 Hello and 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 Management versus Data Strategy. It is the latest installment in a monthly webinar series called Data Ed Online with Dr. Peter Akin 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, he 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 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 tie icon in the bottom middle of your screen for that feature. And to continue the conversation and networking after the webinar, just go to community.dativersity.net. To answer the most commonly asked questions, as always, we will send a follow-up email to all registrants within two business days, continuing links to the slides. And yes, we are recording and will likewise send a link of the recording of this session, as well as any additional information requested throughout. 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 and 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. The most recent is your data strategy. Peter is experienced in more than 500 data management practices in 20 countries and consistently named as the 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 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. Hi, Shannon. So great to be here with everybody. This is a makeup session for an error when I grabbed the wrong slide deck and didn't notice I had grabbed the wrong slide deck. So we are going to do Data Strategy versus Data Management today. And what we're going to talk about is, first of all, place both of these two really key topics in context here. And one of the challenges you have to understand is that because we've had a really severe lack of educational correct focus on these things, as we're really teaching stuff either wrong or we're not teaching them these things exist, that has made it a very uphill battle because you would assume you were reminding people of things. Remember when you learned accounting? Remember when you learned page layout? These kinds of things. That doesn't exist here. And as a result, there has been confusion between business and IT and where data fits into that whole process. So then we'll dive in quickly to the Data Manager first. What is it? Why is it important? And what is the current state of the practice? These are some functions that are required for effective data management practice. So one of the cool things is we do have some objective criteria that you can apply in order to do this. Then the second part of this is Data Strategy. And again, we'll look at a structured approach, but understand really the need for simplicity because it's about making sure that everybody is rowing in the same direction there. There are some foundational prerequisites that are important for applying data strategically. And that most organizations aren't really aware of them. There's a well-accepted theory, however, called the Theory of Constraints from a book called The Goal. Some of you will recognize it when we get to it that we'll use in order to talk about this. I leave this last section in here because most of the time people take, okay, Data Management, Data Strangely, let's look at them in action. It's really more important to look at them in concert. And that's what we'll spend the last few minutes talking about on this, looking at some particular coordination prerequisites. So let's jump in and get started. Get a little bit of context. IT, even today, often thinks of data as a business problem. And they approach the process because of that mindset as glibly described here, if they can connect to the server, that everything must be fine. And I'm oversimplifying and hopefully not insulting anybody, but usually I get people that go, yep, I don't know what's going back on the pipes, but as long as they can connect to SAP or whatever it is we need to connect to, my job is done. The business thinks that IT is managing data adequately. The idea here is of course that, you know, why else would you have somebody with the title Chief Information Officer? And truly those titles in the past are problematic and continue to be today. Probably most CIOs are Chief Information Technology Officers, Integration Officers, several things, but they are not the person who is managing the organization's data to its highest and best use. Not because they're not competent or not doing things, but because we're asking them to do so enormously much as we go through this. So the consequence data has fallen into a gap between data, excuse me, between the business and IT on this. And this gap needs to be repaired because if we don't repair it, we have no hope of actually achieving what most organizations are attempting to do, which is do better with their data. So we also have to understand two really key things about data. Obviously better organized data increases in value if I don't have to put my hand on it. And if you have doubts or trouble describing these concepts to your colleagues, just imagine being handed the pages of a book that weren't randomly placed in a stack as opposed to stapled together, held together with a spine and ensuring an architecture that says I'm going to start at the beginning and go to the end, which is the way most people want you to read their books. So bad data management practices are costing organizations much in the way of time and effort and it's even more compounded because minimally 80% of your organization's data is rot. Now that sounds terrible, rot is an acronym that stands for data that is redundant, data that is obsolete, data that is trivial. My wife corrects me on this and says actually it's incomplete, obsolete or trivial, whatever we're going to do, the question is which data does one eliminate on this. Because if we got rid of four fifths of our data and by the way the only argument I ever get on this specific point is that people say it's not 80%, it's 85%, it's 90%. So we're in the right direction here for sure. Now 80% of the data being rot, we should get rid of it and try to work with the rest of it. The rest of it does have some different characteristics. Data is the most powerful underutilized poorly managed organizational data asset. It's the only asset that you have that isn't depletable, it doesn't degrade over time and it's durable in nature and notice I'm comparing it to other types of things up here on the screen, financial assets, real estate, inventory, et cetera, et cetera, data assets outperform all of these. They really do win when you compare them against these other assets. There's lots of other comparisons that can be made but many people then go too straight. Data is the new oil. This is frankly one of the worst ways to describe data because people think of it as a consumption function and something that goes down. I like to replace the word oil with the word soil and say there's two things about data that are critically important. First one is you don't just randomly take apple seeds and spread them around the yard and expect that you're going to have an apple orchard. You carefully prepare the soil, you nurture it as it's grown and this includes the second mention that is also different. In the soil example it takes time. Nobody expects you to plant apples on Monday and eat them on Friday and yet we do have unrealistic expectations about how long and how much effort needs to go into the various data projects. Now if you need to call it baking in your organization to make it work, that's okay. But data as an asset, as your soul, non-deplatable, non-degrading, durable, strategic asset, data your new soil, data deserves its own set of strategic considerations and it deserves attention on par with similar organizational assets. And finally it does require professional administration to make up for the past neglects that we've had in this area. Now just to show you this is reasonably important is that our former governor of Virginia put forth pieces here that says look you've got to do this stuff because it's critically important. He also says it imposes on people's privacy and we need to come up with common vocabulary. So these concepts of best data management practices that we're talking about in this whole series of webinars are becoming accepted in state and federal government and that is important because it will literally get us on the same sheet of paper because right at the moment most organizations have little idea what data they have. They don't have an inventory. They don't know where it is. They don't have a map to find it and they don't know what their knowledge workers do with it which is hugely problematic. So let's sort of summarize on this. First of all if we're going to do data we understand we're dealing with volumes. These volumes are large and quality engineering and architectural work products simply don't happen accidentally. They have to be properly developed. And yet when we look at this data management happens pretty well at the work group level. It is in fact a defining characteristic of what a work group really is. Without guidance however what are the chances that all the work groups learn the same much less all the individuals in your organization. Imagine trying to learn French by randomly calling up people in France and getting different French accents. It wouldn't make the easiest way to learn the language get on the same piece of paper. Especially also consider the time that is spending our organizations as Tom Redmond calls them the hidden data factories that people have created over time to deal with problems because we've been dealing with these things informally and we have to be formal in the way we deal with them or it will not be effective. Data chaff then becomes sand and it prevents smooth inter-operation and exchange of organizations literally leading to death by a thousand cuts in many organizations. Twenty to forty percent of all IT expenditures are working on these hidden data factories. The reason for that is because organizations and individuals lack the knowledge they lack the skills and they lack the data management know how in order to put this in place. They also don't have a strategy because if we don't have a full strategy set up for these individuals then how can we tell them what they're doing is correct or incorrect. So let's move on to the second component of this that was data management. Mostly it is misunderstood. Organizations oftentimes make fun of us in the data management community and perhaps rightly so thinking that we're librarians and going to calculate all the books in here and everything that we find. Or if we just label everything this actual farsight cartoon was given to me by somebody at the Portland I believe. Data chapter many many years ago but it's kind of fun as well and Microsoft was even making fun of us back in 05. Here's some data modelers and look where the only place they can get their white board is it's not exactly painting us in the most flattering ways. So to talk about from a data management perspective what we've got is the ability to look here development and execution of all of the architectures policies, practices, procedures that manage the full life cycle of data. An easier way to think about it is quite frankly that data needs to be managed from input when it's first captured to the processing that it has all the way throughout until it's used in analytics or otherwise type function on this. So input process output and what we're really doing is we're processing data but nobody will use those terms because they somehow seem bad on this. This is our proper definition for this and yet it doesn't really talk specifically about technologies and that's really important because this discipline is really about architectural focuses up here however it does extend down into the data pieces. Nevertheless there's still problems with this because if you go to the colleges and universities what you'll see as an example like this being given by students the problem is these discs no longer rotate and the students understand this and realize that professors don't really know what's going on out there in the world. So there's an enormous disconnect between what people need to learn about data and what they actually learn about data in college and university projects. I'm speaking broadly of course there are pockets of excellence but in particular they far are in the majority of these. So let's take a look at a data model in particular here and the idea is that a software program may have a specific focus and that we may produce a data model, a data structure that would be used to access the data of these programs. A program can access zero or one or more different data feeds that come into it. Another underutilized effort would be here but if we're going to have all of these programs, this family of programs, programs A, B, and C all access the brown database then it would make sense to look at them as a unified whole and the data architecture focus can focus in on all of these programs collectively. When you're marketed enterprise research planning software, the ERPs of the world or any commercial office health software, these are all marketed as similarly integrated however they are likely internally integrated but between external entities you need to make different connections, different arrangements. So the better utilized, more effective, better leveraging effort is to focus all of these in the broader service over here as a family of programs and that the architecture component would comprise program A, B, and C in this case but really within a group we have to make sure that somebody is doing this at the higher level so that the orange and green databases are also connected similarly on here and our data focus really should be broader than either software or database architectures, the entire scope of the system as we work throughout the entire to have it and focusing specifically on problems caused by data interchange or interface problems. So the goals of the architectural effort are much more strategic than they are operational. So the key for data management is to retain an organization wide focus, make sure that is your requirement to understand and understand means machines understand the same as people, business people understand the same as technical people and let's take it another way forward to not just understanding but also understand the current and the future needs because the goal of course is to make the data more efficient and effective and be better able to leverage this in support of other activities. Notice that the little diagram to the bottom right hand corner includes also people, process, and technologies in here and our most recent surveys, thank you Randy Beam, published at Forbes this year of January 2019 shows that organizations are now rating this as a people in process problem and that's 95% of their problems and technologies are only 5% of the problems that they have in the data environment. So again data management understanding the current and future needs of the enterprise and making that data effective and efficient in service supporting the organization and yet when we come back and say what is a knowledge worker, my definition of a knowledge worker is somebody who uses data and of course what we teach them is generally nothing and yet what percentage of them deal with it daily the answer is 100%. Similarly when we teach IT professionals about this they get usually at the most one course in how to build a new database. Ladies and gentlemen if we do not need any more database programmers on the planet we could survive for the next 20 years with no trouble at all. No disrespect to anybody that is building databases but it is not a skill in demand when you compare to the proper application of the other aspects of data management and by the way our database programming staff is the best source of information as well as expertise. They make wonderful additions to the rest of the data management team. Now more importantly than this if I've only known as I go through a good college and university program and they tell me data build new database that's great and our bosses go through this as well we've done an incredible disservice to literally humanity. Sorry to get so morbid about this but these people understand that we only need data people when we're building a new database. Now two things happen here one these people being taught this is a hammer and if the only tool you know how to use is a hammer every problem looks like a nail thank you Abraham Maslow or our bosses we're not needed for many of the kinds of things that are appropriate and important for us to use here. So there are tools out there besides databases but our data people are generally not taught them and our bosses don't know that we should look to our data community for expertise in this. Now we've got a little beyond this we're at least better at a higher level now we understand that this is what we mean these functions that we have here are the data management functions we put the first version out in 2009 we've updated it recently in 2017 you'll see there's an addition of one component over here on data integration and interoperability but the rest of the body of knowledge is still fine which means we can now comprise architectural specific objective criteria against it because data management has been expanding in the early days when I learned it database operation database development were the only components of it but then we've expanded to data administration enterprise data administration looking further and now since 2000 so we've still tried to get better but we're not better unfortunately we are not getting better because we're not teaching this because it's not being researched because we are not putting time and effort into it I get people who can work on these efforts part-time as opposed to the full-time effort that we need to go back and undo the past damage that has been done to our organizational data inventories let's move on from data management now to data strategy data strategy here is talking about the ability to understand where this goes and yet what I see time after time the number one two and three strategies that people tell me they are using in their organizations are data science big data and analytics well I don't know if you've seen or not but data science is becoming more and more problematic because it's defined at the long level of abstraction happy to take that discussion offline data is producing the same results as general IT project work in general which is a 70% failure rate and analytics nobody really knows what it means so it's very hard to point it out as a strategy the next ones by the way that I get down there are these technologies which also are not a strategy either so people do not understand data strategy at this point and let's again see why organizational strategy it made sense to make an IT strategy and from there we made a data strategy I'm sorry but it's just simply wrong and the reason in particular is because data is a program and data has to exist at a level at least on par with the IT strategy and in most organizations who preceded it it's not that IT strategy is not important but data strategy will drive aspects of IT strategy and that's how you get that 20 to 40% savings in IT strategy now let's just take the word strategy for a minute and focus in on it we had talk if you haven't had a chance to see it I've got the reference here obviously on the slide is Simon Senex how great leaders inspire action and the two minute summary is that human beings are pretty good about telling you what you do but we're less good at how we do it and we're not very good at describing why we do things now it's just one of those things that it talks about what are we doing and people end up being inspired that's the definition of inspiration so it's not what you do it's why you do it it's catchphrases in here imagine if Martin Luther King had given the not I have a dream speech for the I have a plan speech it's just not the same thing so that inspiration that motivation that why that is strategy that we're trying to get to and it's got to be simple if we look at strategy the use of the word strategy originally came from the military but around 1950 the business folks got involved in it and started taking it into the business world but the best definition by far is by Henry Menzberg which is that strategy is a pattern in a stream of decisions I'm going to give you three very quick examples first one Napoleon is fighting a bad guy excuse me and a larger army at Waterloo in here the British are in the orange the Prussian are in the black and the French are in the blue so the question comes up how does one defeat the competition when their forces are bigger than mine the answer of course is define and conquer a quick look at his approach to this problem which is still taught today in military academies all over the world what Napoleon observed at this point remember he's in blue the British are up in the left hand corner in red and the Prussians are in black and the British had their supply out of a stem because that's why you have spies so they go in and find out this information and the Prussians were supplied out of Liege now I want you to imagine if you get hit in the face do you fall away from your food or towards your food I know that's kind of an abstract question so let's put it in tangible if I hit these two armies exactly right Napoleon thought I will cause them to become discombobulated getting hit in the face is no fun at all and when they become discombobulated they are more likely to go towards their supplies then go away from their supplies this of course being brilliant thinking and at the same time perfectly good common sense so his first approach was let's go in there and attack them and what I do I need all my army to turn around to the right and attack the Prussians and then turn to the left and attack the British because if half of them go for the British and half of them go for the Prussians I won't survive this of course you guys know the outcome of this he did not actually survive this nevertheless it is still taught as a brilliant example of strategy now let's go a little bit further one of Canada's biggest athletes is Wayne Gretzky and he has a definition of strategy in that he says he skates to where he thinks the puck will be if he's following the puck the puck is fast they're both on ice good luck catching up there's a wonderful articulation about this entire process at Wayne Gretzky's Wikipedia entry but again a very simple description of strategy that everybody likes our third example of strategy is a Walmart former business strategy that you all probably know because Walmart's former business strategy was known to their customers to their suppliers to every vendor that dealt with them to all kinds of people because if you didn't understand that every decision at Walmart every pattern in a stream of decisions defaulted towards providing the customer with lowest price then you really didn't understand how Walmart worked and they are assiduous about this process they did a phenomenally good job on this now these three strategies are simple and yet what happens here is that many organizations actually screw this up because they'll say okay great well I know we want to do some strategy work and strategy always results in one of two outcomes sorry one of three nothing that's the first potential outcome but if you get some good outcome it either improves operations or it results in innovation we've studied this for years these are the two definitive outcomes that happen as a result of strategy so if we take our standard four quadrant chart of v1 like the lower left hand quantity organizations without a formal data strategy are going to be pretty problematic again we've seen that already so we'll accept the premise that Walmart is phenomenally talented in terms of its logistics areas and that they are very good with increasing organizational efficiencies and effectiveness and let's take in quadrant three the apple quadrant of apple making innovative products and when apple makes innovative products they use it to create strategic opportunities as in inventing iTunes and many other aspects of society that we currently take for granted today now the point here is to say that we shouldn't try to do both of these things at once I want you to imagine Johnny Ive who is the very bright British chief architect forward chief architect for Apple who would come on and do the narrations of the new iPhone or iPad videos again talking about what a beautiful product is and say I'm sorry Johnny Ive you've got to be increasing efficiencies and effectiveness here it would blow his mind and similarly if you took the really excellent people at Walmart who are phenomenally world class best at squeezing efficiencies and effectiveness out of their supply chain and said I want you to be creative it doesn't work so it makes sense to sequence these things and unfortunately most organizations say let's try to do both at the same time with the same group of people again it doesn't work strategy has to be simpler so the proper approach in most organizations is to use the V2 quadrant to squeeze out some extra savings in there use that savings to fund the things that are going on in V3 recall again strategy is a pattern in a stream of decisions it means everybody has to understand this all the way from the top to the bottom and yet we don't understand that data programs need to proceed software products this is the whole reason for these two strategies being separated because data evolves at a different cadence a different rhythm a different pace than does applications application development processes are very good at creation activities and sorry I had a third joke up agile is one of the more effective approaches at doing this kind of development work and as much the question I get asked over and over again is how do these two interact and the answers they absolutely do not one process takes literally months and quarters and years and some organizations decades and the bottom one is trying to be accomplished in two weeks spreads so they have to be separated and made sequential data is not a project I'd like to show this vision of my barn couple years back now I'm what's called a horse husband and part of the deal is that I built a barn with my spouse here she actually designed it and I provided the yes on this but one of the things we did was take pictures of the barn here people stop at this point and say I'm sorry I thought we were talking about data management strategy and the answer is yes they are perfectly complementary in that they need to be integrated in order to do this particularly so the barn example here is that the bank also did an interesting thing with this which we should probably adopt over here especially when we're thinking about development projects so the barn excuse me the bank the bank required that I have a foundation inspection before I go on and complete the rest of the barn and more importantly before I get the rest of the money that they are loaning me to build this particular barn so the foundation can be inspected according to some criteria it was actually passed by a building inspector for Hanover County Virginia which is the law and when I had that certificate I was then given the next set of money why is this important well the bank understood explicitly that if I built a good barn on a poor foundation my spouse would have me pay for the vet bills before I paid off the bank loan so the bank thought that was very bad for business so the bank said we're going to guarantee you that you're going to build on top of a good foundation which is our best hope that you will do exactly what needs to happen correctly in here and interestingly enough there is no IT equivalent for this and because there is no IT equivalent it amounts to introduce this incredible confusion into what's going on in the whole IT spend again returning to Oslo for the second time in this webinar let's just review Matlow's hierarchy of needs and then I'll relate it to data management technologies and technology and practices in general so Oslo's needs are physiological needs if you do not have food clothing and shelter you will be unable to be safe these are necessary but insufficient for requisites if you are never safe then you can never be part of something that is bigger than yourself and if you're not part of something bigger than yourself it's very hard to develop a set of identities your own esteem and of course all of these are necessary but insufficient prerequisites to actually getting to what everybody wants to do self actualize or what is currently called slow in today's environment now I bring up Matlow because data management is very much like this what you see advertised talked about blogged about linkedin about are all these wonderful things if I was going to bother redoing this chart I would add two concepts datecoin and blockchain to the top of this technology piece because these are the things that people are talking about and this is driving discussion and while that is good and wonderful those things represent just the tip of the iceberg and if you don't understand that you need to put in place good foundational practices then you will end up focusing in on the 5% of things that are data related per Randy Bean's early article on this and ignore the organizational capabilities which are much more important to your organizational success with data so in many ways people trying to do bitcoin or blockchain or MDM even are trying to self actualize without sufficient basis in here and this accounts for in my estimation minimum of half of all failed IT projects throughout the entire world we are always asked data blueprint yeah yeah Peter I know you heard you do that but can you do it by Friday can you do this faster and I say yes you can accomplish this faster but if you do it faster it will take longer if you do it faster it will cost more if you do it faster it will deliver less and if you do it faster it will present greater risk to the organization than if you crawl walk and run your way up the data management maturity curve so thanks again to our colleague Melanie Mecca for developing this work and our partners at the CMMI for helping to put it out there in the public domain the five data management practices areas are the ones that I had at the bottom without those foundational pieces and that means that you are then able to manage your data coherently according to objective criteria you are able to manage your data assets professionally that you are able to calculate what is fit for purpose so that you can scope and bound specific investments in data improvement that you do it with right technology and with the right processes because if you don't and don't have the support of your organization you will not be able to succeed and more importantly here data management practices architecture is only as strong as the weakest link so in this thing on the lower left hand corner there's a weak link on the left hand side and I'm going to show you a very specific example of a weak link here in the sense that for some form or another this organization has determined that data governance, data quality data platform architecture, data operations are all operating at a level three it is for a different program to describe what a level three is but let me just say that it's a very good achievement for organizations and yet if they don't have a strategy the entire performance of their data management practices will be only at a one level because data management is only as strong as the weakest link and consequently a day's strategy that winds up on the shelf is not useful at all so let's talk about how the theory of constraints comes into play here based on one of my favorite books that I've read the goal by Eliah Golderat and I've sold quite a number of excuse me quite a number of stories books for Eliah Golderat out there you can buy it on Amazon yourself for $1 plus the $3.99 shipping but the goal tells the story of Alex Rogo and Unico Corporation and it presents management in a view that says look something's happening in your environment that is most blocking you find it fix it and repeat because the chain is also no stronger than the weakest link notice how complimentary these two philosophies are and this weakest link is what really does permit us to understand how to utilize our data in its best way and we need to apply this in a repeatable process as opposed to coming up with a grand plan where most organizations pay for a large amount of money to be spent creating a 90 day or 500 day or one year or a five year plan for how they're going to do things with data from this point on out at the beginning of their journey where they know the least about it the least effective way of doing this instead where it must be approached is through a repetitive process so that is theory of constraints so I'm going to go round the thing for you once with this in the generic fashion which is to say to identify some specific constraints components of the system that are limiting our ability to achieve our strategic objectives and then make quick improvements to the constraint using existing resources if you are able to solve it at that point then go back and identify the next constraint if you are not able to solve it at that point it means it's a bigger problem than you thought and you need to review other activities in the process to facilitate proper goal alignment and support of the constraint as clearly this is the only thing that the organization can focus on until it is fixed or decided it's not all that important if the constraint still persists we have to identify other actions restructure in other words in order to make ourselves available to address the complaint and repeat until that constraint is eliminated so that's the generic version of the goal let's look at it from a data perspective which is to say in your organization understand best how data can support the organizational strategy and identify the one thing that is blocking you the most I'm going to give you a quick clue in this it falls into a talk that I do called triple play that talks about leadership talks about valuation talks about specific understanding of how to apply strategy so one of those three things I guarantee you will be a start for helping to eliminate the thing that is most blocking you in attempting to achieve your data data desires your data specific constraints need to exploit the constraint find out what it is try to rapidly correct it operationally in your organization it is possible that you can do this I've seen organizations literally make good decisions on the wrong data resulting of course in a bad decision that will be a problem if we can exploit that constraint let's do it however if we cannot exploit that constraint by sorry if we doing that does not fix it then we need to restructure our approach to the problem by subordinating all activities to improve all data evolution activities to ensure singular focus on the current objective let's give a quick example if I've got some data in a data warehouse and I'm trying to get some better results on the use of that data in terms of helping to improve sales and the data in the data warehouse doesn't seem to be effective I could put a lot of time and effort into a data quality initiative around that however if I don't also put a data governance initiative in place to ensure that the data is not fixed just once but that all data on the input is fixed and that we apply those changes that are needed back into the production systems that are there we will not be able to support mate all of those constraints in order to do that so again by focusing on one thing which is the outcome we need to have the data that is better supporting the sales cycle we can get that if we can't get it from that then we need to go back I'm sorry I went back too far we need to go back and elevate that constraint such that nothing else becomes the focus again three steps of increasingly importance on this and if those don't work then we need to repeat the process now I'm going to give you an example of this I've got an organization that I've worked with for a number of years that had a Y2K problem although it wasn't really a Y2K problem it was a Y2K like problem the problem was that this organization as good as they were and as smart as they were had managed to restrict themselves to a three digit number and to identify location in their business and this location number was embedded throughout hundreds and hundreds of systems in their operational environment and they were running out of location numbers now in the process of running out of location numbers it's worth it to go back and look at the process because it's illustrative of how this was well intentioned but badly handled in terms of the implementation so they handed an individual a spreadsheet with 999 numbers on it so these are location numbers please give them out to anybody that needs a new location so you're a new location person when somebody comes to you and says give me a new location ensure that you give out the next one in line sequentially but also ensure that you never give out that number twice and that individual did a great job of doing that the only problem was that individual did not understand that as work transitioned from work to telework the associates in this environment found out that they could create a location for themselves which means they could get stuff delivered on company mail asking for a new location so somebody would come along and ask for what was the location number 342 and they would look it up and say well it looks like a residence and they would find out yes there was a programmer or an analyst or a business person who had decided that they wanted to have their own location in there and this only came to light because they were suddenly getting to the end of the number of locations that they had to give out they couldn't open a new store because they were literally running out of location numbers and this was going to be a huge problem I want you to imagine going to the board of directors of this organization and saying to them I know we don't have 999 locations but we're almost out of location information it's kind of like Y2K we need to do something very very different in order to move forward in order to build the organization that we want to build and literally this dumb Y2K like problem is preventing us from opening up new stores to achieve our sales goals even though we know there are lots of people out there that want to buy this stuff that we sell. So the process of identifying this was to say hey we've got a constraint it means we're not going to do this at the same time we're going to implement a planned metadata initiative it means we are not going to do this as we were planning some improvements to off the wall commercial off the wall software that we were purchasing it means we are not going to do this at the same time as many other IT projects and they needed to be put on hold because after exploiting the constraint of saying hey we just need to find out where all this is fix all those places they realized that was a very large number so they had to subordinate all of the non-constraints and say these things are going to be put on hold and so many aspects of what they were doing were simply put on hold so that they could devote the remaining resources that they had to the initial effort. It reminds me of a Star Trek analogy and I apologize for those of you that don't have any idea what I'm talking about but Scotty I need all the power at the engines right is what we were talking about and that elevated the constraint which the organization then restructured to address and luckily they didn't have to repeat the process because it did solve the problem and they also understood now better the amount of governance that is really required in order to properly do this because this is of course the moral of the story the orders to the individual who is handing out the location identification should not have been just give out a new number and ensure that you don't add any duplicates but also make sure they have a valid reason for doing so which is one of the things that data governance will help us in the long run in terms of doing initiatives around this. So the process is really lather, rinse and repeat and I know that sounds awfully glib but doing data strategy it's much more important for your organization to focus on getting better at the process than it is to focus on the exact specific constraint that we're talking about in order to do this. Again lather, rinse and repeat according to the theory of constraints as adapted for data. So that's our data strategy. Now let's look at how they work in concert in here. The first thing to understand is that for years and years and for most organizations it is still the case that business decision makers are simply not data knowledgeable. They haven't been educated. We haven't taught them in college and university. They think data is something that the IT person needs to take care of and we've already seen the IT person have no more ability to gain insight into what's happening there because they don't understand how the things that are in their technology are being used to the business and most importantly they have no interest in doing this. So I'm going to tell the story of the Oracle DBA and I'm going to tell it because you probably have some in your organization. Oracle DBAs are great people. They do a wonderful job but what they have to do is remain I'm going to say loyal to Oracle. Loyal to the technology because loyal to the technology is required because Oracle is constantly upgrading this and making sure that they have a vast amount of people who get paid a good amount of money to make sure that Oracle's fine products continue to perform well for organizations. But when the Oracle DBA has a bad day with the boss and tells the boss to effectively take this job and shove it, they can walk down the street and get the same job they had for their previous employer because there's a need for good people doing Oracle work. And if they have this ability to do this and they get a raise out of it, where is their ability to be loyal to the organization? It is not a good situation here. So the business decision makers are not data knowledgeable and the technical decision makers couldn't care less in most cases or don't know that the business maker decision makers are not knowledgeable about this. Is it a surprise then that bad decisions are made about data? And that results of course in the poor treatment of organizational data assets and poor organizational data quality which leads to poor organizational outcomes. Again, 20 to 40% of your organization's IT budget is being spent in this bad data decision spiral. We've got to break out of the spiral. So in context what we're going to do is look at data strategy as a way of focusing the limited efforts that we have on developing the organizational capabilities to do data better. And our data strategy is done in full support of the organizational strategy. There is no other reason for it to exist. But when we look at data strategies related to data management we talk about what improvements to be made need. Excuse me, let me do that again. We talk about what improvements need to be made in our data management practices and which areas we should focus on for this cycle. And that's the way we will get better about data. And the feedback that goes back says how well is our data being employed to support the organizational strategy? This is really, really key because if we don't have this ability to go back and forth in understanding our data and how it's being used we have no ability to be useful at the strategy level for our organization. In Peter's World IT projects support data management activities because if I'm going to save 20 to 40% of IT costs, and again I guarantee you that is the range that we're talking about here, these projects will be very differently implemented. And the characteristics of what we do with IT and the organization will be redefining how IT supports the strategy. We need to fill in a little bit of context here just to make sure we've got some feedback loops on this, but this really should give you a good idea of how data strategy and data management interact. So let's talk about a couple of quick takeaways in here. First of all, this discipline has not had the 8,000 years that our colleagues in the accounting profession have had to formalize generally accounting principles. We are working on it. Folks like Dave McComb, myself, Tom Redman, John Ladley are all working to try and come up with these accepted disciplines. And there's some very good efforts that are being made. We need to unify all of these around a single effort so that we can more quickly make progress in this area. But we have been around probably 250 years. I'd like to trace the founding of our discipline back to 80 levels. Data Loveless was the world's first programmer who understood what programming was before computers had even been made. 8,000 years versus 250? Yeah. It's okay for us to be immature. It's not okay for us to stay immature. Second, come to grips with the fact that likely your data is a mess and requires professional administration to make up for it. Most of you fall into that category of professional administration which you're lacking are the resources. And the resources can be better easily attained by understanding the relationship between data management and data strategy. The folks that are in your organization outside of your group do not know how to do this because they've learned it on their own because colleges and universities have not provided the level of data education that needs to be implemented. You likely need to implement a completely new business data program. And I use those three words explicitly because this program has been reporting historically to IT and look what it's gotten us. It's not to say that IT is bad but IT has a vast, vast remit. And they're dealing with many, many, many, many, many things. And it's for security lies. All those say that we should have data programs reporting to the business. And the other part of this is it has to exist at the programmatic level. You will no longer need your data program when your organization no longer needs its HR program. Equivocate those two concepts in your leadership's mind so they absolutely understand specifically what we are talking about here. It needs to be a program. It needs to exist with its own budget. It needs to report in outside of IT to the business in order to make this work. Business data strategy and data management components are major components. But they only work well in concert. They are not separate siloed initiatives and consequently it is critical. They must focus on in order by section, by subject area, by however you're going to divide it up but don't try and eat the entire elephant all at once. Improving your organizational data because it has suffered from a lack of professional application. Not that you're not professional on providing it for your organization but that it's very likely your organization hasn't recognized it and it's been done in spite of rather than because of the support that IT has provided for data in the organization. You also need to improve the way people use your data in your organization because none of them unfortunately have been trained to do this. They don't get it as part of data science programs. They don't get it as part of statistics programs. They don't get it as part of analytics programs. They need to understand how to use data in a very basic way. And if you have any doubts of how to do this, how to check it out, how to illustrate it for your organization, just ask people in a room to raise their hands when, if they learned Excel. You will see the vast majority of people in your room will raise your hand. And I'll ask them to keep their hand up if they also learned when they learned how to use Excel. If they also happen to learn that Excel has a built-in capability that will allow them to repeat unerringly, unerringly repeat their spreadsheet machinations automatically, unerringly. In other words, did they learn about macros when they learned how to use Excel? And you will see that fewer than one in ten keeps their hands up. And imagine all of the other things that you know about data management that your organizational colleagues in their various work groups do not understand. We have to improve the way people use their data. If we don't improve that, we will have no abilities to use the improved data that we currently have. One more component of all of this, which is understanding how people can better use data to support strategy and understand full well that because the data is considered of such poor quality, we actually are saying one of the biggest problems in artificial intelligence is not the lack of algorithms to train it, but the data, the lack of good data to train it. This is why Amazon is mining all of your conversations while Google is listening to all of the conversations that occur on your Android phones because there is simply not enough data for them to work on the AI algorithms. And they're trying to capture this in various different ways of doing this. A wonderful read on that topic, which is way off a topic here on data management versus data strategy, but worth shouting out is a book called Rise of Surveillance Capitalism by Shoshana Zuboff. She has done a marvelous job of putting together the argument for this. Because if you improve your data, improve the way people use that data, you're presenting technology in a very neutral form and you're showing people how to better use data to support organizational strategy, which means they now can do what they're trying to do more effectively in some ways. This is kind of awkward, but that's the way it works, right? This can only be accomplished using an iterative approach, focusing on one aspect at a time and applying formal transmission methods, transformative methods. So again, our little icon on the bottom right-hand corner tries to take off slowly and then gradually attains altitude and gets going, crawl, walk, and then run your way about trying to do organizational strategy to influence your organizational data management. So we spent the last almost an hour now at this point talking about some important data properties and talking about the fact that you can't count on people to have had any data background in their college and educational pieces. There's no reason. There's lots of good reasons. There's too many reasons for badly understanding the difference why data has been ignored by both IT and the business. It kind of got dropped in the middle of them. But hopefully now you have a good idea of what data management is. A coordinated practice discipline of formalized methods that we can objectively determine whether they are being practiced or not. But uniformly the state of data management practice is low and needs to improve in such a way. The functions that are required for effective data management are well specified, well understood, but not well instantiated in our organization. So it's going to take decades for us to go back and correct that deficiency. It can best be accomplished by using a data strategy, a structured approach that allows for a very simple way of improving specific data management practices. And each time you go through that cycle that I showed you for the theory of constraints, each time you go through it you are building more organizational capabilities. You have the chance to involve more or improve the people who are doing it's knowledge and skills and getting better about the entire process. Because this foundational data management practices are foundational prerequisites for organizationally leveraging your data. It is absolutely critical that they exist because if not, whatever you're doing will take longer to cost more and deliver less at greater risk for the organization. I've showed you in the example of how the theory of constraints is used to work. That is instantiated in the book that Shannon was talking about Enterprise Data Executive and Data Strategy. So our final takeaways are that the two of these are not sequential activities. They are continuous activities. Data management you will not be finished with. You will be finished with it when your organization is done with its HR finish. They need to be done in concert in coordinated fashion and these coordinations are a necessary prerequisite to getting all of these things done. So we're finishing up here just at the top of the hour. I will turn it back over to Shannon. Peter, thank you so much for another great presentation. Really appreciate you doing this. And just a reminder, we will have a follow-up email with all the information and to all registrants within two business days, continuing links to the slides and links to the recording. Peter thanks again for doing this and hopefully all can join us next month. Thank you, Shannon.