producing something that doesn't produce something of business value the organization is right to ask the question why are we doing this but the second thing that happens as a result of this as well is that every exercise that you do should have increased documentation reporting where when you go in to change something to move something to enhance something keep that metadata and add it to your corporate business glossary to your repository whatever types of things that you have in place that make sure that this reference material is available for the next individual I can tell you a story of one organization that I worked with that literally had gone out two of the developers had gone out to lunch and figured out how to do some tax calculations and they brought it back to me because they said hey Peter we know you love metadata here's some metadata for you and they threw down a napkin literally stuff scrawled on a napkin and you know what we took a picture of it and put it in the metadata repository and it was the most highly accessed piece of metadata that we had because it was describing a really important process and everybody went whoa they've already done it once we don't need to do it again so we've actually got something that we can have so the outputs from these cycles are increased knowledge of your people which means you have to have a baseline knowledge in there we're going to get into that in a little bit another whole category that Shannon and I are plotting around to do this but what is the data literacy in your organization up to a minimal level where at least the decision makers who are making decisions about this understand what they're doing so output has to be a combination of business value some additional metadata clarifications more detail about what you're doing and also some improved knowledge around people who are making these decisions absent those three things yeah you're right it's a very tough process I love these discussions and great questions coming in here if you if you're using data centric strategies get projects done about 50% or more faster then won't you grow a value grow as well yes although I'm not sure where the 50% did I say that in there I don't know but you're right put a put a number on it anyway yes and that's the whole point we want to make things faster better and more functional for our organizations and so the idea is if we can decrease the time it takes us to do these things again my example that each of the 50 states wasting their time relearning the same lessons over and over again that they could easily share among themselves banks are doing it already they're being able to focus in on these things the idea is if we can increase the rapidity this actually moves us more towards the goals that people say when they say why aren't you doing agile now again agile is a good way of developing higher quality software faster but agile is a project mentality and only when you have good known data quality elements that you're starting that agile sprint with should you proceed on the sprint because if you move down that sprint without good data the only result is more small piles of data and again guaranteed employment for all of us on the phone call here got to love that so how would you demonstrate that a decision is a data decision if the organization views it as some other kind of decision that's a great question so the organization is looking at buying a new software package a important process of evaluating software packages should be that your organization should request from any vendors that are proposing to sell you some software that they also give you a logical data model of the package because that should be part of your evaluation process so while you may buy a cheap package that doesn't cost a lot of money it may be incompatible with your existing data and will actually the total cost of ownership go way way up instead of buying a package that was more compatible with your organization and having the integration costs be much much lower that's just one example there's lots and lots of them and again this is the purpose of our community here so that we can all contribute these examples so hopefully everybody on the call here will come away with a lot more knowledge about what it is we're trying to do hope that answered indeed and would you say is accurate to state that data management is the umbrella and broader category under which data strategy and data governance fall could you repeat that one and make sure I heard it correctly sure would you say it's accurate to state that data management is the umbrella and broader category under which data strategy and data governance fall yes 100% so again I don't tend to pick nits about the labels but I agree with the questions labels here on that which is again to say that data strategy is one of the important things that data management does I like to just call it data at the top because that's usually most everybody's understanding of it I mean again several people have referenced to our little bubble that we live in here in the data world which is a good place to live and we're making some progress here but most people wouldn't know the difference between data management and data science at this point and that's sad that's how level how wide the misunderstanding of this area is throughout the entire world and frankly I've seen a lot better results of places that did not have the education systems that we've had here when I go overseas and work with groups that are there they do it because it makes sense but we've actually taught them wrong in the traditional western style of describing IT projects to them absolutely and Peter where can one find an example of data strategy well if you google the phrase data strategy you'll see several organizations including parts of the federal government that have lots of data strategies on there I would also say that if it's longer than a couple of pages you might want to read it once but not really do much with it a good data strategy should be a very clear statement about what you're going to fix next at the programmatic level so that you can do data better in your organization let me go back to that one slide where I have the actual specifics of it on there it's just very important to not try to over plan these things because everybody has a plan until they get punched in the face as Mike Tyson has so hopefully told us our data strategy should be a very high level guidance we're going to do this we're going to start to learn about how to do master data from a strategic perspective that's very different from we're going to implement an MDM project right it should also then be focused on specific business goals not I'm going to do this for the sake of coming up with something nice but I'm going to do something in a way where somebody in the business says oh my that was extremely helpful to us by being able to do that and then finally of course you're going to have guidance when you're faced with a stream of decisions if I'm a soldier in Napoleon's army I know the first thing I have to do is hit them very hard and the second thing I have to do is turn to the right and the third thing I have to do is turn to the left if I get any one of those three steps wrong I will probably in this case die of course we don't want to have that happen dying on our data strategies but we do need to make things as basic and as simple as that because everybody's got to get the understanding of it I love it and if you have a question for Peter feel free to submit it in the bottom right hand corner of the screen we have everyone a moment here as we that has been all the questions so far and just a comment Peter you're talking about practicing and you practice your music that's crazy of course I do alright well that is all the questions that we have for today Peter thank you so much for another fantastic presentation and kicking off the new year with this topic again just a reminder I will send a follow up email by end of day Thursday with links to the slides links to the recording of the session and hope you all have a great day thanks for the patience as we got started a little bumpy this morning but we got it going so it was great thank you and thank everybody and we'll talk to you guys next month what's our topic next month what is our topic next month let me pull it I got it there we go that's a good one looking forward to it thanks everybody bye