 Hello and welcome and there's been a question out there already for asking who was at DJI key last week. So much fun getting to meet so many people in person that I've seen online for so long. I love it. It's great. So much fun. And thank you all and I got some feedback at the conference that I speak way too fast in these intros. So I'm going to slow it down a little bit. Let me know how it's going to make sure you can understand everything that I'm saying. So here you go. Hello and welcome. My name is Shannon Kimp and I'm the Chief Digital Officer for Data Diversity. We would like to thank you for joining today's Data Diversity webinar key elements of a successful data governance program sponsored today by precisely. We have a couple of latest installment in a monthly series called Data Ed Online with Dr. Peter Aiken. 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 section or if you'd like to tweet we encourage you to share highlights via LinkedIn or other social platforms using hashtag data Ed. You can also share with each other. We certainly encourage you to do so and to open an access either the Q&A or the chat panels you may find those icons in the bottom middle of your screen. And just to note the Zoom chat defaults to send you just the panelists but you may obviously change that to network with everyone. 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 we'll likewise send a recording of this session as well as any additional information requested throughout the webinar. Thank you to Matt for a brief word from our sponsor precisely. Matt, hello, and welcome. Thank you. Sharon screen real quick. Looks good. So welcome. This is a very good opportunity for us to talk to a whole bunch of people about, you know, what data governance is and what data governance isn't why we do it. What we would like to make is data governance for data governance sake doesn't really work. What we're really after is changing business outcomes. And when we talk about that. There's four components we believe from a precisely perspective are critical to a good data governance program and start with the top right business accountability business not taking ownership of the data is the critical and maybe it's like, don't pass code like $200. We're at a standstill if we can get the business to take accountability for the data that they require to run operations analytics finance compliance. Going down around the circle to ensure that every data component is being treated. The same or treated correctly, you need a decision tree. What data what does this data drive what outcomes it's supposed to support. We'll make our governance decision based on that. Going around the circle. If we have an architecture and a lance application landscape that supports what we're trying to do from a governance perspective and from a business perspective. Most importantly, are we getting the right data to the right system at the right time that is fit for purpose. If not, we have some issues to deal with. How do we measure, how do we measure, are we moving the needle from a data quality perspective, and a data reliability perspective and an integrity perspective to ensure that the business outcomes we said we're going to drive, we're actually hitting the mark. One of the other things we want to get into make sure that everybody's understanding of this is governance is not a one size fits all endeavor. Not everything it's in your landscape has to be governed to the same level of rigor that other things. If you think about it, you have the top of the phone this is all the data we have available to us. Of that you probably use anywhere between they did number says 40. I've seen it goes high as maybe 60% and that's really an almost an outlier. 60% you have, you may govern 10% of it, which means you've made a conscious governance decision on that element. All that 10% and this is with an airing Lee consistent accuracy. Most groups identified between 100 and 200 data elements that are critical to how your business operates, whether it's reporting compliance analytics or operations. That is it. And those are the elements you want to make sure you are making a very conscious governance decision on, and how you're going to operationalize that governance to ensure that those business outcomes are being met. Now, who are we, we believe in data integrity. Our software data enrichment products and the group I run strategic services deliver accuracy consistency and context to empower confident decision making. Every journey to data integrity is unique and driven by business initiatives market trends as most of you will know are accelerating really quickly. And the need for integrity of your data is getting more and more paramount our company with our software components and our services groups help you get there. What my individual group delivers. Four things. What your data strategy should be how your organization, and I know Peter's going to get into this, how your organization needs to look. Because once again I said earlier it's not a one size fits all endeavor and if anybody's telling you that run quickly. How you operationalizing the data. And then most importantly, and this is mainly for all your C level people. How are you showing the value you guys are delivering to the business with that Shannon back to you. Thank you so much and thanks to precisely for sponsoring today's webinar and helping to make these webinars happen. And if you have any questions from at you will like guys be joining us for the q amp a portion of the webinar at the end. So feel free to put your questions in the q amp a panel. I want to introduce to our speaker for the webinar series Dr Peter Aiken Peter is an acknowledged data management authority and associate professor at Virginia Commonwealth University president of Dima International and associate director of the MIT International Society of Chief Data Officers. For more than 35 years Peter has learned from working with hundreds of data management practices and 30 countries including some of the world's most important. His 12 books are many first starting before Google before data was big and before data science Peter has founded several organizations that have helped more than 200 hundred organizations leverage data, specific savings, which have been measured at more than 1.5 billion US dollars as latest as anything awesome. And with that, let me turn everything over to Peter to get his presentation started hello and welcome. Hello everybody it's great to be here with you Matt thanks for a great little wind up on this because I think you're right we're entirely complimentary. In our perspectives on this I look forward to welcoming you back here in about 42 minutes, sorry 52 minutes I can count my. Anyway, the key here this is one of those talks that if we were dishonest we would say it's the secret sauce that goes into data governance programs. What we're really going to talk about here is how to identify that subset of the data that does in fact need governing and that of course was critical to that funnel concept that he described to you. What we're going to talk about today then is, first of all, the data has some confounding characteristics that if you fail to take into account will cause you nothing but grief in the future. I'll mention to that then there are four key items keeping your data governance practically focused on strategy, making sure that your data governance program has the required commitment. That was just talked about again a second ago it's got to exist at the programmatic level on this, and then to just simply add ingredients gradually don't dump the entire spice in there all at once but try a little bit and see what works and then add a little bit more as you go forward. Just learn how to do this with storytelling because the more your team knows how to tell stories about what it's attempting to do the easier will be for everybody in the organization to figure out what is going on and I can get behind you because they can see tangible results in action. Let's just jump right in confounding characteristics I type these words key elements of a successful data governance program into one of those AI generators and it came back with this so clearly it doesn't know. It's scarier this was another one that was suggested or even this one so nobody seems to know what these key elements of a successful governance program might consist of and I think part of the reason for that is because data has been approached from a number of different perspectives correctly over the years again just like the blind people coming upon the elephant for the first time from different perspectives, people have entered data from this and this is more of a problem that most of us will admit. It means that people think that data is this and only this doesn't include the other five things that are in the picture or whatever the number is that we finally agree on for these things. There's always been confusion about data responsibility it thinks that data is a business problem. If they can connect to the server my job is done whereas business thinks that it is managing the data adequately after all title is the CIO chief information officer what else with that individual be doing turns out the answers a lot of things. And in fact, so many things the data has fallen into an enormous gap between business and it. It's hard to us to repair that gap as a team and cooperate between us, because the lack of understanding about data leads to people saying well I don't understand what the problem is if we have a flawed data foundation for our project and I immediately put the princess on the P. Thank you Hans Christian Anderson the P is course down here at the bottom of these mattresses and the princess is up at the top here sleepless because of the P and similarly a data flaw will be in there for locking in the imperfections for the application and it restricting future data benefits, decreasing the ability of the organization to leverage around this it accounts for 20 to 40% of it budgets spent migrating converting improving data all the way around here and just simply lack of data and that's what it means, causes everything else to take longer costs more deliver less and present greater risks. Thank you Tom DeMarco for those wonderful words again going back to the concept of the funnel that was presented just a minute ago, separate the chat is a critical component of data governance here and the first thing that you have to convince people is that well organized data is inherently worth more than less organized data. If you have trouble with that one there's a wonderful person named Abbie covert who's done a great job of saying that before the information age occurred we still organized books in this fashion and Abbie has under wonderful series of YouTube videos and more important is how to make sense of any mess. And of course if you took the spine off of Abbie's book took the page numbers off of it and handed it about the yard it wouldn't show us much of anything and this allows us to understand that 80% of our data is redundant, obsolete or trivial and we should only concentrate our efforts on the focus on the very narrow 20% that's there or perhaps even less in some organized organizations on this and who's better qualified to make these decisions of course than the specialists who have been working with them in this area. If we don't have specialists and it means we have people that are completely confused and the data just becomes ephemeral as a result of that adding up to a concept we call data debt and data debt is the idea of getting back to zero so that we have an absolute starting place where there is no data debt so the machines are well honed and ready to go, but likely you're going to need skills in order to undo the data debt about this just to give you an example of how pervasive it is. Back in 2020 American Airlines and United Airlines were valued in an article in Forbes magazine that I've given you the reference there at the bottom of $6 billion and $9 billion respectively but the data was valued. Tens of billions of dollars higher to this. This is something that you better believe that those CEOs would like to unlock the value of that data there because most of them get paid based on worthwhile of the corporation if they could have unlocked double the value in United and maybe triple or even more the value in the case of American here. This is what data debt adds up to and why people are having trouble with this. So you've decided or been told to do data strategy excuse me data governance around this and you've got to get started. One of the first things to realize is that it's not an old profession and it does require these bespoke solutions that are described and others just got to fit your organization so unlike the accounting profession which has been around and selling alcoholic beverages to each other for 8000 years. We've only been at this a couple hundred years or so. Still gleaning value from insights such as oh my goodness if I look at this weaving loom I realize I could do mathematics with this weaving loom and well that's a wonderful concept and thing to think about. It really leaves us in a terrible situation when we look at results we've been asking for years. Are we driving innovation competing on analytics managing data as a business asset creating data different organizations for doing data cultures. The answers you can see here overwhelming it is no no no no no no no no. And the most important question on this particular survey which we're grateful to Randy Dean for recording for us. I have an idea that which is more problematic in your environment so I'm focusing in here on the bar graph. That is a vertical in this case instead of horizontal and you see the number 2018 and then year was 19% 80% 2019 was the 10% 5% I mean the numbers don't change. There are largely problems that are people on process based and the question you have to ask yourself is whom else in the organization is in fact going to address these problems if we don't start on it right away with us focusing on it. So let's look at governance for perspective governance is pretty straightforward most organizations do it it's all about the bottom line. This is sometimes you're seeing some societal implications get into it but all governance generally exists at the corporate level. Similarly, when we look down at it it said we better make sure we're aligned with that and provide something that we can provide measurable value on and I noticed none of them have anything to do with data on this and here's some data governance definitions and I'm so sorry but my friends and colleagues who put these together. They did a great job of them but I don't like any of them and I can't imagine getting on an elevator and trying to explain it to a manager. I'd like to use the term managing data with guidance and the idea is would you want your soul, non depletable non degradable durable strategic asset managed without guidance. I'm not interested in that of course it's typically now and consequently people start to move forward as I'm moving up the food chain in organizations the definition changes just slightly by adding the word decisions so managing data decisions with guidance around this because managers are making data decisions they just don't know that they're making it in fact organizations make lots and lots of bad data decisions on this and this is of course because business makers are not knowledgeable about data as technical decision makers are not data knowledgeable and this leads to bad data decisions these bad data decisions result in poor treatment of organizational assets and poor quality data. This then leads onward to bad organizational outcomes and the cycle is one of these lather rinse and repeat cycles. My sample of Morgan Freeman they're thinking Morgan Freeman yes that absolutely it is wrong. Let me just give you the most recent and egregious example that I've seen over and over again in the past couple of years which is that organizations are moving to Salesforce by an it driven deadline as it goes to a data quality driven deadline, and this has hurt Salesforce immeasurably over the years because Salesforce customers are very much not able to distinguish between Salesforce good software filled with bad data, and Salesforce not functioning correctly and Salesforce gets a bad reputation because it's rushed into production without data quality applied to the process and consequently organizations make bad data decisions about it and have bad experiences with good quality software. In fact, moving to any quality software from a data perspective there's sort of three questions that we should address is the quality of the data in the new system forecast to be a better quality in the data than the old system and if it's not certainly lifting and shifting is not going to improve data quality and that's a motivational question as to why would somebody do that. Number two, are we able to or formulate plans to obtain significant new value and others if we put this transformation and put the data in this new system. Are we going to be able to obtain new value if if we can't plan anything at all we don't have specifics enough around that. Does this give us an opportunity to consolidate data and other data types it's not enough that the software meets the requirements. It's that the software must meet the data governance requirements that are occurring within the organization and as again was stating at the top of the hour. If we govern all this data many organizations look at this and say well we'll start with the A's and we'll get to the Z's right it's just not going to work. Instead the question should be is this governance is giving me this data that I'm looking at right now. Should we include this within the scope of our current data governance practices, and that way you're adding a value component to it and saying is this item valuable enough is it non rot. I'm going to go back to that and doesn't matter what the answer is document the decision why so that organizations can really follow this and get the benefit of understanding the reasoning that occurred at that particular time because after all something may have changed such as organizational strategy. Let's dive in what is strategy. The references to the word strategy occurred before the year 1950 which was the year that management consultants discovered the word strategy and came up with these grand plans and while PowerPoint hadn't been invented 1950 you can still bet there were 100 page strategies on there and that made strategy in the modern era more of a thing a noun on here yes what is the strategy, but I like to go back to the original definition, which is derived from the military using this and their definition is a pattern in a stream of decisions and that instead of being a thing is a process let me give you three quick examples here. I'm not telling you anything you didn't already know because they did a brilliant job of making sure everybody in the world understood that Walmart's business strategy depended on every day low price, being in the minds of their consumers, their suppliers, their you name it people understood this again marvelous job in terms of that Wayne Gretzky the great soccer player again has sorry I talk you what I say soccer had a great strategy skated to where he thinks the puck will be if you're chasing the puck and the puck is faster than you on ice, you will never catch up a lot more at his Wikipedia entry on this topic it's a very interesting example here's number three for the good guys on the left and the bad guys are on the right. We're going to employ one strategy if we're here, we're going to employ a different strategy here if we're up at the top of the hill and the bad guys are down at the bottom. Then we will if the bad guys are at the top of the hill and we're at the bottom so pattern in a stream of decisions should make sense now and none of the hundred page PowerPoint decks or anything else are important relative to that piece. One thing also to note here about the importance of strategy strategy guides work group activities. I know I read the words there and that may not mean anything but you think about it when a work group is looking around to each other as to what to do. They go by what they believe collectively strategy isn't so it's one of the most important leverage points that you have in organizations from a governance perspective. Let me show you a relatively straightforward. Bit of governance proposal here this is something that's prepared to advise the top executive in the organization what you're doing around these disparate pieces it's a government example so it certainly will have those aspects of it but I hope the message that you're taking from this is that this is a complex and to try and start with something this complex. It's got to show substantive value in order to be able to sustain itself in the long run or it will succumb to all cost cutting measures as they inevitably do. Let's take a look on what a data strategy is then data strategy is the highest level of guidance that's available. It focuses on specific articulated business goals and provides guidance when faced when they disissures of decisions or I'm going to have a good provides guidance when faced when they stream with decisions or uncertainties. They most usually articulate how data can support the strategy how it can better support the strategy. And there's a balance of these remediation and proactive activities as we've described as well. The organizational strategy is of course going to guide the data strategy and the data strategy exists to support the organizational strategy both in terms of ideas of data as well as actual data that is provided up there. The data strategy coordinates with the data governance group to say what the data assets do to better support the strategy and the governance group is closer so they will do a better job of implementing that and reporting back in how it relates to data strategy working on this particular approach. Their main lever that they have from a data governance perspective are the data stewards and what's the most effective way of employing those limited available resources who reply with plans, problems, etc, etc. We always want the data strategy to be expressed in concrete measurable business goals, and that the language of data governance should be the language of metadata and the more we repeat that process with the data stewards, the better communication we will have all the way around add to that finally a trusted catalog or starting to use controlled vocabularies in here. And now we're really starting to make some progress. Let's observe how this role occurs over time again. Trustee catalog may start out with a couple of entries literally, but then grow over time and the typical processes there's some new leadership that occurs in there and that they start working on data governance activities. And finally, somebody realizes that the amount of time it's going to take to improve the water quality falling over Niagara Falls depends on how much effort we can put into changing the upstream flows of water. And they don't perceive that often as slow. And so they go we've got to do something that is faster than that and they start things, at least the US Army calls them data improvement projects which they analogize by putting wheels on the sleigh there snail sleigh. Data improves as a result of very specific focus in the so sort of a proactive thing we were talking about before and as you're getting your infrastructure in place with the stewards the community participants etc. In here, we can start to achieve some tangible results in there one of the things we've been good about, but not great is saying when something happens in the data world. There's no relation to that in the organizational world we've been too good about celebrating things happening data wise but not celebrating them enough as we've moved into the organizational side of things where we'd really like things to happen. All of course during this time we've been building up in the trusted catalog in order to do this. There are basically four functions that were previously done part time off the side of somebody's desk that can be incorporated in data function and that involves these four which are really support for organizational strategy, better it outcomes helps from a planning perspective the operational part is being handled fine but the planning parts are the parts that have been challenged, better managed high quality data assets and finally a focus on the ability to improve the productivity of our knowledge workers and I mentioned, these have been typically handled by a group that's kind of gotten confused and mixed settings and it's been coming out so well so we're going to try and concentrate that down into one individual and that one individual will be enabled with the ability to track things down and do system analysis because almost all of the data challenges that are encountered organizationally are filtered through some combination of an IT system or business practice, and that one individual will be able to connect the dots, understanding the commonalities of those core results resolved to a single data element in many cases root cause analysis is absolutely part of data governance and more importantly, if somebody offers you 10 people 10% of the time. You can turn around and trade it in for one full time because only by specializing in these skills and developing this ability to create this repeatable process and to say sustain organizational skill sets. Will the organization be able to attack their challenges in a manner that's both proactive as well as reactive, depending on what they find in all of this. I find that data governance best thought of in an analogy that's very much like a firehouse that you that are familiar with them understand that there's times when you're responding to fires and you grab your coats and jump in the trucks and go fight a fire and save lives around that you never know quite what you're going to happen to run into I love that analogy for a data governance professionals because we never know. Exactly what we're going to run into very much like MacGyver they're diving into things and certainly there's amount of time where you have some downtime at the fire station. But there's also time we're going around and doing data education and fire prevention and changing batteries out of smoke detectors and all sorts of other things in here so some of the section again the governance is important because it costs millions in productivity and siloed efforts not being eliminated poorly thought out hardware and software investments and badly and things into the cloud you're seeing a lot of pushback on that at this point delayed decision making around all this I've already mentioned the 2040% of IT governance data governance must exist at a programmatic level in order to do this and the key first of all is to understand that data is not a project. It has gotten extraordinarily good at the process of creating new capabilities through functions we have some proven methods that are working better than we've been able to in the past. And that's a wonderful thing but as you can see from the top of this data is not a project, it evolves, and it changes. One of the nice things about being as old as I am as I get to go back and visit people I worked with literally 40 years ago at this point. And I can tell you that the vast majority of them are working with the exact same data and data structures that they were working with when they started. So what needs to happen is not something anything new I've called for but many have called for that is to separate next to make external to and precede system development life cycle activities, where the data programs are driving the IT programs because if we don't have that will be in a constant state of attempting to patch our systems to work with our data, and it just doesn't work that way it wasn't ever designed to work that way. And it, we can never expect it to come up with a good result in that same fashion. I'm spelling programs the British way the European way because I do want to distinguish between data programs right and software programs. And here so what's the difference between data governance and data management well first of all governance is going to be at the policy level, setting general directions or guidelines. For example, you might never have had a policy prior to this it says all information not marked public should be considered confidential that may have a tremendous impact on your organization the way it's going on. So keep in mind this firehouse metaphor the idea that there's a variety of different things that need to happen in each one is likely going to be unique in terms of the encounters that go into so what is data management then what's the business function, planning for controlling and delivering the information assets that are required in this case, and delivering data to solve the business challenges around this. And quite frankly, if we haven't messed up sort of the angle we had gotten into in this it wouldn't require so much remediation but we've got so much data debt accumulate now that it really does need to be factored in as part of the understanding that said, most on the outside have no real interest in what we're talking about here on this webinar and I thank you for your attention on this but this is what happens on the outside we've got lots of things that are going on in the data world. So for data management now we're trying to talk a little bit more detailed about data governance. That's what everybody else here. So just call it one thing, it's part of our data program, because our data program is what we need you to focus on and we haven't been doing a data program to the right perspective that like drummer the right way if you want to do that. One of the differences is that we've traded data as a project, because data has been thought in the past to be part of it. It is not part of it, because it doesn't concentrate on the value of the data it concentrates on delivery and piping and infrastructure and aspects. So data really is a program, and what I want you to ask your organizational representatives that are having trouble understanding this concept are a couple of questions one do you think they're going to be more data interactions between your team or less the answer of course is going to be more. And then do you think that there's going to be ever at a time when we're not going to need the data so your data program is going to last, at least as long as your HR program, in order to do this this is an important point to drive home in people's minds that gosh I'm going to have one of these things around for a while. I'd better have a good one and better do it right. So let's dive in and look at how to do from a governance perspective relative to something that's near and dear to my heart the DIMBAC. For those of you that are meeting it for the first time, I'm going to go back outside there. The DIMBAC is a representation of possible practice areas that are inside of data management created by data management. The DIMBAC is a national back in 2009 I believe we've got an updated version on it, and each of these represents in one or more practice areas that we have within there so we might be trying to do something that you can see on the bottom left designated as a combination of governance warehouse and quality. And the reason I bring it out in the structures because most of these things don't work very well on their own that it's really best to approach them as combinations of three, but not seven. Right, it's just a few number but a couple supporting each other, we might for example in the first iteration of the problem, take one experience point by trying to do some data quality, some data warehousing and some data governance in there in order to do that and that leads us to a reformulation of the problem in a second pass out of the second cycle, which is by the way just making the same thing real that happens in all it things anyway we always work on the third version and later. The version of this we've shifted from data quality to metadata in this focus. And again notice we are up to two x points in data governance and warehousing and finally the third version of it one that finally finished the project correctly we went to reference and master data, but we now have three x experience points and three x experience in warehousing in this we've gotten better at those activities and some experience in some other aspects of things. So instead of focusing in on a data governance I like to use something that was created a while ago called a lighthouse metaphor, and that is the idea that there are sort of three things that we can overlap and try and find the intersection of all of them that's really the best the first thing is things that help our organizational strategy whatever we can do to improve data and improve the performance of the organizational strategy be a data means. Absolutely great that's a good universe of things to look at it intersects with data that is used by your business and needs improving so in addition to furthering strategy if we can find something that also helps to improve the quality of some specific measurable identifiable business data items that's great as well. You may need to practice data skills remember I want you to have full time staff doing this kind of work as opposed to 10% of 10 people part time in order to do this when you have the intersection of all of those three places you have a really nice sweet spot that allows you to focus in and work on very very tangible data practices that are able to be focused from a government perspective and understood by the outside as helping in all of these different types of areas that you have in order to look at these. Let's take a look at another thing that we're going to be facing no matter what and that is the idea that we've done. So many of these things in it and what we call an it project or an application centric development perspective which is the idea that we started out and said strategy and then we should add some it to it and then after we add some it data and information or kind of the tail wagging the dog at the other end of it, which leads to some challenges and problems in those areas and there's a wonderful book by our colleague Dave McComb on this. And then if you haven't had a chance to get into it's a real good read for this, but it really is the wrong way to think about organizational and data strategy and that's the idea that it drives through and filtered through an IT piece again Morgan. Thank you, sir. Very much love suppling him in that capacity because he says it so well instead, the right way to do it of course is to have a parallel component for your data strategy and your IT strategy and your governance group is going to have to fight for that doesn't exist already there. Similarly, you can see I'm indicating as well that the data strategy has an out to influence on that overhaul process. So what I've done here is simply flip the idea that we originally started with data. Excuse me so we started with strategy, and then went to it projects and data and information and we just so just have to make that little change now those of you that have been in the business for a while. No that is not a simple, simple change in order to do is a very fundamental change, but nevertheless it is the way in which we need to think about things and governance needs to push forward and make sure that does happen if it's not able to do that it will have trouble, always going forward around that third component is to add ingredients slowly but steadily in here and I just want you to do a little bit of a digital insight with me courtesy our friend of Mark Johnson. We're doodling on a zoom call one day and he doodles and says you know I see that if I subtract data from digital I'm not sure what I get. But I do see that if I subtract digital from data I still have the data left over. And that's because while everybody wants to go digital now nobody really has a clue as to what that actually means and you cannot do it of course I'm talking to a group of data governance professionals here without the idea that you're doing a lot more than just spelling data. So pieces again a wonderful cartoon I wish I knew the origin of it because it's a very apt one shows that in this case some random person in Nebraska has a manual process that would watch the entire House of Cards or the digitization effort. You know is is dependent it requires much more work if we're going to do that, because we all understand the principle of garbage and garbage out and while we do understand that principle. We don't need who do not do not understand the same thing with the folks that are implementing Salesforce and deciding to optimize by a delivery date instead of to optimize when the quality can be of a certain level in there. And because it doesn't matter what you have in the middle this garbage in garbage out model is going to be true, whether the thing in the middle is a warehouse whether it's machine learning algorithm business intelligence machine AI MDM data governance analytics technology, or literally all of them at once I've seen in some organizations. Now the challenge of course around all this is that we have to go through and understand because it's always going to be true that you're going to have poor results in order to do this. Let's start by harmonizing existing data flows and so many organizations spend so much money sending so much stuff. The same stuff back and forth and back and forth if there's money in them. ETL flows that you can get out of it sounds crazy. Now you unblock the good quality data make it run all the way through. And let me focus in for just a minute here on machine learning, because once again governance is going to be the area. You can have some data scientists perhaps in your organization, or not now but soon if not now, and it's going to take them three years to learn your business and to learn what's important about that. You can help expand that trend really fast by, instead of them trying to say I have an algorithm where's this data, you simply say to them, here's this data design the algorithm in order to support this. There's several of the more embarrassing AI oopsies that we've had recently here's one example a newly released chat button people know I like chat bot so hey, they come to me so I ask it the question this is me and the black there so what are you. And that comes back and says I am a chat bot and I say what is a chat bot, and it says a chat bot is a program or an artificial intelligence which conducts a conversation via auditory and textual methods, and I say why would I want to chat with a chat button it says. It looks like I'm stumped, please submit a ticket below and we'll get back to you at our earliest convenience this is such a perfect metaphor for unfortunately, how so many AI resources are prematurely deployed, and can be better focused in the organization because the thing the organization is going to depend on is something I call the data sandwich, they're leveraging high performance automation in one degree or another, and that high performance automation depends on a combination of literacy, a combination of data supply, and data standards being applied within and across the organizational ecosystem because all three of these things have to work together in order to get the organization to perform at the level that it has with the error rate that is acceptable, something perhaps avoiding Southwest Airlines debacle that they had across the holiday season this year just toss in one of the examples that many of us experienced around that. Notice that the statement that popped up on top of that one those three pieces those three architectural blocks merged so seamlessly together and it says that this cannot happen without investments in engineering and architecture and I found that I'm going quote on a cash register that this English, excuse me if this Indian tea farm that I was on the denture seeing out there and it was just a wonderful thing I appreciated it but appreciated finding it. And that particular junction the better and it should make sense that these quality engineering and architecture work work products do not in fact cannot happen accidentally and if we insert the word data into those sentences as we're in data governance here, we have to make sure that the governance folks know how to understand incorporate and push for advocate to these concepts as well. So what I'm going to talk about is a little bit about frameworks and for those of you that haven't seen this picture before that is literally the location I am speaking to you now in fact about the weather that is outside. At this point in time to that of course is not worth looking at the thing you're seeing in the front there is a barn foundation and what's called a horse husband for those of you that don't know. What on earth does it have to do with taking a picture of a barn foundation and data governance well, first of all, consider the circumstances of the photograph. I had taken out a bank loan for the barn construction in there I'm not a constructor myself, although my wife did design the structure, and the bank gave us exactly this much money to do before further construction must proceed. And it was to be a foundation inspection by the county that they would submit to the bank, and then and only then would additional funding be produced well. That makes good business sense does it not knowing anything about horse husbands, we know that the vet bills are going to come first and if I built a poor quality barn on top of a good foundation it would be okay but if I built a good quality barn on a poor foundation I would have injured horses and I would be paying their bills, and then I would pay the bank secondarily and that's not good business sense but there is no it equivalent for this and we need to put it in place. And governance is one of those structures that fits into here so these governance frameworks are really a way of guiding analysis organizing project data, mainly think of it as trying on ways in which you might use these to get some guidance from here for example don't put up the walls until the foundation inspection is passed because there is no point in it. Or once you do get the walls up with the roof on as soon as possible so that we can have inclement weather protection, given a physical building structure, these are a number of these. So if you're on my website as well you can surely take a look there, but this is the one from Dama which is the input process output thing and I'm not going to go through and there's a lot of words here I'm certainly not going to read it these are mainly for you all remember you get the slide so that you can download them and use them at your leisure. Here's Gwen Thomas's data Governance Institute and Gwen is back on the markets good to have you back out Gwen. And, and see at the EGI Q as Shannon mentioned we were all there last week and had a wonderful time with a sold out event. Bob Sinner's Kik construction again, sorry construction, construction consulting sorry Bob marvelous job in terms of this lots and lots of areas these are old diagrams that he said I'm sure he's upgraded since then. IBM had some offerings in this area the point is try these things here's one that I like that had a lot of robustness Jill Deshaix from SAS, her name had put this one together in order to do this. Some people are very clever with these things don't worry we're not going to dive in too far but the American College of Personnel Association decided there was a boat. Right well. Let's let's hold off and look at the way I look at it from a framework perspective and then try those to see how they look at yours. In order to come up with this. First of all, it clearly has to play a role we can't exist without them they play the same foundational role that they always played for everything that happens in our organization. And of course I use for quadrant consulting diagrams and the left hand side here on the left hand side of the orange line, the domain expertise of these individuals who are on that side is less and the roles are more formally defined. On the right hand side the opposite is true the domain expertise is greater, and the roles are less formally defined I said it was a four quadrant, so we have the origin, or excuse me orange piece, and the brown piece the brown one shows that the things on the top half encounter governed data more directly, and that more time of theirs is dedicated to the topic of data governance, as opposed to the top half which the encounter the data that's governed less directly, and less time is dedicated to those individuals and he's got the four quadrants that are there you can see I've already populated it in with the four components we've got leadership which are the data decision makers the stewards, they are the data trustees, the others data makers and consumers all of them are the data workers so they all play a role, and we have some specific subject matter data experts that occur within the framework in here and they're heavily utilized. Most organizations draw a yellow circle around the left hand side of this diagram, but I've seen other organizations do it other ways remember there's not a wrong way. It gets things done, so that you can use data to apply more effectively to the process of the organization, achieving its strategy. So left hand side of this diagram is being designated the data governance group for this example, only there are other options. The leadership is responsible for acquiring resources, listening to feedback, and understanding it, making decisions that then we hand to the stewards to make the decisions they make an action plan, and require changes, and things that go into the others. And there's again data feedback ideas stuff that comes into it guidance and gets us started in order to move. And you don't notice I've taken the quadrant diagram off this is the one I would use to show with people if you need to share them in there but again you can get the idea of how that works and more importantly you can talk about how things would work you can desk check it program it, try it out kick the tires that's what frameworks are for see what they work and and and what works for your organization, your style your communication, your solutions to helping get data governance moving in a faster direction. Get started with data governance the thing I found is the hardest for organizations is simply the getting started part. It can seem like a daunting thing to have to put together all of this stuff and look at it and the next couple of slides once again our reference material for you. It's not required for you to to absorb on here and I certainly don't intend to because I'm going to downplay the importance of them so use them as checklists from that perspective so once again. Goals and principles that you may have in order to do this one of the types of deliverables see whether these work for you. What are the responsibilities here's a master list for you to take a subset from a scorecard. Things that are important. Yes, I've seen a lot of organizations get graded on the numbers decisions that they make but it's probably better to tie those decisions to business value. And that is a checklist for you that you can use to put together and the reason I'm going so fast through all of these is because I look at them as a set of components that need to be addressed lightly. I don't mean that in a bad sense, but it is important to understand that you're going to do these things exactly one time in order to get started and yes it's important to get a process going. I would also suggest that your organization should be open to the idea that this is going to evolve over time. Again, I tell organizations when they appoint their first round of data stewards, make sure that they advertise it as appointing the first round of data stewards, because so many people get left out of those first rounds and if you appoint the first round. They will expect there to be a second round that will populate along with that so yes these things are important but spend some time. Again, I like to use the concept here that it's going to occur once and if you've got a process to get better at something, get better at the thing that's on the right hand side there if you will, because executing the plan evaluating results and utilizing it and applying change management should sound awfully familiar. If it doesn't it is the plan to check act model from again the Deming quality cycles that go through this so I find that most methods, people like to call them fancy things like methodologies a crew up to this sort of a plan to check act thing and almost anything that can work within there. So, don't be paralyzed by the left hand side of this diagram and saying oh my goodness how am I going to get through this and and all of these things must be perfect no instead they're going to evolve. And it helps when you have some guidance that you can get along the way in order to, to look at that that would be great but nevertheless, the part that's the important part is, what am I doing to achieve results. I have lots of organizations that have got data governance charters. I have many fewer organizations that have achieved sustainable business value from the process of getting better at governing their data, then they can in order to create the most perfect starting point that they possibly can come up with. So, let's take this now to the last chunk of this which is storytelling. It's critical that everybody in your organization be aware with and be quite frankly tired of and able to finish your sentences on your data governance stories. The reason it's important is because you want them to get to the point where they say, you're not going to tell me the chocolate story again Peter are you. And I say, yes, in fact I was just getting ready to do so but since you've recall it very well why don't you tell it yourself or if everybody's got it we just say we've all got the chocolate story and let's move forward it becomes a part of the organization's culture so that they understand this and in the context of the chocolate story. Absolutely that is one that is understood. So here's a great one to start with many organizations got heavily into the idea that digital also involves some aspect of cyber currency and they're or have keys or something along the lines. The story is quite fun this was Jack Dorsey's first tweet in 2006 on March 30 for 21. He tweeted just setting up my Twitter and abbreviated it in a format that somebody who me like 64 years old can't quite figure out what that word is but I guess it's Twitter. So, yeah, Jack Dorsey's tweet and it came by the way it was sold for $24 million when it was originally sold. Unfortunately, I'm sorry to know it sold for $3 million and the guy who said it was going to who's going to buy it seen a stop I was going to sell it and he thought it would get him $48 million and instead it got him $2280. So that is an astounding amount of oopsie to be playing with in there and while he was going to have to charity decided not to tell it in that context. So many organizations are discovering that the volatility that's associated with the crypto markets. It is just an unacceptable risk around there from a business perspective and this simply illustrates one aspect of it if your organization has direct focus in there use the direct focus. Here's another one is a wonderful healthcare company that worked with, and they kept telling us that, wow, our getting data from our organization is just like that Catherine Zeta Jones movie, where she has to get through all those lasers and some of you may remember she Sean Connery were involved in a robbery blah blah blah wonderful film but also wonderful evocative language and when management found out about this instead of wow, why would we want to make it difficult for our knowledge workers to use data as part of their knowledge. Again, very important use of this very important aspect from a data governance perspective quite useful to involve and show that there's a complete understanding and leveraging ability around the organizational culture to do that. Charlie's bank has a well determined set of governance articles around their spreadsheet. Here's why when they were buying Lehman Brothers they were having their arm twisted into buying Lehman Brothers is what it was. There was a final contract and of this thousand plus row spreadsheet there were 179 that were tried to eliminate they said we're not going to buy those no matter what you bring to us because they're so bad in there and they handed the spreadsheet to a first year associate who went home after midnight and reformatted the spreadsheet and unfortunately, unhid all the rows that were hidden, which were the 179 that they didn't want to buy but the judge said, too bad, you get them all anyway the sale is now closed. Again, governance around spreadsheets. Absolutely. One last spreadsheet example on here and that is back to coronavirus while we don't necessarily like to talk about it certainly seems like it was an avoidable self inflicted wound to equip our health care professionals with a database technology that would literally drop the rows of data and without any warning to the user that they were doing it because they were using a dot xls file instead of a deck xls x file type. Again, what a terrible thing what something unfortunately the data governance is in fact going to get tarred with if they're not careful from that perspective one last component on this. This is an example of taking engine manufacturer here. And they had in general the old days one piece of censored on there and it would sit into the fan and it can come up with probabilistic formula maintenance forecasts say you know we need to maintain this engine every 16 takeoffs or 10,000 miles or whatever the number of, you know, measurables that were involved in it but now the idea that we can put sensors literally around the engine and they can continuously send back to the organization, 9 million data points per minute. It is an astounding amount. They know so much about these engines that there is, as you can imagine a superb safety record around about this. Certainly in addition to having a better safety record the manufacturer was also able to reduce storage costs, handling opportunity, making a total savings of a billion and a half for this organization, in order to do this data governance process with them for the year. We're going to get to our overview and takeaways here as we do again data has some very confounding characteristics which means you have a very uneven understanding and if you're trying to explain it to people in an elevator. You shouldn't be surprised that people have fractured views and that they are unfamiliar with the increasing organizational data debt that occurs, but by keeping data governance practically focused on strategy. Every day low price is a really easy target when you're in there. It's recognition that this is a young profession and we're going to support the strategy by improving the data and its use in short term and long term and that means we can't just improve the data for people we also have to improve the way in which people use that data. Data governance must exist at the same level as HR we need to develop specialist effectiveness. Data management and data management are both centralization to excuse me central to digitization efforts around that. And again, decoupled entirely from it strategy finally gradually add the ingredients with the idea that we're trying to enable this high speed implementation employee frameworks to refine focus around that, but keep an idea in mind that this is a exercise of plan, do, check, act, and that sense here and learn to get better at storytelling it's simply a good idea that you regularly means to take a minute or and practice telling each other data stories. There's no harm in it and it's a really good exercise and you'd be amazed at how many people discover something as a result of something that is actions like that. So we've talked a fair amount here I'll give you a minute to think about some takeaways and Q&A and say that the date for data governance is increasing because of the increase in volume and practice improvement needs to get better as well. It's a new discipline it's got to conform to constraints there is no one best way and in fact I think it was said at the top of the hour by Matt. There is absolutely if they tell you it's a cookie cutter thing walk away and I like that that concept there. It's got to be driven by these four elements that I've talked about keeping focused on strategy programmatic instead of project specific focus gradually adding ingredients and learning the value of storytelling to improve our data governance initiatives. I'm a favorite book of mine by my colleague John lightly a couple of reference for you all and we jump in back here to the Q&A part. Hey Shannon went over today didn't I. Oh, you know your perfect timing as always Peter it's really so impressive that you time it so well. If you have questions. If you have questions for Peter feel free to submit them in the Q&A portion of Matt. Yeah, I have questions for Peter and Matt so I'm sending them in the Q&A portion of your screen and just to answer the most commonly asked questions just a reminder I will send a follow up email by and a day Thursday for this webinar with links to the slides and links to the recording of the session and our many reiterations of pictures there. So diving in here. So Peter and Matt so some global companies have their own data governance framework. If someone has a DM box framework, how to proceed a cardling in the company without being a rebel. Well, I think that the questioner has answered the question themselves and that maybe you want to jump in here as well but the, the idea of trying to oppose, you know, any sort of thing that's for the organization is clearly a bad careers over kind of maneuver in there. It's not prescriptive in its nature the Denmark talks about what needs to be done but it really doesn't talk about how in that sense Matt again I'll toss it to you on that so that a good good spot for you. No, I completely agree with you. The whole thing about the framework is it to me it really doesn't matter which one it is as long as you have one you can use it consistently. There you go. As long as you are being disciplined in that. You could have one that my kids did when they were five. I does that part doesn't really matter to me. Let's actually tease that a touch further back because it's 100% alignment with what I'm trying to get across here as well. I may be familiar with something called the trolley problem and there's a whole beam on the internet about the trolley problem including someone who's two year old solving it and AI solving it as well. But the idea is that you've got a chance of sacrificing one to say five and just at a very highly simplistic level. If the organization says always do that thing you're going to have people that have a predilection towards doing that. What you want to develop here is that same kind of an ingrained focus here to say, let's do something from the data perspective because it has been ignored for so long and the idea of going against your organization is not what you want to do if they're moving use the demarc to complement what they're doing out there and I think you'll find it's quite easy to do that because the demarc as I said is not prescriptive matter your services group gets into that business as well does it not. Yes we do and we, we basically come up with a decision tree, and it's based on business impact and more importantly the timing of the business impact will generally drive the governance decision. And so get back to what I said earlier just be consistent to that. And you'll be fine. The where groups start to work things start to get off the rails is when they start doing their own thing that's where you start having people with their own Excel data stores, for example, and when you try to figure out why from maybe from an analytics perspective, two people aren't lining up or aren't agreeing on what the value is. That's usually your problem. Somebody is relying on content that has been managed outside of the framework. And that's what gets you into trouble. I have a question in there and I appreciate the into in little team with which the question was asked I think it was need. So, can you either of you give examples of storytelling. One of the things I gave at the end there were stories I mean I want you guys to be aware of the cryptocurrency story at this point and the, the ideas around that Matt do you keep a focus on that or is that my own imagination is no no Absolutely we do because and as data people we get accused of this in the past and it's baiting around them and glacier. We fire out words that other data experts know, inherently, and we lose half of our user group because they haven't figured out what the word means. We talk about one of the best stories we use is when you get a contractor in. You don't really care that they're using this type of saw or that type saw or this drill bit or that drill bit. You just want the closet remote. That's what you're after. So the storytelling is for us. There's an analogy that everybody understands at a fourth grade level and getting away from some of the terminology we get rightfully accused of using that adds to the confusion rather than clears it up. I want to do with you more the idea of course, you know complex concepts like normalization are important to understand and that they are addressed, but that's not what you want to talk about at the boardroom table and we'll get you tossed out of there and your ear real quickly. Great question. Thanks for it. So what are the differences and what the two roles represent an enterprise where data is broken down by domains. Can we associate the data domain manager with a data owner and a subject matter expert with a data steward. I'm going to go out on a limb and say I don't like the idea of anybody owning anything around data. I like the data to be owned by the organization and if somebody insists on owning some aspect of it. I would give them ownership of the requirements of the data while it is in that version of the credit matrix that their domain allows them to work with them. So that's a dodge on that question and mainly to provoke Matt there to see what he's got to say honestly. It's a dodge but it's not completely inaccurate, I would say. We define the data owner is the person who defines what good data looks like to make it try to make it, you know, going back to the whole fourth grade level, what, what data needs to look like to do what they need to do. We don't really subscribe to but this happens especially with like really cross functional data components so I'll give you an example the SAP material master and some instances that you don't have, you don't really have an owner but you have a lot of people who care. And so, to get to an owner you want like this one throat to choke kind of a theory. And it's really tough to get to with the really cross functional objects. That one over from an ownership perspective always gets a little tricky and because in some organizations engineering says we'll take it and some manufacturing will say yeah we want it and other times it's supply chain. So, the hard part with those cross functional objects is is sometimes it may be easier on the organization with it. And I don't subscribe to this but sometimes it's easier for the organization to get past it if the governance group actually handles the ownership and is required to go deal with all the shareholders appropriately. You guys tell him right. Yeah, kind of. So, again, good, good third date so far that is great so yeah the idea of ownership is really problematic and one of the tools that will help you all out with that is something that's looting to but it's just a racy chart, showing that there would be in fact one decision maker over that and having multiple decision makers is really problematic. But the same time getting to that codification can be the challenge for people. And I've just found in my own practice over the years that allowing any group inside to quote own just the data is, you know, even for certain product product or domain areas is really the most problematic aspect of it because it incorporates the worst behavior. So you want things that complement the organization's culture that will do this and so absolutely somebody insists on owning something show them that what they need to do is understand where the data comes from what happens to it while it's under their care and where it goes from there, and that while it's under their care they can absolutely own the requirements which is a long fancy way of saying exactly what Matt said. With from that perspective, you know they're they play a really key role because they're determining viability for the organization so that's usually enough right there and they have enough responsibilities they like yeah okay somebody else. I mean think about who would own the data of accounting. They're coming from everywhere else in the organization so they're, they're interesting roles within all that and as I said it's a relatively new discipline. One of the things we can look towards though. This is something that most people don't consider from a data governance perspective is that it's been the law since 2018 in the US federal government to have what they're calling best data in the US and everybody's interpreting correctly I believe as the DIMBOK in terms of the high level guidance that I mentioned before it's again not a prescriptive document. That's where you're going to need some more specialized, but they've been practicing this area and therefore five years at this point and are achieving some very very nice successes so there are some lots of parts of the federal government that are good to each other parts of the federal government, how to do this, and that we are seeing in fact a very good microcosm and no not everybody is the federal government but gosh, some of the practices are truly universal and you would want to be different just because the government doesn't that way. Maybe that's another topic Shannon we dive into and do a federal government one. And especially if we go back to DGIQ East in December. Sorry, I'm babbling now Matt, would I add anything to that question. I think you're fine. For sure. I love it. And the call for presentations is open right now for DC, if you all want to get in on that. So I sold out last year so yeah, don't slip. Yeah, yeah, exactly. So, do you think there's any management difference between text and numerical data. Or what we call tabular and non tabular or what everybody else called structured and unstructured right man. Yeah. At its basic level. No, every, every data component and I use that word carefully stands on its own merit. And what that data drives should drive the governance decision. Whether you're going to do something with it on ingestion after the fact, or not do anything at all and let a group of experts deal with it, you know there's certain. There's certain financial components that because it's so centrally maintained that you can handle it via a work instruction is how we say, but yeah I don't believe from a true governance perspective and how you were going. Now some of the. Some of the tactics you use could differ between structured and unstructured, you know what what tools in the toolbox you have to deal with it, but at a base level of how choosing to govern it. I don't believe there's any difference, or there should be good governance principles apply across and I find that to be also true. Real key that most of the unstructured is doing is that they're still trying to turn it into a binary event. And I think one of the questions to ask is, is that in fact appropriate this is where quantum computing make it very interesting and things that are 20 years from us at this point that probably more everybody to tears if we started talking about that map but the idea of course of, you know, looking through a video tape and trying to identify a break into a truck, which is one of the projects I've worked on before. The whole goal for that for looking through that was to try and determine was the truck being broken into or not. I know that's not very interesting, but it turns out to be a really challenging AI type activity in there and so working with that data was you benefiting from the decision tree that Matt was talking about before and it was, you know, it was. The measure of breaking this was a confidence measure rather than an absolute measure so it was in binary around that but it still ended up being you know if five of the 10 conditions are met we throw the switch and you know call the guard over to double check on what's going on in this particular sector. So, it gets there yes, it's a matter of tempering expectations as a matter of I would add one thing to Matt I get hold to jump in here but what we see so often in the AI community is a determination that they can in fact tell everything that's going on. And one of my favorite examples is a Tesla encountering a horse and buggy for the first time if you just be able to find the video out there, what should be happening there is that unstructured data that's coming through at that point in time again wrong term it's non tabular data, but coming through the AI should have said, I don't know what that thing is, and having the category of other, as opposed to it must be a semi trucker it must be a person or it must be, you know, again, different things, really is a crippling architecturally deficient component. You know what goes on and this is where data governance must get involved and apply these ethical frameworks in the same way as we were talking about the data governance frameworks here. That's more of an operational plea that has to do with that but certainly a component anyway, jump jump in No, I think it's funny I think we answered this question in the chat about AI and machine learning and I think there is a place for it in data governance and we've seen, and we've actually developed started develop prototypes to do critical data element identification, or at least propose these elements seem to be really driving what you're doing from in a lot of them first prototypes from an analytic perspective. But in some cases, we've had to like my groups had to get involved to say, this is what you're looking for and why, and I don't think that can ever be conjured up by AI. We're still going to have to point AI in the right place and tell them what it's looking for. It might be worthwhile, for example, to use AI to help do that investigation to determine what are your current but it's not quite exactly as you said it's, it's going to be the same thing where it discovers that wolves are things that look like dogs that have snow in the background and the criteria that it understands, which is of course we know is incorrect in terms of differentiating wolves and dog that are there. Yeah, I do want to double back in and again we could probably do a whole webinar and just your identification of the critical data elements there but I want to emphasize what Matt said there is not all of your data elements. I mean, SAP starts off with 20,000 tables and 200,000 data elements, you better believe that that same number of refining and getting to the critical things occurs in there and that in many SAP systems, it's the same number of couple hundred critical data elements that are what you need to control and govern. And even if you say, if management pushes back on it says no you've got to govern everything right because I've seen that happen. It's silly but it does happen to say good these are phase one and when we're done with phase one will, we'll start working on phase two and three and four and get to it, but if you want to make a difference with your governance initiatives and I'm sorry. You know, we're going to have a recession of some sort here it's just the nature of the business cycle I'm not imminently predicting anything it's just knowing the longer we go without it the more likely one is to occur in there. And these are the initiatives that are going to get cut on the other hand if they look at your data governance efforts and say, yeah, I gave them, you know a million dollars last year to invest in things and they clearly delivered $10 million worth of value from it. So you're not going to get cut when that happens coming up that you'll be able to have that sustainable effort in there so the focusing in on those are your best opportunity to really achieve leverage within the organization. Again Matt just coming over to you if you would add in on that but I just think that's a really key piece of your, your method there. It's all about delivering a business outcome. It's, it's, it's, I can't say it any more simply than that governance for governance sake. Who cares. And I think the, the rules start to get a little different from a governance perspective when you start getting into like self service analytics for groups that want to get into selling their own data product, because now you, you really have to certify a complete data set. Instead of saying the out of this data set these 10 elements are critical, you're going to have to go certify it all. That's the only chance time when someone brings up the thing well we got to govern everything. If that's what you're if what you're doing yourself surface analytics are selling that data. Yes, now you've got to go govern everything because at all you don't know what a consumer of that is going to use and why. So that that's where it becomes a little different, but most of the time, you stick to the CDs first. The other thing that you can do in your face with that self service criteria is to say that we're going to have to branded types of data in our organization. And one is, one is data that is of known quality, and one is of data that is unknown quality and probably you shouldn't mix the two. If you're getting ready for a board level presentation on, you know, future trends of the organization just to give a precise example on that back over to you Shannon we should probably kick this one to death. It's a great conversation and great topic great question. So, you know, I'm moving on here we've got about 10 minutes left. Any advice of a business glossary that helps to simplify our language. We always go ahead. We always start with the role names first. Everybody's starting to use keep calling people. I don't care really about accuracy more consistently throughout the organization everybody know who's doing what. And one of the bigger hurdles we've always had is some groups use the term, or the role name data steward for somebody who's going to be putting in material master records all day and that's not the. I don't believe that's the case. We believe the data steward is a governance professional who ensures that the processes are being adhered to consistently and ensures that data is being. It's it for purpose that those that first hurdle starts to get everybody in line around how we're going to talk to each other who's doing what way. For that, the glossary helps I believe in like a very disparate landscape where group, not groups that I just read the chat question, sorry. Where different the same element is called very different things and creating that connection between the two helps groups figure out where things are coming from especially from a lineage perspective. So Matt I'll jump in then and say first of all to everybody listening here again thanks for these great questions. Matt interpreted the question that was asked as where in our vast array of start potential starting places should we start and his answer was focused on the roles. I interpreted the question differently Matt and I interpreted as what products should one purchase in order to get the capabilities in there. So I'll answer from that perspective. Right. No answers are wrong right. No, no, no, no, no, not wrong at all I never even thought of it that way. So my guidance here is don't buy a product because then you have to ask permission, but instead you can do something that I call the Nokia term bank. I worked for Nokia for four years and watch them go through a very interesting transition of literally transforming their business from purely a European company into a worldwide company that all spoke English and that English was spoken at all things. And they created a literal a term bank that if the group was sitting around and had a conversation and didn't understand a term they would look around and ask what does this term mean if they didn't know they would consult the glossary and if glossary didn't have it in it they would add a submission to just film the glossary but the glossary was simply based on reading a web page so once a week they would reformat a web page that was searchable with a browser so we're using zero technology here that could allow the organization to start defining as as Matt said the roles. If that's the starting point that you want to have and there's a good sense for doing roles because absolutely it's something that everybody understands to a large degree and also helps to build on the existing culture. I think one of the key products is my guidance for starters instead build your own work with the components of having a unified glossary a controlled vocabulary for a bit and then you'll have a whole lot better conversation when you are ready to talk to the vendors about specifics on that so two answers for the price of one there on that. Rules before tools people. Absolutely. All right. No important as racy in data governance and can it be modified based on organizational structure. So, you go first. I don't believe that the concept of racy is is something that you necessarily would be able to improve on or somebody would have done so. In the subsequent years it's such a basic useful tool but does, are you using racy and the whole chart as itself as a way of evolving your understanding of how the organization should govern its data absolutely. I think the tool is a very good way of doing that from a structured perspective. So Matt how did you hear that question because it might be completely different. Now I heard it pretty much the same way and there's two things we, when we work with a group that's starting down the governance journey. And for the for us, those those groups are the most fun. Two things we get them to do pretty much right off the jump is create a charter and do a racy. And it can evolve over time and that's completely fine, but the charter, especially, and the groups that have been successful getting their program off the ground have done the charter, and have gotten by in, and are able to make progress in their in what they're tasked with doing is because new group when a new governance group gets put into an organization. Sometimes they become the dumping ground for work that other groups don't have the skill sets to do and don't want to do. Those tasks. People trying to be good corporate citizens, but those tests sometimes can take them away from what their real purpose is supposed to be. The charter kind of protects them. The racy ensures we're not overburdening the same groups of people, because that also happens for every individual initiative that comes up. So I personally believe they're both critical. And it's a, if we didn't hit it right, reach out to us individually. We're not hard to find. I'm happy to follow up with you. I'm not. We've got three minutes left. So I'm going to slip in as many as I can here. So many great questions coming in. So from a capability standpoint, do you think data governance can take some cues from Six Sigma given the higher, the high context of data in Six Sigma. Once again, the discipline around that is admirable. I was, but again, not that you want to resolve on one data point, but I was introduced to the quality movements in the Department of Defense, which was rightfully trying to implement this disease, but was not really capable at least at the unit that I was at of implementing a discipline Six Sigma piece so I'd say ish and directionally and perhaps maybe a realistic version of it but that's a short answer Matt I'll turn it over to you. The one thing. And I read this question when it came up and triggered something the one thing that you can take from Six Sigma is the continuous improvement. We talk about it is how governance groups, once you fix an error, meaning you've remediated the data, you've done your data cleansing. The continuous improvement is putting something in place from a governor perspective whether that's a data quality report or a data validation check inside a system. And that's the part for me of Six Sigma that coincides very well with the governance program because you're continuing once you do that you're continually improving the program and the organization's data. So I do think there's some parallels there. Any more. See, can they do it. Well we've got about 30 seconds left unfortunately, but I don't think we're going to be able to get another question in but I want to say thank you so much for all of our attendees for being so engaged in everything we do. Again, it was so nice to meet so many people in person last week. Always love that networking capabilities and Peter and Matthews thank you so much for this great presentations for thank you to precisely for sponsoring and helping to make today's webinar happen. I appreciate you joining us in the conversation Matt, it's been a pleasure. And I hope you can all join us next week or next month on the second Tuesday of the month for next Peter's next webinar. Again also to follow up email with links to the slides links to the recording and the additional information requested throughout so by end of day Thursday. So thanks y'all hope you have a great day. Great discussion thanks so much for participating. Thanks Shannon by everybody.