 All right. Hello and welcome. My name is Natalie Raymond. I am the Digital Program Coordinator for Dataversity. We would like to thank you for joining today's Dataversity webinar, Data Stewards Defining and Assigning. It is the latest installment in a monthly series called DataEd Online with Dr. Peter Akin. Just a couple of points to get us started. Due to the large number of people who attend these sessions, you will be muted during the webinar. For questions, we will be collecting them via the Q&A section. If you would like to chat with us or with each other, we certainly encourage you to do so. Just a note, Zoom chat defaults to send to only panelists, but you may absolutely change that to network with everyone. To open the Q&A or the chat panel, you may find those icons in the bottom middle of your screen. To share highlights on your favorite social media platform, use hashtag data ed. As always, we will send a follow-up email to all registrants within two business days containing links to the recording of the session as well as slides and any additional information requested throughout the webinar. Now let me introduce you to our speaker for today, Dr. Peter Akin. Peter Akin PhD is an acknowledged data management authority and associate professor at Virginia Commonwealth University, past president of DAMA 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 in 30 countries, including some of the world's most important. His 12 books are many firsts. Starting before Google, before data was big, and before data science, Peter has founded several organizations that have helped more than 200 organizations leverage data-specific savings that have been measured at more than 1.5 billion US dollars. His latest venture is anything awesome. With that, let me turn everything over to Peter and get today's webinar started. Hello and welcome. And welcome to you, Natalie. I hope the weather is pleasant out where you are. At this point, we've got a nice day in the Lower East Coast. I understand they're getting trapped with the snow up north of us here. But anyway, welcome everybody. It's a pleasure to be here and to talk with you about a subject that I've had a lot of work with and had some very good experiences. Literally, I'm recovering from jet lags from a trip from Zurich and then South Africa in the same week, if you could imagine. Also made it back in time to see Elvis Costello concert on Friday night. So how's that for the week. But anyway, you don't want to hear about that stuff. Unfortunately, though, I've had the title data stewards defining and assigning into one of those AI generators that generated this. I think it's kind of got a little bit key on the defining and assigning and perhaps not really quite understand what data stewards are. Let's see how we're going to attack that over the next hour. First of all, why do we need a role force data stewardship it's not something that we've necessarily had formally in the past, although that's the clue to it it has been done it just hasn't been done. It's been done informally around that. What are the stewards supposed to do that as far as being stewards for for the data of the organization in certain categories, and then the process of assigning stewards is a good place to ease into this whole thing, but an hour from now to the top, and look for your questions and answers, which are always lots and lots of fun around that and again driving forward. Why do we need data stewardship as a role well first of all we'll do some definitions what do we mean by stewardship data stewards and data debt. We'll talk about the world strategy as being intertwined with the word data stewardship because there is no other place in the organization or data strategy to have a bigger role in what's going on in the data picture. And then we'll look at this as an architecture as a general tool for the stewards. If I were in person at this point I'd ask how many are starting their data steward journeys versus restarting it and it turns out to be seems to take about three times for everybody to get to the place where they have a good rolling initiative I don't mean that to be discouraging it's just it is a new field and we're still loading our way through it but a good way to think about it and perhaps even more importantly to communicate about stewardship is the idea of actually musicians now as musicians we want to be singing on the same beat of music. And that's what we want to do if we're going to leverage our existing data and change our existing practices to something that is more supportive of organizational strategy and I'll start up with a quick definition here that I don't really like but I want to sort of address it head on. Absolutely there are people in your organization who are key to this but of course that's bad business practice to be dependent on a single individual and also to count yourself as perhaps more than other members of the team. So, while it's not a great definition here it's a good starting place and then perhaps a place to introduce some misinformation yes data stewards are important and not everybody can and could be a data steward but it is a really key role. So let's take a look how that would go from a definitional perspective. So, steward would be a person employed to manage another property is the easiest definition to come from there a custodian a caretaker steward of the estate in that and that gives us steward so we're okay with that. And that is stewarding is the process of managing or look after. Again, they have in parentheses another property will call it another data. Therefore data steward is managing the data assets on behalf of stakeholders in the best interest of the organization. And the Gilbert gave us a wonderful definition many many moons ago and it's no reason to change it at this point. He has to represent the interest of stakeholders and take an enterprise as opposed to a localized perspective and high time dedicated enough to be accountable and responsible for he in the concept of data stewards is trust the element that we have a belief in the reliability of someone or something again definition from Google there. And lastly introduce the word fiduciary if you're not familiar with it, you have a fiduciary relationship between certain types of individuals in your existence typically it would involve on finance matters, health matters and perhaps legal advice, I would be looking for starters around that so if we look at this in totality we're talking about a well understood role here that is to manage on behalf of somebody else in the interest of the organization. And in this case the data assets not all of them of course but a portion of them that are reasonable in order to take a look at let's step back just a touch. And from a steward perspective one who actively directs from Marion Webster and therefore data direct component is one who actively directs the use of organizational data assets in support of mission objectives now. Unfortunately, this leads us to a couple of things to take a look at this before it's a new concept around this. And this is a Gartner chart that's well done but I think it represents sort of a talented the role of steward is defined as data analytics, the word in here and they've done some color coding if you can see that they data and analytics role so a key function in order to do then there's also something called a citizen data steward which I don't quite understand, because it's defined a business role so that that's confusing to me, and neither of these are must have roles I would respectfully disagree with this. And you can see that there are a lot of components in all this. I will argue that the knowledge possessed by data stewards in a well run organization are the key differentiator between success and failure of governance and therefore data programs all the way around. Let me now move on to a slightly controversial topic as well. So that is the idea of data ownership, very big challenge. If you have it, most organizations they start to say this is my data and it's mine and I need to have it around here. Instead, the terms to do theory relationship or if they insist on owning something, give them ownership of the data requirements within the defined portion of the life cycle that they are looking at. In other words, there are upstream activities of which they do not own those data requirements, and they're downstream activities that probably should be taken into consideration as you're defining the local ones in here again. As a matter of we put our controls the better off will be able to do this if you get real pushback on the concept of data ownership. You can just instead say what data would accounting own, because accounting is a an area that inherits data from other parts of the organization, all the way around This is unfortunately a result of what has been considerable confusion as to data responsibility so imagine having to take your new stewards and their perspective is perhaps that I think data is a business problem if they can connect to the server my job is all taken care of. And the business thinks that it is managing the data adequately after all what else would be the title of chief information officer be, and as a result of course data has fallen into this gap between business and it where we need to reestablish the patterns of cooperation in these areas in order to repair the great amount of data debt that organizations have that aren't quite as visible as that slide would give it to you data debt is things that get in the way think of it as a clogging the veins of the organization. So this progress that decreases quality increases our costs and presents greater risk to the organization. Let's zero in on the role of this to do theory piece for a starter what is the purpose of it is the person who's legally obligated to act in the best interest of the individual company and this includes again as I said before lawyers, but these doctors, other types of activities up to and including for the legal guardian, the real key to this is that there are three duties of the fiduciary relationship and one is to act in good faith as a very clean in that sense, the duty of care that you have a duty to care for the data that you're looking after at this point, and your duty is that the organization as a whole is your loyalty around this area. The fiduciary, they have left the legal standards they must adhere to, and not knowing this, we also end up with the sort of perhaps the biggest business talent that we have around data matters that we build data things. It's often the case of the princess on the key I've just circled in yellow down there, a key at the bottom of a 20. You have the princess up here at the top it's sleepless because of this inherent flaw. Similarly data flaws that are built into new systems have extremely challenging areas that are locked these imperfections in for the life of the application it restricts the potential investments that are going forward and decreases the ability to leverage the organization data, 20 to 40% of it budgets are devoted to maintaining migrating, converting and improving data as they evolve through the process. It takes tremendous amount of savings and it also improves your speed at the same time. Bad data models bad data, he's caused everything else to take longer costs more deliver less and present greater risk thank you sometimes Marco for that wonderful bonus of articulation of an article in Forbes at the height of the pandemic in July of 2020 gave the value of American Airlines at $6 billion and the value of the United Airlines at $9 billion. But the reason they picked this as a contrast is that investors were apparently willing to value the data in there between 20 and 30 billion in the case of American and at least twice what United was worth as an airline if this has been true wouldn't have been more sensible to buy the airline throw away the airline bits and just keep the data. I don't know but I think this represents an underutilization I firmly believe that if the CEOs of these two airlines understood how to unlock the value of that data they would have because they have every incentive in the world to do this. This gets us to the topic of data strategy in the context of seward so strategy again just very briefly is typically done as an organizational strategy and then it has been filtered in many cases in the past through it strategy in here and I simple Morgan Freeman here because he says it so well this is wrong and the reason for that and the way it should be done is that data strategy is distinct and probably more influential on it than it is on strategy around that again not to pick on anything particular but when you think about the types of assets and things we have been doing over the years, one of the fortunate things about being able to be active in this field for 30 years plus is to observe that these organizations have not changed through the course of 30 years and that is a very significant thing they're still managing the same data sets that they were managing in the first place perhaps with different technology for us in different places. But that's a much more solid component. The strategy of a concept wasn't even really introduced to the general public until about 1954 and it represented this master plan after the management consultants got a hold of it and said we're going to put together a thing for you to think differently about this because I believe that the management consultants have hijacked this term and that the proper way to think of it is strategy derived from military use after all this is where it quote has been successful and definition here is a pattern in a stream of decision. So this makes strategy much more of a process and much less of a thing. Let me give you three brief examples Walmart business strategy in the past life was every day low price and they understood this and their customers understood this and their suppliers understood this and their critics and their adherents understood this it was a unifying feature of the organization and a very clear and successful implementation of an organizational strategy, not just in the organization but throughout society in this case on here. An example of a strategy here is Wayne Gretzky, and he basically says I could either chase a hard rubber plastic ball around that roots across tiny ice and very slick ice and try to take it or I can skate to where I think the puck will be raised my hand and say I'm open, and then I will score as I move in on that. And at Waterloo is our third example here and the question was how do I just eat the competition when their forces are bigger than mine and the answer of course is divide and conquer. Now let's look a little bit very closely at this one. Napoleon observed that the British troops were supplied out of a spend and the person troops was supplied out of lead and being a good battle commander he understood that when faced with significant harm and peril, a troop has a tendency to run more towards a supply chain that away from their supply chain on the so he came up with the divide and conquer plan, but which is still starting in US military schools today. The idea of course is to hit him exactly the right place so that the red troops move off in the left direction that they're headed and the black troops move off in that direction, then of course very quickly turn around and conquer. And the idea was of course first of all go to the right and remove the front and then go to the left and remove the British. Again just very very simple strategy but on the other hand let's think about it perhaps it seems simple. Hit both armies really hard at the right, then turn right and defeat the British. Excuse me to write and beat the pressure center left and defeat the British. I know by the way please do this, while somebody is shooting at to the entire time. It makes absolutely good, I think, evaluation to say that this would represent a complex strategy particularly considering that most of the folders didn't want to be there. So what we've spent a bit of time talking about strategy and the idea is that strategy when it comes to the data level would be a simple concept, my friend Don loudly get a quote that I think I'm going to adopt. Thank you very much, John with full attribute and that is that the data strategy could be the chapter of your organizational strategy that deals with data. I like that a lot. Let's look at a particular other type of perhaps complex governance that can imagine being a steward in an environment that was this complex. And I'm not saying this is good or bad or even what it's particularly for but I think you will again all agree that it is complex. I'm going to try and work on this. So, let's go back to our definition of strategy of pattern in this frame of the savings and understand that strategy guides workplace is a work group activities. The idea that within a work group, we understand our common goals and objectives and move forward in a unified passion is a very, very important concept around strategy, and that is what has to be communicated ultimately down to our stewards. We understand governance and architecture again why are we doing this well corporate governance has been around forever, except that we had before the pandemic the idea that perhaps wasn't all about the bottom line and we'll see whether that goes anywhere but it was, nevertheless, a robust start to that effort. Of course, if we have corporate governance we got my IP governance looks at measuring results addressing key areas, again the five that have been identified here are in blue. We have two data governance definitions and these are seven of them we have used over the years, sort of foundational in nature around this. And of course you all know what an elevator pitch is, you get on the elevator and executive steps in and look over this. Oh, tell me what these data stewards are doing and what is the data governance and all this other you got 20 seconds or so to not look like a fool and give you an advertisement for your program of course. Well, what I'd like you to take away from here is perhaps a definition of data governance. That's a little bit different from the ones that I showed you I like to use managing data with guidance. And the reason is that guidance then becomes the predominant term and governance is not the predominant term. And of course also you might ask the question in reverse, would we want our soul non depletable non degrading durable strategic asset managed without guidance. I think not. And again if we look around the organization stewards are the individuals, the job class that is going to make a difference in these areas. And when we're all speaking to upper levels of management we change this definition just slightly with the addition of the word decisions managing data decisions with guidance. Very important in that concept shares well because most managers make data decisions they just don't know that they make data decisions. So let's look at how stewards fit into the composition of a data governance organization in here. We're going to have leadership we're going to have data subjects matter experts. Me as we call them oftentimes we got everybody else and of course we've got our stewards that are really focusing in here as prestige. The organizations will take this sort of left half of the diagram and that's our data governance organization in here and we're going to find the role of acquiring and maintaining resources. The leadership which means there has to be a positive return on investment again who's best does that come out of the stewards right. We've got the feedback and decisions and the stewards are the ones that are charged with guiding leadership through the decision making process, as we're implementing the results of the decision which means there has to be some actions taken we have to do things differently than we used to in the past, again as an ongoing feedback ideas concepts and guidance type of role in all of these things. Now we're going to talk a little bit about architecture and architecture is just about things the function what those things do and how those things interact. You can see here these are two very silly things that are interacting and it gives us the ability to define a controlled or a common vocabulary, and that's the best example for most organizations because they start to come back and say why are we doing this architecture stuff and the answer is, we need to get all the music on the same sheet of paper. This is our common or controlled vocabulary that we're pulling into place so that we stop the siloed towers of Babel in here and understand conceptually as an architectural component. There's connections and the commonalities among the data pieces in a way that's shared by three groups the business users of that data the technical group supporting that and the systems as well this is what we mean when we say the language of data governance and especially data stewards is in fact metadata, because unfortunately most organizations have architectures that are not understood, not documented, and therefore not useful to the organization. Let's now drop into our. Oh, I'm sorry, I want to say data for that as well. Let's drop into our what are we supposed to do. And again here. The idea is to look at challenges that are bogging us down because most people are unaware of the effective data debt, and it just makes everything of this and before faster, excuse me, slower, take longer, deliver less, present greater risk to the organization around that we'll look at a framework for stewardship that is derived from a well established framework and it's a very good way of thinking about this. Then I'm going to relate our activities as stewards to the role of fire station, and we'll talk about that. Then we'll look at stewardship in the context of overall data governance around. Let's dive in. We've already said don't say no reference instead this fiduciary relationship and it's just appalling to do this absence from that and I know it comes up twice in here but it's still fairly important and I just went to the absolute wrong slide. Here we go. Did you get when you type the wrong number into the slide by the way any guesses in the chat how many slides I'll get through today. Around all of that. Really sort of fun to push the limit there we go. So, I'm so sorry. Typically, what happens in organizations is they recognize they've gotten to the point where they have data but they're not able to access the value. And the reason for that is because we're overly dependent on human beings or the wet where that is in between the years the knowledge workers and sort of informal communications is often described as the weakest link in this stewardship is designed to formalize these roles to be available if the organization decide to measure the activities in here which means they can be supported and an organization can then determine an effective ratio of data that could be covered by stewardship around all of this. The first most united purpose of stewards is to help the organization use its data better to support remission. And I mentioned, or at the top of your framework idea which is the idea of guiding analysis organizing the project data, enabling the ability to make priorities, assessing progress, coming up with rules if we're building houses that don't put up the walls until the foundation inspection has been passed. But once you do that put the roof on as soon as possible so we can work in inclement weather, and again make it all dependent on funding sources continued funding. So the framework for stewardship was a wonderful contribution to the intellectual literature here which there is a personal mastery a vision around this, there's a component of mentoring that comes into it obviously diversifying opinion, but gathering the vision in this and a certain aspect of risk taking and experimentation vulnerability as well as maturity on this again making mistakes as a human activity so we would expect that but overall we would be delivering results around this and so with that sort of framework for stewardship I present to you this case a framework for data stewardship which is the idea that there will be organizational data challenges. Sometimes it's as easy as complying with battle three. And I don't mean that lightly at all it's just that has a well articulated set that other types of organizational data challenges may not come as well defined as compliance types of activities. Either way, we have to make a decision when we encounter these of some sort of strategic nature, we're going to put them in the bucket and address them some other time or we're going to address them now with our data stewardship engine and that again, regulation and policy of these stewardship activities would fall underneath it, but always falling into sort of reactive and proactive and we'll talk about the ratios in that that will then yield value and some of that value is monetary some of it is non monetary. The key is of course to get good at the process of the value of the results become increasing up to the point of diminishing returns in which case we have adequately buys our data stewardship organization. I mentioned before the firehouse metaphor around this. And the idea is really of course if you know anything about fire departments, their professionals and volunteers, then part of their time waiting for something exciting to happen so they can go out and help the communities by saving lives in the face of these kinds of threats, but fire departments also do an awful lot of proactive work, educate them in our schools and community centers, where they are helping people to understand things like the biggest cause of residential fires in today's world is an incandescent bulb. That doesn't mean the incandescent bulb has gotten worse. It means that we have made our houses safer to that standard and that is now the greatest threat. It's similar to the risk taking pieces as well. We used to have a TV show called my guide. It was the guy that could fix everything with duct tape. And that's an important consideration for us that we will have novel situations that we're facing and therefore novel approaches that we'll have to take in order to address these I like to reference data and duct tape and where I a songwriter I would have you a song about that but I'll share you that. The last piece of this on the role here is the idea that of course a organization can take a group on as full time or part time. And given that set of choices. It's real easy from my perspective. And I'm back to pulling the driveway alarm. My wife is back in her truck up anyway, if you have a choice of 10 people 10% of the time, or one person, my advice would be to take the one person. And the reason for that is very simple. And if you have the ability to get into this in a more in depth way 10% of 10 anybody's time or even one third of their time is still a difficult series of choices that the organization has to make but if you're full time. This is your job you're professional about it and you're going to try to improve your ability to serve in this role in the organization. It's an easier way of starting to co here that you need to have in order to have your organization start to move forward on this. Now, next chapter around here. I'm not reading you guys a book, but a very well run survey over the years has been run by our colleague Randy Bean, who have all these results at the new vantage website down at the bottom but the question is, he's asked people years after year after year, are you driving the organization with data and most are not are you competing on data and analytics, most are not are you managing data as a business at that most are not are you creating a data driven organization are you forging a data driven culture, again, most are not and while these are not great statistics, there's some real interesting patterns in the data have another slide for you. Yes, the question here over the years, are your problems primarily technology focused or people and process focus and in 2018 you can see it was 8020, it never got back up to that level 2019 5% 10% 7% 8% in 2021. It's clear that the data challenges around our organizational inability to utilize data are people and process based and data governance is the only resource to address these challenges and data stewards are the only people in the data governance organizations that have to do any work. Well, I'm sure I'll get some questions on that but let's bring them on and have a discussion about my just the more I see this is the more I'm convinced that this is the right approach to this. For data, you see manifests itself in a number of ways that your data stewards just naively won't understand, but almost all data problems are filtered through some sort of an IT or business process combination or combination the example I'm showing here is very simple, because I'm only going one still praise and around this. Many times it goes very far downstream before we actually do get the results of the state of being incorrect. And only when we are able to connect the dots all the way around can we understand that these poor results have a unified cause, and that is that the idea within organizations, when you look at trying to do this individually by everybody from the outside trying to fix it at the at the edges, instead of having a dedicated team to focus on this. It's just folly. And no matter what happens we continue to see these things be problematic around here. Again, the idea here is that we need to be able to have the organization be able to get to the critical math in order to do this. Let's not talk about the role of data governance in here. And again this is to put the stewards in context. Most organizations have sort of vague ideas about what's going on in these areas. But of course, eventually the data governance organization is formed and data starts to improve over time. However, the perception oftentimes on the part of management is that this is low, and having to make the progress in quarters, bugs them, much less the fact that the real impacts may not be felt for years in some instances around it. So what happens is that organizations say, yeah, you need to do something a little bit more proactive with some wheels on that snail there but and let's get that rolling out and make the data improve as a result of focus. Now this is active data governance and I don't see a lot of organizations indicating that this is something that they could do but I see a lot of organizations practicing this. And so I think we're beyond the point of this is an introduction this gives us the ability to get better feedback and to start employing a professional class of individuals that will exist here from this point onwards, as our data professionals, one component of our data professional staff in here. And just a quick side note here that while we're pretty good about articulating when data things happen, we're not quite as good or we haven't been quite as good about making sure that those data things that are happened are tied to organizational things that happen here. And so the more we can practice at this particular activity, and make it easier and this is really the only reason I find an organization so they have trouble with this is because they haven't practiced it, which is just in terms of what they're attempting to do so it's very possible to put together very, very big returns on investment on these. Inicutives in here it's not that some unfunded mandate is definitely not the way they should let people think about it but in fact a strategic investment in your ability to use data better in your business let's also then pass to another component here which is that the role of the data stewards is in fact generally larger than we originally described it early on here, and that within that role of the brown box here they are really responsible for making sure an awful lot of these things happen within their areas so the scope of stewardship efforts is quite large in most instances around here. The key to this is to understand that your organization is going to have to take a both proactive and reactive role in addressing these topics and that that proactive and reactive role will also include some things that are not considered part of data that you should ask is, well, who else is going to do them, because if I don't have a group focused on these people and process problems. I will not in fact be able to overcome the good things that data can do just by throwing more data in front of good people and I'll take a quick side deviation here to an idea of a slide that I didn't have a particular slide for, but it's the role data literacy plays in all of this and of course the idea here is that storage need to be more day letter than everybody else, so we can clearly see there's some gradations that we could pay attention to, from that perspective as well. So now you have an idea of what they're supposed to do they're supposed to take data and move it in a way that will increase its ability to support the organization's ability to achieve its strategic objectives in here. We look specifically around several things which is data depth and what sort of a framework for this, whether you have a ability to distinguish between these reactive and proactive folk that you're looking at. And finally how the stewards play a role in all of this and it's a very encompassing role. Think about it just who else is going to do it. So last section of this we're going to talk into assignments of data stewards the idea that you're going to start the process. And, first of all, there should be a very much of a concept here that says these are going to result in tangible improvements. But even when you make those improvements it's important to document those improvements because those improvements, while they will be seen simple for two others. They'll not have a good idea of how much they actually had up. I'll give you just a very brief example, which is that I had an organization at one point that had a query that ran a billion times a day. And the idea of this query running a lot. This gave me the question of I wonder, the person that wrote it did they in fact understand query optimization. And this is not necessarily a topic that stewards should get into but necessarily, they should understand that there are always better ways of doing things and we should be able to identify what they are and evaluate them and respond accordingly. And it turns out that the person who had written the query and not really understood the process of query optimization it was possible to rewrite that query, but that we were able to reduce it by a quarter. Now that's not a lot 25% is a good amount but if I take a quarter of a billion. I'll take that any particular time, as opposed to trying to find out the underlying causes of why we can't reduce costs around that area. I'm going to dive into our last section here now which is again the idea of assigning them and going to give you some advice which is hopefully makes sense. It is that you should start simply, and if you can follow and understand to make sure everybody understands the differences in cadence between IT projects and data governance approaches, which is then talks to the need for a different structural approach. So that's the foundation for requisites that we'd like to get to around this. And then finally, simplicity agility and practice around here. So let's dive in again and just say that the sources of data are many in just a health care environment and this is like if you are looking at the source being an order catalog and the steward for that is probably a member of the IT staff. The point of this is that it's a lot of different definitions and while there are, in fact, a lot of different types of stewards. This represents the thinking of an organization that has matured into this setup because it works for them. This is not the result of everybody tree planning and attempting to guess in advance what's happening here. It's more explicit about it. So while these are excellent definitions of stewardship and knowing that there are multiple roles in this, it's probably not a good idea to introduce all of this at once to your organization. My colleague, David Loshan, has written a wonderful book on data stewards and I highly recommend it to you, but he's describing here a number of different types of stewards and I'm not going to read them to you because it's not the time to start thinking about them from that perspective. Because of course if you get this many different types of data stewards that are in here, then of course we need an auditor and then a manager of some sort in order to put them together. So it's never kill, confusing, and remember that most people are less observant and care less about this than we do. So we have our definition of stewards, one who directs the use of data assets and supportive mission objective. Okay, well, problem is externally this is not well understood. They sort of data stuff that people here and then the data management data governance and what's the difference between those two again and oh my gosh never going to introduce this concept of days stewards in here as well. It's not not possible to most people to do this so keep the comprehension at the programmatic level. What's going on here we have a data program doers are part of that data program. If they want to ask you a follow up question. Great by all means speak with them about it, but it's just so difficult to try to get this even understood within the data groups much less across them in organizations. So I have a whole series of charts that I wish I knew where it was from because I would love to give them credit that just appeared in my inbox at one point that what we're really trying to figure out is what do they do in our organization they need to improve data value and help the organization to achieve that, but they should also be the primary evangelical first inside the organization to increase the scope and regular data center practices and ensure efficient and effective data management practices unfortunately these last couple are not typically included in there. It's just a, well it's a wonderful idea it's not been practical up to this point for most organizations to do this let me take you to the concept of what I call an organizational data machine now this is the external perception of the organization by outsiders whether they're business partners or customers or regulators, all of your input their data and all of your output their data and so everything that goes in is somehow transformed added value and the question is, if we add too many controls and standards into the process. That becomes expensive and low. And if we do too little miss opportunities. So the answer is clearly in between these two extremes, the people who are best able to judge where that balance should be and more importantly, where you should start in your journey along these lines, so that you can be most effective around this they'll understand inherently the dependencies are your data stewards on this. And let's be right at this point, you're probably in a situation like many organizations of not knowing as much about the data that you'd like to will say it's a known quality and 20% of it and unknown is 40% of the 80% around here. Well, similarly, over time we'd like to improve what is known and how it's known around these topics in here, and eventually we'd like to get to the point where we realize we don't understand anymore and that in fact that one fifth of the data we might be able to dispose of entirely after appropriate vetting, which could result in a tangible reduction in our cloud storage bill, which is a very nice way of starting to decide this in terms that others will understand the key determinant of all of this of course is interoperability. And that is what gives us the value in this particular process. Now some of you may have encountered the term systems thinking before this is a wonderful quote from fruit health Capra. Sorry, I want you to the next page that one more time. Oh, there we go. We got that. And let's just hear on my part. Well, we'll keep it up here for a second. Hang on. We go back and put it up because this is important stuff. Yeah, is that the only way to fully understand what's going on from this series of perspectives is to understand the relationships as well as context with everything else. So again the forest for the trees, or to understand that relation in part of the whole, and we're looking for system thinkers in data stewardship, so that we can have these individuals take a idea. from upstream to downstream in this logical progression of value addition. Now I mentioned before the organizational strategy provides context for the data strategy and ideally the data strategy is that chapter in the organizational strategy that deals with data in there. I hope that to have that that's wonderful. If not, again, reconsider in terms of how much effort you're putting into the actual document, and then balance that out by the ability to practice what you're doing more around a series of strategic cycles. The stewards are the implementers, and they are what we are implementing so data strategy must be expressed in tangible concrete business goals that everybody else understands within that, and that the language as I said before of data governance is metadata and therefore to appear highly in the stewards area they're going to be working in these areas in a way that it's going to allow them to continue to add value from a very comprehensive perspective remember in the Peter's world the data stewards are full time even if you have just one it's still a great place to start, because that individual will be so visible that they will have no choice to succeed massively. So let's just component off by coming back and just making sure we've got feedback loops within the entire process. Now, looking at this and saying the role of stewards they look like they're at the bottom of the sheet thing but actually you need to invert this I could probably make a diagram of this at some point that the stewards are clearly at the top of this framework because they have all of that apparent knowledge, and they are tasked with making sure this is documented now that's to say that you want to support your stewards with appropriate resources if they are working in a case tool environment that's terrific if they happen to subscribe for meetings and things like that whatever it is they need to do so they can focus on their goal which is to take the value of the organizational data and make it even better than it currently is. So I would like to point on this, which is kind of related to scope activities. Again, most organizations they hand the stewards and say okay you go start with a new do the bees new do the feast. So this is the wrong question. Instead, what we should ask is, is this data item within the scope of our stewardship practices, and the answer may be next year, or five years from now or never or absolutely. But those are very different from trying to do everything. Let's get to a couple of dirty, specific secrets, 80% of the data that you having your organization is rock. This is the nature of data debt, you have multiple sources, if your organization only has customer data in less than 13 systems, you are above average, our standards are absolutely low, given this context. So, rather than how are we going to manage it all, let's instead find some things that if we manage them would make a direct impact on the organization, or an easy song like top six, but let's play that song and learn how to do it together around. Finally, document either way, because that documentation has to be the corporate record of what's going on in the data world I've already told you the story of the billions of queries that run in these instances. And that's because what we're doing here requires what I call a highly tuned data sandwich. The idea is that some components of literacy perhaps uneven supplies of data and standards, application of the organization are kind of clunky and causing stand in the gear. And of course is for customers to have seamless experiences and the only way we're going to do that is by improving literacy supply and standards and once again stewards are the source of that expertise. This is what they have to learn how to do to get these pieces to work well together, and this can't happen without investments in engineering and architecture. This is all the way to India pandemic to see on the tax register of the tea farm that I was visiting the wonderful demo and quote quality engineering architecture work products do not happen accidentally. And of course the same thing is true for data pieces of it as well. Because of course data is not a project it's a durable asset. It has a life that lasts a long time on this more useful than one year. I'm going to get into some of these topics. Again, probably not a direct theory topics that certainly of interest is the monetizing components that are apparent in some types of data. And Doug Laney has done a wonderful job on his infonomics book of describing some of those components. And here the cadence that was involved in this. So certainly we invest in our customers and hope that their data will have an asset value of greater than one year if you think back to the example from the Forbes article on the two airlines. Clearly, at least valued by some investors as being more values for their data that they had, then in fact the operation of the airline I don't know about you but if I could get just the data and let somebody else run the airline. It's a tough job I'm glad some people do a great job of it, but it's certainly not anything that I have any expertise at on this. We're talking about the cadence of this and that while it may be reasonable to do two weeks front score, you know, increments around the day instead what we need to think of is that data evolution is going to take much longer because data has to flow through the systems. And so the evolution can be measured in as long as years, because data evolves it's typically not created and that's an important thing to consider because many people think of data stewardship and data governance as project. It is a project that it has a start and a beginning and objective start and an objective and the data of course has neither and consequently because it evolved. It is more stable but it is also not suitable for project management, and it is requisite on the organization to put in place a program you will no longer need your data program. I'm not on the right side there. You will no longer need your data program there we go. When you're no longer need your HR program, right or your finance program as you're thinking about it. And that's a really good way to think about it and the arguments are straightforward. Do you think there will be more or less data in the future. I will say more and do you think that we will have more or less expertise in the future. We'd like to have more expertise but on the other hand what stuff are we taking to develop that expertise. And this is where the investment in stewardship is probably an even stronger indicator of potential success than the existence or nonexistence of a data governance program in and of itself. Back to the slides just bring two points here that I jumped ahead to and I apologize for being a little confusing but that's the fun parts about a webinar right. Yes, your organization is going to be creating agile topics and agile is the best way we have come up with of improving the quality and the speed with which we create good component software on this. Data requirements are a prerequisite to agile developments and if you recall a couple slides back I had to all thinking that not we don't want to own data ownership we don't want them to own it but we can own the requirements and those requirements. Play a direct role here if you don't have those data requirements well objectively understood at the start of your agile print, you need to print off in a different direction, because the only possible results and this is to create additional data files again. The part of data debt is for years and years we have taught all it students computer science computer engineering and information systems that the answer to every data problem is to build a new database, and therefore we're surprised when we have these these those that have been bringing up all around us and it's the need for a program in order to do these things. In fact, the way to think about your organization is perhaps changing as well. In most organizations, there are plans that are at the top, and those become data on their transmitted down through the ranks so that others then get some instructions and in some cases now. You actually manufacture things. Of course, if we know anything about the various manufacturing we're manufacturing components and then maybe assembling them today again it's going to be different for every organization but increasingly, you're going to have more of a data focus. Your organization is all about data until it's not about data. And that's an important consideration to think it changes the mindset around this I was talking to a logistics group on South Africa the other day and then they had good expertise in these things that they also were good at understanding the marketplace and we talked about them prep providing advisory services into the rest of the marketplace a question to ask is what business. Are you in and it's a very good question to take a look at. Let's also say that stewards have an important role to play here as well, which is the organizational data practices. Lots of buzz going around the chief analytics officer and all of these and I have nothing against them, but I will say that the data is a necessary and insufficient prerequisite to good analytical practices around this. And that, while we have lots of models of organizations doing really cool things taking it into March and then data and then flopping it back out on to a series of mini March and accesses and different types of dashboards and all the rest of this is fine. We learn a lesson here when we discovered data is wrong because we're looking at it in the dashboard here, the feedback and learning tends to be focused within the right hand side of this diagram. Whereas I hope you clearly see that an investment back into that black box that we call data management would be a more effective way of solving this in the long run stewards are the only people who are going to be able to come up with the usage. And that's around this that talks about how long it's meagre and what we should do with it. Another area here that steward focus is will agree that along these two dimensions here we can either improve operations or innovate there are no exceptions in between. And then if we put folks in there without formalized steward programs, we also agree that's wrong and hopefully if you get nothing else from this hour that you're spending with us today. And I hope that you agree that Walmart has good reputation a well earned represent reputation for being able to increase the efficiency and effectiveness of its operations and that Apple perhaps has the innovation buzz around all of those. I want to imagine taking Johnny I was individual who used to do the Apple product rolls out roll out and he would, you know, just have his head explode if you drop them into Walmart's environment to do it cheaply. Right, that's not the way he does his creation. Similarly, I imagine if you took the Walmart people down there at the bottom right and drop them into the apple quadrant and said, I want you to be creative that you like no no no we optimize that's what we do we it's in our DNA. We're good at that so there's a sequence in here that we want stewards to look at as well. Many organizations they do both at the same time. I would also argue unless you're really special and a lot of organizations and they are but it really are. Take down here and earn money by helping improve things again my total is more than $1.5 billion US that I have helped organization save over 30 plus years. And so use that to then create the conditions where you can do this more correctly because only about one in 10 organizations are in fact working in this area starting the stewardship journey at all and that's very sad we'd like more of them to get involved, but we also need to prove ourselves better as an industry. So let's recap a bit and we'll do some takeaways while you're getting ready perhaps for some questions on this. Why is it that we need a data stewardship role well we have strategy that we want to do, and we have data that's been employed to do it but there's nobody who is specifically employed in our organizations to take data and make it better. We have had lots of people doing it in the past, sometimes informally sometimes a little bit more formally where you could call all of these topics data administration, but we need to modernize around this. So, the idea is what is the organizational strategy can we employ data in support of it, so that our targets are easier to achieve, etc, etc. And within that, our architecture is the strategic tool that we're going to be using we're going to try and improve the architecture around this through active as well as through types of measures. And what are the stewards supposed to do well, there are going to be some challenges resolving from data depth and that's unfortunate, but if you ignore them. I don't want to paint a bad picture in your mouth but imagine trying to go in and work on somebody's cavity, but that person didn't discover brushing teeth at that point so they had a very bad sort of dentures. You need some sort of framework for stewardship it is a professional role. It's well defined well under good. Although most organizations don't consider that they need to do reactive as well as proactive types of activities and think of stewardship as the main people who get things done in the data governance area on this. So start simply assign your first round of stewards announced that you're signing your first round of stewards I've seen an awful lot of organizations do it poorly, but because they announced the first round when they people who wanted to become stewards worth appointed to become stewards sort of a head scratcher right there but different different topic on that. The idea is to say, let's get all of these people on board but maybe that's the next round and they won't quit and find out there's a professional class of David stewards that's starting to be there. So we'll have some some ideas. The idea that stewardship has to be sort of a buffer between the cadence run by it and the cadence at which data needs to evolve gradually over annual periods as opposed to weekly periods in there. But there are some foundational prerequisites around stewardship and finally start simply get good at it and practice because if you continue to practice you will do better around all of that. externally again as I said before, people don't understand and now we're adding stewards under this as well. Very complex. Keep it at the data program level, let people really get it from that perspective. And with that Natalie we're right at the top of the hour I will turn it back over to you hopefully we've got some good questions. Thanks so much Peter we do have some questions. If you could put any questions you have in the Q&A can also upvote existing questions. I'll be reading them out in order of upvotes. So the first question that we have is, are defined business processes required for data stewardship to be successful. Good question. They are not required what to focus on is problems I just arrived here and I'll put that slide back up so that everybody follows with that let's see. There's a little bit of fumbling earlier here, but I can change my house. Well, there we go. If the question is very good one does the process architecture need to be there. The answer is no it does not but if it is there it is super helpful because of course if you have a group of people already advocating an architectural approach to business. And then they're going to have much easier time selling the data architecture time, then they will, then they will the other but you do not need them to have it what you really need people to understand is that data problems show up on the other end of processes and systems. Works them perhaps necessarily, but more importantly it doesn't lead to a common understanding that's that one problem can fix a multitude of things out there if it's approached from this data center perspective. And again the guidance here is that we have one group is doing this they're practicing getting good at it. They repeat the process and they develop sustained organizational skill sets around that. So thank you for allowing me to reemphasize that point I just think it's very important that people approach this as an organizational capabilities activity. Great, thank you the next question is one of the greatest challenges is establishing domain data stewards data governance tends to have more silo data stewards for different orgs. What's the best approach to motivate data governance councils to establish domain data stewards. I agree that that identifying domains that people are associated with is an important characteristic here and put all these up to see them all up and you'll notice domain data stewards here are defined as across multiple business areas. For example, as if we're going to manage customer data we can anticipate that several different parts of the organization will approach that. If your organization is ready to add these qualifiers to the types of stewardship. And then it does make sense to talk about them but let me just very briefly warn against trying to do anything other than describe what a steward is much less the difference between a business data steward and they domain data steward. These are good definitions they tend to be used within specialized areas, but not necessarily understood widely to somebody business data steward may feel that's actually with a metadata that should be shared across multiple business there in here. The important thing is, yes, you can't attack all of your problems and data at once for starters. So to subdivide the problem it makes absolute sense to say, great, we've got another level of division down here. Again, you know customers over here and manufacturing over there and research and development over there. Those are great ways to start the process, but don't try to pull it all together and have a perfect comprehensive architecture at once. Instead take one of those areas, particularly if you have limited resources. And just a little quick side note here. My federal government customers that I'm working with at the moment are very fortunate because of new law that was passed in 2018. There are getting million dollar implementation budget for some of these activities. And so we're going to have a very nice perspective on the federal government. Having done perhaps more work and faster work than the business sector in this instance around here. But again, our question was about domain data stewards so I wouldn't start out by identifying anything other than a few people who are going to be data stewards. Let them spend some time see if they can determine easily where a source of good investment might be on the part of the organization. Carve out that piece, give it a candidate's temporary test, you know, name and say, but you know, what's the alternative to this to moving forward and iteratively improving our practice the alternative is, I'm going to sit down in advance. I'll tell you exactly where everything is and what it's going to be called and how it's going to differentiate extra mile. Now, are there people in your organization who have been thinking about these problems for a long time. Yes, did you make use of their expertise. Yes, but don't implement their plan directly in there without trying it out in private and seeing how it works. Oftentimes, while these things work terrifically from one perspective, they work less well when actually forced to go across and work for example on the side of somebody who's a data exchange partner or regulator which is even worse don't want to put those guys off. Anyway, great question and I do appreciate that domain data stewards tough things evolve them gradually rather than try to put them together all at once just as I would never introduce any of these qualifications. I'm not sure if you're in the word data steward I just want it's going to be hard enough getting people to know what a data steward is much less differentiate them. Just one more point not only before I get back to your next question on this. I forgot to take this point earlier but the legacy data steward is the fun one. What do you define as legacy, either data or legacy systems, and I say anything that's in production is legacy so that's actually certain that particular issue, as opposed to figure out whether it's like your cobalt or whatever it is that you want to do. Anyway, Natalie that's why I like doing this and we've got a great group out there asking great questions. Yes, thank you. The next question is in your experience, which is the best data governance framework guidelines best practices and so on. I don't think there is a best. And the reason for that is because I call data governance a bespoke offering. I don't mean that in the sense that you have to go out and buy a very expensive well tailored suit. But every organization is different and let me give you what I consider to be the most powerful example around that. I remember a couple of years back when we all used to travel through Heathrow and other airports around the world we would see software vendors having signed on the runways and then the terminal and things and say all the best companies use our software whatever that happened to be. No, no bad ideas around good good software, but if company x and company y are competing on the market and they're both using the same architecture it means they're using the basic same process architecture and the only thing left to differentiate their offering is around business in there. So it's a very tricky process to look at that and say, hey, within that context, how are we going to do this better. And the idea is, if we can figure out what works for us, then that will work better. So, again, I use the word bespoke. And it really means your organization is going to have different priorities in your competition and then your other business partners. The stewards are the people who get to know these better. And so when people ask for a framework there are seven good ones out there I think I have a YouTube video that you can go get access to take a look quickly at what those look like in there and they're good places for you to balance but I do recommend in a governance group. And obviously if it's governance, there should be involved but if it's a governance group to look at each of these things and try them on for five and how would that work here. Again, different presentation on that. I forget what we've done at this year or not around here but I think I know there's one out there on the YouTube videos. So, when we say bespoke, we're not saying expensive bespoke but we're saying fitted in a way that works for the organization. Again, I'll just give you one more very quick example here. But I was fortunate to be invited to participate in the army's role as data governance many years ago. And the interesting piece of it was that the army actually has a culture of governance. Your shoes are governed, your manners are governed, your weapon is governed, your job titles are governed. Everything in the army is governed and when the army discovered something wasn't governed, they went, we have to fix that as a gut level instinct. In any case, we were able to make the army's culture work very well in its favor. And let's be very frank here we were able to take also an increased troop deployment rates through better governed data through the use of army data storage. Again, wonderful stories I haven't had a chance to write that one up yet but it is a good one and certainly is something that we'd like to do, at least right here in the good old us of a. That was a great question. I love that. Thank you. The next question is who leads the data stewardship group when formed. Great question. Let's put it back to the right slide so I can address it. The idea is, really don't have to be a lot of different players in this scenario I make it very simple, particularly again at first just because it's easier to explain to people and get them to do the things that we want them to focus on around this. There we go I finally got to my slide. So who could be your leader in this area. Well, one of the, those of you that know me know I like to travel. And one of the most enjoyable characteristics is to come to an organization where people go. I call them gray beards without being gender specific around that, but people who've been around the organization for a long time who really understand this data and understand what needs to be done and they hear management come up with an really simplistic approach of, we're buying Tableau, and I don't mean to be smirch Tableau it's fine software, but Tableau suffers from the same problem everything else does. If you have bad data, anything awesome is not going to change your results to be something other than awful results around that. So, there's got to be somebody in this leadership role, who is able to say, Ah, yes, these are the priorities. I know how to get resources I know how to keep people on these tasks, I know how to do the things that need to be done for this organization and that's not just looking at a textbook and the textbooks not going to tell you what your organizational data challenges are. I translate these concepts in here so I've seen the words that were a de facto leaders I've seen them kind of like if you remember back to the old program I'm really dating myself, but most of you remember math. I'm really, you know, right there whispering and whoever the commander here was that they would need to happen. Oh, I thought the leadership is capable of this, but it's more rare that within the data community we find the leadership, then finding a business user, who really understands the value of this again recall the slide that I showed you that had data things happening, and then wonderful things happening over there. Give your organization some time. See who steps up to her various meetings you'd be surprised. I've seen wonderful people over again by my 30 plus year career here, who have just made amazing transformations and then guess why there used to be a data model or at one point and nothing wrong with that. They've actually blossomed in this leadership and they are able to help the organization really move forward around those areas. I've also seen, I ran into a fellow who used to teach information engineering at one point. I came back to an organization that I was still working with. And I said what are you doing he said well I got tired of that nobody was listening to me so I'm not teaching requirements analysis. Okay, fine. I know that was good not an individual is going to go off and become a leader in those areas. I do think that we run into the same kind of challenge that we've run into from the CIO community and that is that when we put first puts the is out there and again just not don't blame me but was part of the group to put together the original law that mandated CIOs in the federal environment. And our first question was always to put the most knowledgeable techie person in there well. That didn't necessarily translate into good leadership characteristics I think a part of what the questioner is getting at. But in addition to that then we went with MBAs and said we should put a man into it. It hasn't worked out as well either so now we're trying to figure out what makes a great CIO and we have the same challenges around data as well. And again, sometimes it's going to be down to a characteristic of an individual, perhaps in using the same word I used in the previous question be spoke the ability to have something that really works well for the organization that may not work as well for another organization, in which case again the goal guidance for organizations to nurture this talent to make it known that being able to do this I mean just imagine if your group are better able to comply with regulation if that was a deal that your organization had. Wow, then when we get the bottle for sorry didn't mean to give everybody a nightmare around that yeah when you get the bottle for you'll be better equipped to do it and perhaps at a rate that makes you competitive with the competitive with the competition which means you can put more money into that quadrant of kind to develop new and innovative things again great question thank you very much for back to you Natalie. Thank you the next question is what are the main stoppers you have faced in data governance implementation and how have you solved them. The real key there is value organizations tend to profess value. So being able to link the outcomes of these activities to some tangible outcomes is absolutely key to making sure that happens. It's not an easy process but it's actually not that hard. And again just to have to hook a particular book, but I do have a book out there that is focused specifically on understanding data values that come out from here so when you're looking at this. We've gotten good over the years of making data things happen. We now need to get good and making organizational things happen and those as I mentioned before could be increased deployment ability on the part of US military has considered to be a national priority. A bone marrow center that we worked with for many years was able to increase the number of successful bone marrow transducent from 3000 annually to 6000 annually. These are the things that move the needle these are the things that help the organization keep its eye on the pride. Any organization that doesn't believe that data plays a role in achieving its strategic objective. I just simply haven't encountered it. Again, I will go further all organizational problems, challenges, whatever you want to call them have a data component to them. And that data component is something that organizations are not quite aware of and that's really the essence of getting data centered because it means you start with the data perspective. And instead of solving one problem in an isolated instance, even though it may be painful, they're able to solve classes of problems and that's a huge value in there. So get to the point where you can practice I mentioned the book it's called monetizing data management. That's available at Amazon it's less than 100 pages. It's got 17 different cases that you can go in and look at to see if any of those patterns are useful for you so I don't mean to sell you all books. My goal is to give you guys data around this and that's exactly what I intended. So again, thank you for the question value is where you get that I am if you don't have value it just becomes an unfunded mandate and scares people. Thanks so much. We do have a couple of requests for the data steward template that you mentioned. Yes, in the slides. Absolutely. That'll be part of the slides. Awesome. The next question then is, I appreciate the reference to the importance of data management. It is so often overlooked. You can find data trustees and stewards by functional business area HR finance registrar, etc. But there's a big gap between these people and their expertise over the data they steward. These are the folks that are ultimately responsible but they don't have the knowledge or expertise to actually function as a steward. How can organizations address address the data literacy gap of their stewards. That's a very good question and a difficult question because unfortunately it involves the need for more now. I would first of all say that while the words need training my order of precedent in an organization is slightly different. I start out my data literacy efforts with organizations at the executive level, because they make decisions. They don't understand that they're making these decisions and consequently they don't understand the impacts the long term consequences. A very simple example of that is an organization that installs sales force. Now you all know sales force. It's wonderful software. Excuse me though. There's no way that an end user being presented with bad data in sales force understands the very fine distinction between sales force working perfectly to deliver bad data and sales force working as expected. And there and so their first impression from sales force is that sales force is not good. And that's not a happy situation. And all of that. Yes, these starting out with the executives is first, then absolutely the data stewards need a literacy piece. And again, I'll just make a stainless plug for my book on data literacy on sale right now. It's a great place to start as an audio version of it if you're interested, which I've had several people compliment our colleague Susan Hansfield for doing a great job on it. You don't have to listen to me the whole time. And then finally, I think data literacy will be best successful in organization that the impact of a more literate workforce will be best achieved if the third group that we make data literate are our knowledge workers. And we've been a day when HR given two candidates who are of equal category equal skills and knowledge will be able to select for the one that is more data literate as opposed to the one that is less data literate. And I also anticipate that on a regular basis you'll have some sort of data literacy reinforcement the same way we currently have security practice reinforcement. And there's the fixing and things like that just a little reminder once a year to keep all knowledge workers up to speed after of course they've already been exercised in that so yes, data stewards need to understand. They all the data stewards they just first of all they own the data right now I don't agree with that everybody agrees with me on that as well on that and that the data stewards then are just supposed to make sure that the data gets delivered to the tableau experts in the tableau experts do wonderful things and discover the people who fly into the Richmond Virginia airport on Thursdays like to have white convertible instead of yellow convertible I'm totally making that up but that you know that is a good piece. And we all understand though from a VIP expertise exercise, but the best answer from any sort of VIP expertise, excuse me, expertise VIP exercise is another question. And that becomes problematic in a number of different ways. And again, as I mentioned on this slide, in particular, the loop that is too close that only ricochets among the right hand side of this diagram and does not go back and achieve the true leverage that is possible within the data management practice area in there so that's something else that stewards have to look for and say hey, if we're starting to see things get corrected but let us correct it in the source systems that you all never have to correct it anymore. What a wonderful thing. Again, again, thank you for giving me an opportunity to expand upon that I hope I hit all the points it's not dropping the chat. Make sure I do back to you Natalie. Thank you the next question is there should be a difference between data requesting regulating organizations and data providing reporting organizations and in general between public versus private sector companies. So how to define best data stewards tasks responsibilities according to the type of work. Excellent question are there in fact differences between an organization that is myth and focused as would be most government entities and other and operations that are commercially focused. Yeah, I'm going to give you the answer from the literacy book that not so much from the stewardship practice although I think that the concept is equally appropriate. With a knowledge worker, we can, and you've already heard me say all knowledge workers should be data literate and literally certified at a literacy level and objective literacy level. Again, they're entrusted with concepts of this to do theory so they need to understand what their role is and how they have it, literally a duty to do things the correct way as opposed to, well I think I have to do it this way around it. All of these activities add up with an organization trying to do something. And let's hope that the organizations that will work for continue to try to do the right thing. But if we're going to be an organization that professors to do the right thing, we need to have our knowledge workers educated well enough so that they can catch anything that goes wrong. Now, they are the third level of data literacy as I mentioned before in this, and that there, they can be screened for literacy, open things like that, but they can also be more knowledgeable as a result of all of how do these things work. And true story I had one organization that I was working with was a health organization they were having some dashboard created, and one dashboard would have variable x pointing upwards increasing. And the dashboard right next to it had the same variable decreasing. You know, the point was, hey, folks, management events are going to be asked how can something be both increasing and decreasing at the same time. And not knowing the answer to that question or not knowing that it's even important was in fact the predominant level of knowledge that we found within this organization. So very, very problematic around that. Anyway, without babbling too much further, the early I think that they're going to be different. I don't have a good idea as to what would be differences, but certain priorities in the mission area are going to drive priorities for stewardship and governance efforts in types of organizations where the profit motive is going to guide most organizations in the other direction. Again, the, the, the idea that one size fits all that to be a perfect example why one size doesn't fit all that in fact different organizations are going to be different approaches to this but all of them are going to benefit from somebody who knows a lot about the data within the areas of the organization. By the way, the same specialization pattern that I'm raising here applies to data science I think that the term data science is absolutely useful but if you tell me that you are an animal data scientist or maybe even better still a dairy farm data scientist, then I know exactly what knowledge and skills you need to have in order to be useful in that environment and I think until we take that step, it's going to be a problem. Similarly, here I think until we get all of our organizations more data scientists who will not be able to best identify what works for them to start small, feel your way around, practice some things maybe you like jazz maybe you like doing pop songs maybe you like classical, you know, again different ways of experiencing that music. It's a way of figuring your organization out but of course management wanted to be a project, and what are we going to be done by so it is a tension that I do recognize that the face out there. So that's the question hopefully I answered it. Thank you Natalie. Thank you we have about five minutes left so I think we have enough time for one more question. And when trying to get business units to step up and allocate data stewards, how do you help them estimate how much person time they will need to budget. Yeah, so I think they probably asked that question before I jumped in and we're talking about the full time versus part time on that. I would urge you if you do have the opportunity to play somebody and say I could have two people have time for one person full time. It's the one person full time. And there is just so much to learn so much to learn about their individual data set. I'm thinking a one in 10 feet here may not be that he thought it may not be possible. But how much time well, what shape is your data and if you can't answer that question you have no idea what type of time that you can play. Do you have a bunch of mad scientists that are running machine learning algorithms all over your organization they're screening for training data. Either way that's a very narrow focus and a very good way of focusing machine learning activities is on your data that you haven't analyzed or on data that you actually know as opposed to going out and getting data that you don't know. Great question around that. If you have the opportunity, take a person it's going to be full time around that. Thank you for a good question. Thank you. I think that is just about all we have time for so thank you Peter for this excellent presentation and Q&A. Just to remind everyone we will be posting the recorded webinar and the slides to data versity.net within two business days, and Shannon will send out a follow up email to let you know the links to those and any other requested information that came up during the webinar. Thank you again for attending today, and I hope you all have a great day and rest of your week. Until next time. Thank you so much such a pleasant afternoon. Thanks Peter. Have a great day everybody.