 Hello, look, and my name is Shannon Kemp and I'm the Executive Editor of Data Diversity. We'd like to thank you for joining the current 2014 installment of the Monthly Data Diversity Webinar Series, Real-World Data Governance with Bob Seiner. Today Bob will be discussing data governance and metadata best practice. 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'll be collecting them via the Q&A in the bottom right hand corner of your screen, or if you like to tweet, we encourage you to share highlights or questions via Twitter using hashtag RWDG. And Webex does have a new user interface. If you can't find the chat section to chat with each other and make comments throughout the presentation, just click the chat icon. And again, we do encourage you to submit questions for the Q&A at the end in the Q&A section. As always, we will send a follow-up email within two business days containing links to the slides, the recording of this session, and additional information requested throughout the webinar. Now let me introduce to you our speaker for today, Bob Seiner. Bob is the President and Principal of KIK Consulting and Educational Services and the Publisher of the Data Administration News for TDan.com. Bob has been a recipient of the Damon Professional Award for significant and demonstrable contributions to the data management industry. Bob specializes in non-invasive data governance, data stewardship, and metadata management solutions. Bob will also be speaking at our upcoming Data Governance Financial Conference happening September 29th through 30th in Jersey City. I will be there as well. Hope to see some of you there. And with that, I will give the floor to Bob to introduce today's webinar and get started. Hello and welcome. Thank you very much, Shannon. Thank you, everybody, for taking time to attend this webinar, whether it's live or via the recording that Dan sends out. This is one of my favorite webinars to do, actually, the one that we talk about best practice and sometimes talk about data governance best practice, but also want to incorporate metadata best practice into this session. So this is actually one of the more popular sessions at one of the recent Data Governance Conferences. It's a Dataversity and DevTech International event. And like Shannon said, in a couple of weeks we'll be doing a conference in Jersey City, the Data Governance Financial Industry Conference. So, as you know before we got started here about the upcoming webinars, a governance risk and compliance in October, selecting the right data governance approach in November, and then big data governance, one of the hot topics of the day in December, what it is and why it's necessary, and certainly have some opinions on that and people talk about data governance and big data and whether or not there's a special thing called data governance. So we'll talk about that in the month of December. I also wanted to share with you real quickly, I know I've been talking about it for months, and finally it's here. September 1st finally came and went and my book on non-invasive data governance, The Path of Least Resistance and Greatest Success, is now available through Amazon, and I'm in the top 500,000 books. The Data Governance book is not going to be a best seller, but it's good to know that there's some interest out there in the non-invasive approach to data governance. I also wanted to let you know that the KIKonsulting.com website has been revised by the nephew who is very skilled at that. However, there are some changes that are still being made, so please take a visit to KIKonsulting.com if you get a chance, if you want to learn more about non-invasive data governance. As Shannon mentioned, the Data Governance Finance Conference is coming up in New Jersey City in September, and then the week after that I'll actually be speaking at the Data Diversity Event, Enterprise Data Diversity in Chicago, Illinois. I'll be speaking on the 8th of October on a couple different metadata-related subjects. So this is a subject that's of interest to you. That should be a very interesting event, should be a well-attended, hopefully, event of some people where there will be a lot of discussion, and it's going to be more tutorial-based than the individual hour sessions that you've seen at a lot of the other conferences. Somebody asked early on what the agenda for the session was. The agenda's changed a little bit. There's some things that when I started to revise an existing presentation on data governance best practices that I needed to be spoken about. One of the things that I'll talk about right out of the gate is the relationship between data governance and metadata. And I have given presentations at some of the Data Diversity Events regarding data and data governance being a two-way street. The fact is that if you're doing data governance that some of that information, some of the metadata that you really need to run your business is going to be a byproduct of data governance. And then the metadata itself also needs to be governed. So we'll talk a little bit more about that. As we go through this, we'll talk about when it comes to determining what is the best practice, whether it's best practice around data governance or metadata or business intelligence, what are the criteria that we can use to determine if something is best practice? I actually had a client stop me recently in a presentation of the best practices to some senior management saying what those criteria are really important. We'll talk about those two important criteria to determine whether or not something is best practice. And then we'll talk about best practices for creating best practices. We'll talk about best practices for demonstrating value, best practices for roles and responsibilities, and then applying governance and metadata in our daily life and getting metadata and data governance part of what we do on a daily basis, basically. You know, for a lot of you, again, a lot of you have attended many of these webinars and I really, really appreciate that. But for some of you that these may be new to, I want to start with a definition of data governance. And I'll just go through these kind of quickly, but I have a specific definition that I use for data governance. And I call that data governance is the execution and enforcement of authority over the management of data and data-related resources. Well, the execution and enforcement of authority is oftentimes worded too strongly for some clients. Some clients really like that and use that in their definition of data governance. But some say that we need to tender that a little bit. We need to temper that a little bit and make it a little bit softer of a definition so people don't cringe when they hear it. And I'll share a couple of those with you here in a second. But the data stewardship, which we talk quite often about, the data stewardship is really the formalization of accountability. And if you've listened to me before or you want to go back through the existing version of the different webinars from previous months, you'll know that I have a specific thought process around data stewardship. In my opinion, everybody in the organization that has a responsibility for data that either defines or produces or uses data in their daily job that are steward, whether they like it or not. It's not kind of an opt-in, opt-out type of thing. It's if that person in the organization has a relationship to the data of defining, producing, or using the data, we need to formalize their accountability. We need to know who they are. We need to let them know that what they do has an impact on other people in the organization. So when I talk about stewardship, I'm really talking about the formalization of accountability and then data governance is the execution and enforcement of authority. Because at the end of the day, it's necessary to be able to execute and enforce authority. Ask anybody who has a job related to risk management or compliance or to privacy or to information security. There's no optimal aspect of it there. We know that we need to do those things. So we need to enforce authority. We need to let people in our organization know that they have an impact on the way the data is being defined, produced, and used in the organization. So just real quickly here, what I want to do is I want to share with you some definitions from some recent clients for data governance. One is the use of authority, discipline, and behavior change to ensure proper management of information assets. Again, that's a softer definition, but it says the same thing. It says that we're using authority, we're using discipline, and we're using behavior change to manage our information assets. Another client used the definition almost exactly, the execution and enforcement of authority over detection of data. So that was one of the things that that organization was focusing on. So as a best practice, it makes sense to have a definition for data governance. It makes sense to have a definition for metadata, and I want to share one of those with you here in a second as well. You know, have that so when people ask you what data governance is, there's some consistency in the language that's used to answer that question. Just real quickly here, non-invasive data governance is the purpose of applying this formal accountability through non-invasive roles to existing or new processes to ensure that, as I said before, the definition, production and usage of data assures regulatory compliance, security, privacy, protection, and quality. So really what we're doing is we're describing how we're going about doing it, how we're applying it. We're applying it in a non-threatening way with the goal being to be transparent, supportive, and collaborative wherever we can. So let's talk about the definition of metadata, and the one definition that seems to be used most often by people in the industry is that metadata is data about data, and that may glaze some people's eyes over and they may roll back in their head when they look at the meta and at a description of metadata. So I gave it a little bit more of a complete definition, at least in my mind, where metadata is data that's recorded in IT tools that improve the business and technical understanding of the definition, production, and usage of data, the three main activities that they take place with data in our organization. So let's take a further look at that definition. So it's data. Interesting, way back when, when I got started in data modeling and in data processing in general, we would go to lunch meetings and people would pull on a napkin and they draw a picture of a data model of how here's how we're going to relate the data. We had some pretty geeky lunches back then, but it's data, and the fact is that it needs to be recorded somewhere, so even if you record it on a paper napkin, is that the best place to store your metadata? No, without a doubt, it's not the best place, but the fact is it doesn't really become metadata until you record it, just because somebody knows a business rule associated with the data doesn't mean that anybody else knows or understands or acts consistently with that business rule. So first of all, we've got to record it, we've got to record it somewhere, whether it's in an IT tool, like a data modeling tool, like an ETL tool, like a data reporting tool, or whether it's in a spreadsheet or a Word document, or a share point document, or something like that, it needs to be recorded somewhere, it needs to be organized, it needs to be categorized, but it needs to be recorded. That's the first, the most important thing, because the knowledge about the data, the data about the data really doesn't become helpful until it's recorded somewhere, and the idea is that it's not just a technical tool, that it improves both the business and the technical understanding of the data, understanding of how it's defined, how it's produced, and how it's used. So yes, we can start with a definition of metadata being data about data, but my suggestion is that we take it a little bit further and we take a look at, you know, what are the best practices about when we record that information, how we make it available to the business, how do we use it to help people to define producing used data that are in their organization. So there we go, so when we're going to start with best practices, why don't we start with best practices as to how we should present data governance and metadata to the organization. A lot of organizations sell data governance as being this huge, expensive, time consuming, invasive type of ordeal. The same thing is true with metadata, but if you take a look at it, most organizations are already governing their data to some degree. They may be governing it inefficiently and effectively informally and they may have some metadata. I've had some people that have told me that they don't have any metadata in their organization. Well, somewhere there's either a data model or a database layout or a data dictionary or maybe it's in people's head, but there is the metadata. What we need to do is we need to move from managing that informally to managing that formally. So if we go into the messages, we're not doing any of this and we still need to start from scraps, that's a lot different than going to management and saying we're already governing data. We've got information security. We've got compliance. We've got certain things in place. We've already got metadata. We do data models. We do data dictionary. We do whatever we need to do, but we're already governing data and we already have metadata as well. We can formalize how we govern it by putting structure around the governance of the data and the metadata. We can improve how we manage risk, how we improve quality, the coordination cooperation. The fact is we in an organization don't need to spend a lot of money on data governance and metadata. Maybe some of the software tool vendors will disagree with that, but the fact is that we're really trying to govern people's behavior. We're not governing the data itself truly. It's people's behavior that's associated with the definition production and usage of the data. And the last message of what I suggest to tell management is we need to have some structure. We need to have something that puts in place some structure around those things that we're presently doing informally so we can move to a more highly governed environment. We can move to a more metadata intensive environment. We've got to take the cover off of where metadata is hidden. We can make it available to people. My presentations in Chicago at Enterprise Data Diversity are mostly around metadata management and even the creation of something called an operational metadata store. So that's something that may be of interest to you, but we need structure. We need to really have some best practice around what we tell our management because management is going to believe what we say at least we hope that they do. If we tell them that it's challenging, it's going to be expensive, it's going to be resource intensive, that's what they're going to understand. If we tell them we're already doing this, we can formalize it, we can take advantage of things and it really only costs us the time we put into it, you know, maybe they'll understand that as well. So those are the messages to share with management. The messages not to share with management are that we shouldn't sell governance and metadata as being a huge challenge. We should emphasize that it's really not in its entirety a technical solution. You know certainly there are some tools from a data governance and from a metadata perspective that will help with the deployment of those disciplines within your organization, emphasize that we're really governing people's behavior and not the data and emphasize that governance and metadata is really an evolution. It's not a revolution. So it's not as though we're going to go from an ungoverned environment to a governed environment with the flip of the switch. The same thing holds true for metadata in your environment. So it's one of the beliefs that there are some best practices around the messages that we send to our management. What we say to them about what approaches are available to them. And if they understand that there is a non-invasive approach, or there's a way of leveraging the things that we're presently doing, you know that seems to be what organizations like to hear. If they can do it in a less expensive, less invasive, less threatening way, it's something that is would be considered very highly in that organization. So one more best practice message would be we've got to define what it means to govern something. We can put a definition to data governance, but if we don't define what it means to govern something people are not going to really ever get it. And so the definition, the free dictionary.com definition of what it means to govern something, and what I've done is I've added the terms data to the end of it, is to make an administer public policy and affair. The exercise of sovereign authority, to control speed, to regulate, to control actions or behaviors. That's really what we're trying to do. We're going to try to do it in a non-invasive way. It's not all about command and control. And it can demonstrate to them that we want to achieve what we're saying we want to achieve by governance. And we're using the definition of govern to do that, that we can do this in a manner that's going to be less invasive than what people wish to, which is the kind of hit people over the head with a stick and get started that way. So we're going to talk about metadata as a data governance byproduct. If you are going to govern data, one of the things that you need to know is who does what with the data. If we agree that executing and enforcing authority over the data is what governance is all about and that we're going to formalize accountability for people, then we need to move from this informal, not recorded, people just know who to contact when they've got a question about something. We need to move from that type of an environment to a formal environment to a recorded environment. We need to take the information who does what with the data and we need to record it somewhere. I'll share with you a template and if you've been in my webinar before, you've seen the common data matrix, but there's an easy manner to record the information about who does what with the data. And my suggestion is if we record who does what with the data, what we're really doing is we're identifying who the data stewards are. We're not assigning people to be data stewards. Think about it in your organization, are people already busy 100% of the time? As people have put it to me, they people have day jobs within the organization. So if they're assigned to be a data steward, then that obviously immediately feels as it's over and above what they've presently got going on in their responsibilities. So if we can just identify who the stewards are and help them to understand the impact that they have on the data, that's a better approach than to assign people data stewards. And if you remember what I said earlier, the idea is that pretty much everybody in the organization may be considered to be a data steward, the different types of stewards, but anybody who defines produces and uses data has an accountability for how they define produce and use the data. So the idea is to write it down, rather than best guess, make formal rather than keeping it informal. And the question becomes, well, why is it important to write it down? Well, it's important to write it down because people want to know that information. I've had clients say, well, it takes us weeks or months to get the right people in the room to discuss and to solve a specific issue. Well, we can solve that problem pretty quickly by recording who does what with the data and the different systems across the organization. And again, like I said, I will share with you a template that may help you to do that. So if you want to identify and record the people that have the responsibility for defining the data, we want to identify and record the people that have the responsibility for producing the data and we want to also know who's using the data as well. So we've got to, the metadata without a doubt becomes a byproduct of your data governance program. And the question becomes, what are you going to do with it? Are you going to have best practices around that metadata that's associated with data governance? Are you going to have best practices around governance or metadata by themselves? The real question becomes, how are we going to do this? And what are or what constitute best practices around data governance and metadata? I want to ask yourself a question. Are those people that are defining, producing, and using data? Are they the data stewards? How can we benefit from by knowing who those people are? What's it going to cost for us to do that? The inventory certainly takes time. It takes some effort, but it doesn't have nearly the value that you would get from having that information about who the stewards are. So it makes sense to record who these people are. Is this information about stewards metadata? I would say that it is. It's a byproduct of putting data governance program in place, specifically on invasive data governance program, but using these tools, it's very important for us to know about who does what with data. And we need to record it somewhere within the organization. So let's talk about metadata stewards. So we talk about data stewards as being any person in the organization that can find, produce, or use data. Well, how about metadata? Metadata stewards. If metadata is a type of data, it is going to govern itself. We need to identify the people, have the responsibility for defining, producing, and using the metadata as part of their job function. So people that define what metadata is necessary, people that produce the metadata and enter it into the tools or enter it into the spreadsheets or to the dictionary, people that use the metadata, we need to know who these people are. So we may want to create a common data matrix just for the metadata stewards so that we can make certain that the metadata is being recorded when it needs to be recorded, that it's being produced, it's being made available to people that can get value from it. So we know how that metadata is defined, produced, and used, and we must govern to that. People must be held formally accountable for metadata actions. And the truth is, if you take this slide and you replace the word metadata with data, it's a definition of what a data steward is. So what we're talking about here is a data steward for, specifically for the metadata itself. And the metadata that's the byproduct of the governance program needs to be stewarded by somebody. It needs to be governed by somebody. And typically that's the data governance team that has the responsibility for capturing this information and making certain that it's available or made available to people within the organization. So what I'm going to talk to you a little bit more about here is the best practices for creating best practices. And so that may sound somewhat redundant, but I assure you that it's not. And so typically what I suggest to organizations is they start by defining a limited set of best practice for their organization in relationship to governance, in relationship to metadata. And what I'm going to do is rather than share with you kind of a general list of best practices in this webinar what I'm going to do is I'm going to share with you on the next couple pages some specific examples of what some organizations have selected to be best practice around data governance and metadata. So what these organizations are doing is they're defining the basis and the guidelines for behavior to deliver both their data governance and metadata programs. And they start by defining that limited series of best practices. Then when they have the best practices, first thing you do is define why is that best practice for the organization. What is the organization doing presently that they can leverage to support the best practice? But then where's the opportunity to improve? So rather than calling them strengths and weaknesses, my suggestion is that we look at what are we already doing that we can use to support this best practice around governance, around metadata. And then where are there opportunities to improve? What's the risk? What's the gap between what we're doing presently and what we say we want to be doing in the best practice? And what are the risks that are associated with that? So typically that is what constitutes a best practice assessment. For most organizations is find the best practices and that's where the organization is in comparison to them. And then build some recommendations as to some next steps. So I talked a little bit earlier about these two very important criteria for determining whether or not something use best practice to an organization. The first one is that is the question, is the statement of best practice practical, doable, and will it add value? And I would tell you that you want to be able to answer that question, yes. So if the statement is practical and doable and it will add value to achieve that best practice, then I would consider that to pass that first criteria for determining whether or not something is best practice. And the second one and probably even the more important one is that is the program going to be at risk if that practice is not achieved? So if we don't achieve what we define as being best practice around data governance and metadata, are we going to be at risk? And I would venture to say that the answer to that question should be yes as well. So if we use those two criteria to balance up against each of the best practices, we can pretty much say that we know that we can achieve it and we know that we're going to be at risk if we don't achieve it. So maybe that would be something that we would want to consider to be best practice in our organization. So I think before that I'm going to share with you some different best practices from organizations and you're going to see some consistencies from one to the next. So the first one is from an insurance company that the best practice is there's a high level of senior management support, sponsorship and understanding of the activities of the data governance team or of the metadata team or both the data governance and metadata team. But the question is can we get senior management to go beyond supporting and sponsoring the activity, get them to understand exactly what we are going to do to achieve the goal of having a governance and metadata program in place? Is it doable and doable? Well if it's not then you're in a lot of trouble, I'll be honest with you. This seems to be a best practice with a lot of the top best practices for a lot of organizations. Are you going to be at risk if your senior management doesn't support sponsorship and understanding? I would say the answer to that question is at some point it's going to be yes, that people are not going to give their time. A team of cross departmental resources as well as a dedicated program manager are accountable for the sustainability of the program. Is it achievable? Do you want somebody in the organization that has the responsibility for putting these programs into place? If it's practical and doable well first of all you're probably not going to be able to put a program in place unless somebody has that responsibility. In order to risk if you don't have that team of cross departmental resources as well as somebody to leave the effort I would say in most organizations, the answer to that question is yes. The measurements of success and how they will be measured of the program are well defined and communicated. Again it's very often a best practice for an organization that they want to be able to measure if they're going to spend any time and effort towards data governance or towards metadata that we need to measure how successful or what success means to the organization. The goal is to scope objectives and role responsibilities. I should probably be, I'm certainly how you use the tool here so I should be checking things off. I guess that's not checking things off. But the scope and objectives and roles of engagement of the program are well defined. So let me share with you what another organization from a government agency, and again they said senior management supports sponsors and understands. Now some of you may be asking the question as to why are some of these terms underlined. Well because these terms aren't necessarily very highly understood or very well understood within an organization. So when we are developing best practices, when we are sharing them with people, when we are using terms that may not be in their normal vocabulary, then we want to also provide definitions to what they mean. What do we mean by senior management? What do we mean at a high level or data governance or metadata? You know what do we mean by policies and procedures or resources? You know we need to define that so that people can really get a better understanding of what we mean in these best practices. So again for this government agency they made it best practices that resources are allocated to define, develop and execute the programs. Policies are procedures are sure that it's not optional, that internal and external stakeholders are being informed of the goal scope. So again very often there are very many similarities between the best practices from the previous example that I gave in this one. Let's look at another one from a best practices from an education company perspective. And if you noticed in the first two sets of best practices they're written in the present tense. I'm excited this is kind of a change of heart for me. In the past I've written them in the in the in the future tense. It's really best practice and we're comparing ourselves to it. You may want to consider writing it in a present tense. But when I say senior management support sponsors and understands compare that to the next one that says pretty governance to be successful there will need to be a high level of senior management support sponsorship and understands. Again a very similar best practice but it's done a little bit differently. It's written more as a future tense. And I think it's just a more of a matter of style or preference in your organization. But you know you want to state things that are practical and doable. You want to state things that you're going to be at risk if you don't achieve them. Selected staff must be committed to the definition of development execution very similar to one that we saw before it. The data governance principles will be applied consistently and continuously. It's one that's used by a lot of organizations as well. Is it practical and doable in an organization and are we going to be at risk if we don't apply our principles around data governance consistently across the organization. And let's see and one more from the university. The goal of scope expectations and measures of success will be well defined and communicated. Data governance should be considered an ongoing program rather than a project. Other individuals in the steward role will have business expertise. So you'll see there's a lot of different styles in the variation of best practices. But what do we do once we get the best practice? Well again we're going to define why it's best practice. What is it that we're doing that we can leverage to support it? How much does senior management already know? In what forms do we have available to us to get in front of them and to improve not only their support and sponsorship because we might have that already but their understanding of what it truly means to govern the data within the organization and what it means to do it in a manner that's less invasive, less threatening to the culture. So again it's a matter of preference, it's a matter of what you divide. And if you're interested, I believe that we've shared this in the past so I think we can share it at this time as well, there's a list of best practices that I use that have been written most specifically for data governance but then they could also be applied to meta data as well. And if you're interested in seeing that it's just basically kind of a laundry list of best practices that I've pulled from multiple places at one time. Basically to summarize the best practices for creating best practices you'll find that they very often relate to these same five subjects that senior management, sponsorship and understanding no matter how we say it, resources must be applied no matter how we say it, role responsibilities clearly defined, goals, scope, expectation, measures, those must be clearly defined and we must do this consistently within our organization. So I think you said that as a common theme as we're defining best practices in our organization. So examples of some recommendations that have come from best practices assessments is put an operating model together roles and responsibilities, build a communication plan, staff the team, be constantly defining and formalizing the actions of the people that are going to be participating. And let's develop a documentation platform, a metadata platform that not only includes the metadata that's the bottom product of the data governance program but any other metadata that we have regarding the definition, production and usage of data. Let's take years for a second here. Let's talk about some best practices around demonstrating value from, from governance and from metadata. And it's always been my suggestion or maybe it's not always been but it has been a lot recently or for a time now that if we can get the business people to tell us where data governance and metadata will add value rather than us trying to sell it as the best things in sliced bread or being the silver bullet that's going to solve all of our data problems. If we can get the business folks to tell us where data governance will add value or metadata will add value if they can define for us what they can't do because the data is not there to support it and if they had the data what they could do with certain things I want to share some examples of those types of things. That's where it's weighed in gold when we take it to our management I've done webinars on this subject specifically of getting the business folks to speak of and I want to share a couple of slides with you that I shared back then about what are some of the questions that we can ask to get a better understanding of the pain points that they have of things that they can't do because the data is not there to support it the metadata to improve their understanding of that data is not there to support it. So here are some relatively simple questions that we can ask people we can ask them what data and metadata they use most often to achieve their function and their jobs you know what's the primary source or system that they use their data or where do they go to get their metadata who do they know to to get this information what processes do they work through that consume the data we can ask them questions like this what processes do those processes have data and metadata challenges what is the product or customer impact of this we can ask them questions like that we can ask them questions like well is there an associated risk in our brand and compliance and legal you know what's the operational impact what other impacts should we note have we identified potential solutions you know we can ask these types of questions or we can use them in a different way we can say that we can get them to respond to that the inconsistencies in the definition production and use of specific critical pieces of data prevent them from doing what and here are some examples of what one client told me they couldn't compare costs across regions they couldn't identify the best place to get substitute materials obviously a manufacturer a manufacturing company provide for for sophisticated ad hoc you know i have universities tell me they can't track students to be able to identify who's at risk they can't identify all the customer touch points because the data is not there to support it how does that cause them problems so we can get them to tell us what where inconsistencies in the definition production and usage of data prevent them to do certain things that becomes worth its weight in gold then what we can do is we can share those with individuals we can and share those with individuals at senior management level and we can tell them that we've got a solution a way that we can improve on some of these things but it's not us selling the virtues of governance and metadata it's the business telling us what they can't do because data and metadata isn't there to support it you know data governance provides a formal framework of roles and processes that will enable us to do certain things again here's just a list of examples I apologize for them being so small but governance basically provides that framework of roles and processes that helps us to formally address accountability to formally investigate and report the benefit from gaining consistency in the data formally investigate and report the costs associated with that benefit and the gain from it so we can take a different a little bit different of a tact and really depends on what's best practice for your organization do we just want to tell them why this is great or do we want them to tell us why it's necessary I would venture to guess that getting them to tell us and using that as ammunition when we go to our management becomes an extremely powerful tool for most organizations and in the course of these conversations share with them certain concepts like don't let perfection get in the way of good enough or don't let perfection get in the way of progress that there's a point of diminishing return we don't have to govern the heck out of the data we don't have to govern the metadata the heck out of the metadata associated with the most important data in our organization but we need to at least define a minimum set of requirements that we will achieve and so that's where we're saying let's let's there's a point of diminishing return you know we can try to govern too much we can try to govern the metadata too much but what exactly do we need it's interesting in the webinar that I did with Scott Ambler earlier in the year about agile and data governance I'm sure that that from an agile perspective that the doing return comes very early but you know at least we have to set that minimum set of requirements for the metadata that we're going to collect and how we're going to govern the data so again I'm not going to read through all the key concepts sharing I don't have the time I want only time for questions I see that there are some questions so I want to kind of walk through the the additional slides that I have here but let's keep in mind when we're demonstrating the value of governance to our organization maybe there are best practices around how we go about defining how we're going to demonstrate value in our organization certainly there are best practices around defining roles and responsibilities there's a lot of different approaches a lot of different organizations have there's I have mine in the modern base of approach to roles and responsibilities I as I mentioned earlier I think just that we identify people into roles rather than assigning them into roles and leverage existing responsibility wherever we can that's compared to a command and control approach which you're going to assign people into roles hey I don't I don't care how the heck busy you are you're an Alice doer to this and we're going to give you these responsibilities it feels a lot different than if you identify somebody into the role and say that really what we're doing is we're not changing your responsibilities we're just helping you to be better accountable and to follow the rules associated with the data so there's a big difference in the field a non-invasive approach versus a command and control approach and then the real extreme is the two by four approach is state governance is not optional people will need to make time let's pull out two by four smack people ahead and tell them that it really doesn't matter what they're doing they need to steward the data well well that's a lot more invasive than the idea of identifying and recognizing people for relationship to the data and helping them out so real quickly here the different roles and responsibilities in most organizations when they're defining roles and responsibilities around data governance there's executive level strategic level practical operational support levels and a lot of that is represented in this diagram right here the ones that I use fairly regularly to define the operating model of roles and responsibilities I've done complete webinars on this and I don't intend to to spend the amount of time on it and that is really necessary to define this but we need to follow best practice when we're defining roles around governance when they're defining roles around capturing the metadata making the metadata available using the metadata and oftentimes you know people look at the diagram and they say wow it's really bureaucratic there's a lot of different levels until we can add those things on the outskirts of the diagram and say well you know what our data governance partners that already exist we've got regulatory compliance we've got PMO we've got IT you know we've got a team of people that have responsibility for data governance we've got executive leadership we've got you know maybe there's something that we can leverage as a data governance council you know typically the bulk of the work and I've stated it before and I'm sure I'll state it again is in that yellow section is the tactical level where we stop viewing data as a siloed business asset we start looking at it as a cross business unit asset so there's best practices around how we define roles and responsibilities there's best practices around how we define data stewards are they everybody or are they just specific people now I've been in organizations where they pointed at six people it's okay well these are our data stewards there's a data steward for customer a data steward for product a data steward for finance so the first thing I'd like to do is change your opinion here that really anybody who defines produces or uses that data has an accountability they're by definition they're a steward they're stewarding the data for the organization so then there are best practices around roles and responsibilities and defining the value and you know each is just kind of the generic best practices that I shared a little bit earlier on so here's a diagram that I mentioned earlier the common data matrix and we usually provide this with the email that goes out after the webinar or I don't present this right out of the gate right out of the gate because people start filling it out and stop listening to me but the common data matrix is very important for cross referencing the different type of data that we have in different systems with who an IT has the responsibility for data from a data subject matter expert or from a system subject matter expert perspective who and what parts of the organization use that specific type of data in that specific system whether or not we put an X in the block or we just put a person's name the roles and responsibilities alongside with recording who does what with data really are the stalwarts of making a data governance program effective in your organization so we share the common data matrix and the the templates with each of the webinars so let's talk lastly here about applying governance and their best practices for applying governance so typically in organizations they will apply governance and they will apply metadata in a proactive way and in a reactive way so whether they're going to build it into the process they're going to build it into their methods or whether they're going to use it to respond and in a lot of organizations they decide to take the reactive approach first and in fact with one of the organizations I'm working with now they've decided that they want to be reactive in how they're going to apply governance and metadata in their organization but as they define the process for being reactive they've not built it into their methodology so there's a little bit of proactive nature in it as well I wanted them also to understand that when you're proactive when you build it into your for example system development methodology that you're going to identify problems that have to kick you into reactive mode as well so the best idea here and the best practice around applying governance and metadata is to build it into the process so rather than redefining things redefining the processes my suggestion is let's overlay the roles and responsibilities over the methodology over the workflow one of my pet peeves I talk about it fairly regularly is that people call things data governance processes I like that term I think that any process is governed but just by the fact that it is a process that we know who gets involved in what step what step of the process don't call things data governance processes because it's implying that we're doing this process because we have governance well your information security processes aren't in place because you have governance it's because you have information security your risk management processes aren't in place because of governance it's because we need to manage risk so what we're going to do is overlay roles and responsibilities over existing methodologies and workflows wherever we can and we're going to do it in a non-invasive way so for example if we have a system development lifecycle methodology or an application development lifecycle methodology data development database development whatever you want to call it there are specific steps some requirements analysis design development things like that that are pretty standard in a lot of these methodologies and my suggestion is well first of all those steps are typically pretty formal they're written down they're followed if you do follow a lifecycle methodology or a methodology of this choice what we really should do is we should cross-reference these different roles and responsibilities that are the best practice we should cross-reference them with the different steps of the activity that we are trying to govern and we can identify what metadata needs to be collected in these steps as well so in this example that I'm showing here it's labeled as being a master data development lifecycle and then we've got the steps down the left hand side of information gathering assessor requirements planning and analysis so we're going across the top we have the different roles that we've identified for part of the data governance program and we can be very specific as to what we expect out of that specific role in that specific step of the methodology some organizations have gone as far and this shouldn't be surprising to you of having time offs at certain steps to make certain that we involve the right people in the right way in solving a specific problem another example of that here's an organization that was applying governance to a master data master data project and the pilot was in material and here's what they wanted here's the steps that they wanted to do here was the estimated time period to complete and who was responsible who was accounted if you're familiar with the racy or RASCII depending on how you use it in your organization here's a way again of cross-referencing the steps of the methodology with the different roles that you've defined as part of your organization and so that's best practice is that we are utilizing the things that we've defined as part of our program and we've utilized them for a specific purpose to achieve a specific goal or to govern a specific process better we don't have to call it a governance process it's a process that we're going to govern and that any of there's a big difference another example one I show fairly regularly this is an example from an organization that one it was putting together these six primary activities if you can see them highlighted in blue of reserving or researching information quality issues identifying and monitoring risk monitoring improving quality in life cycle well the way that they had it set up was you just click on one of these and okay I guess you don't just click on them is what would slide in is these are the specific steps to achieve that repeatable action here's the different roles that are associated with it and who's responsible who's accountable who's supportive who we need to consult who needs to be informed while we are while we are governing this process so again there's best practices governing processes and that's let's identify who needs to do what and when let's communicate with them that they need to be involved and let's help them and provide them with the tools that are necessary to be successful in our implementations of data governance and metadata so basically you can do this in a reactive mode as well with a data quality methodology where you qualify and prioritize data issues identify affected domains affect identified students and identify affected stewards conduct the root analysis and all of those steps but you can cross-reference that in many organizations it just looks a lot like the diagrams that I just shared with you where we cross-reference the steps of the reactive methodology with the roles that we've defined as part of our program so basically we're cross-referencing who knows what and when do they do it and we can also add to that what type of metadata is going to be a result of it so just trying to quickly to wrap up before we take some questions and I see that there's been some questions here as well is we talk about data governance and metadata being a two-way street we identified some criteria to determine whether or not something is best practice that being is it practical and doable and are we at risk if we achieve it we talked about some best practices for creating best practices best practices for demonstrating value roles and responsibilities and applying governance and for applying governance and applying metadata to specific processes within our organization just to remind you real quick before we take questions of the upcoming webinars that will be taking place in October November and December and with that I'd like to turn it over to my friend Shannon if Shannon's out there and see if we have any questions that I can answer for you in regards to data governance and metadata best practice oh lots of questions and always great questions and one of the most popular questions of course is a request for the slides and the recording and just a reminder that I will send out a follow-up email for this webinar by end of day Monday with links to the slides the recording and anything else requested to run out the webinar Bob the first question coming in is just to help the definitions you've been using is a steward the same as a custodian? Wow it really depends I mean there's different names that people assign to different roles different names that people associate with different roles and I've heard the term data custodian used often times and I didn't get a really chance to walk through this too much in the webinar but I differentiate between the tactical level of data steward and the operational level of steward in my model so an operational steward could be anybody that defines produces and uses data in their job a data domain steward or somebody that has responsibility for a subject area of data across the organization is more at a tactical level because they're looking at it more across business lines than within a specific business line I would say that a data custodian relates closer to what I would consider a data domain steward or a data subject matter expert had an organization say well really this data domain steward is truly just a subject matter expert right? And then I guess in some ways that's true so it's all right well that's what we're going to call them subject matter experts so I've heard them called many different things enterprise data stewards because they're taking an enterprise perspective I would more relate the custodian as a formal role to the tactical level data domain steward than I would to an operational data steward that doesn't mean that everybody who defines produces uses data in their job is not a custodian of the data so I could see it used either way but I would lean more towards the tactical role to define the data custodian so moving on to the next question how is the establishment establishment of roles defined? Well easy it's typically pretty easy because a lot of organizations look at this operationally tactically strategically and from an executive perspective in anything they kind of combine the executive and the strategic level into a single level so if we want to define roles and responsibilities first of all we want to make certain that at an operational level we identify people that touch the data that are stakeholders in the data that we understand how they touch the data and we know who they are so we can help them to govern the data better you know we also recognize that and I hear this in the majority of organizations and probably in your organization as well where they want to break through the silos we've got silos of data specific to each of the different parts of the organization and we want to break across that so we need to define some roles at a tactical level that start looking at data just specifically within a business line and start looking at data across business lines so that's the kind of the tactical view the strategic and the enterprise view somebody needs to be the ultimate authority somebody needs to make decisions somebody needs to say this is important and that we're going to follow in on it so from an operational tactical and strategic perspective that's how we typically go about defining roles and responsibilities when it comes to to governance more specifically but also to metadata as well but not only are there operational tactical and strategic there's also support roles you know typically I get the question that they want to know whether data governance should reside in business or in IT and I answer that question yes that it needs to reside in business and in IT and there is certainly a supportive role for IT there's a supportive role for the data governance team so we look at things as operational tactical strategic and then supporting and that basically is represented in that peer diagram of how I define roles around governance for more organizations going back to your conversation about the napkin and using the napkin and we have a comment that says I think napkins are valid IT tools especially bar napkins I think that many people would agree with that and from the same time a question came in about how information about data about information about data well you know it's interesting you know a lot of organizations are moving to packaged environment well first let me address the top of the napkin they don't keep very well especially if you spill some beer on it or you get a coffee ring on it so not necessarily the prescribed approach for capturing your data but you know information about data that can be information in your data modeling tool information in an industry standard for a specific piece of data and what the acceptable values are you know wherever there's information about data where we move data from one place to another and what sharing rules associated with it those are all metadata you know how we produce that metadata how we use that metadata what are the rules associated so any information and it's pretty interesting that you know any information we have is information about our data how we can use it how are we restricting this use what are the privacy rules what are the financial regulatory rules around this data what's the quality of the data how many different times do we have this data defined and in how many different ways what's the turnover of the month you know how did an organization that wanted to compare you know wanted to define something as a renewal date for a policy well renewal it was granted it was determined to be a renewal if it followed a specific business rule well you know what there were five different versions five different flavors of our renewal of determining whether or not something was a renewal versus new business for the organization any information that we have about data could be considered metadata we can go too far and we can collect too much that nobody's going to use but if we define metadata requirements and we understand what metadata it is what information their term to use what information about the data do they need to help them to do their job better you know those are typically the approaches I define for identifying well what information about the data is important to us and it's really not that difficult to do when you kind of look to definition metadata production metadata and usage metadata it's pretty simple when looking at things please discuss the role importance and purpose of a data governance office in order to implement data governance with metadata management and data quality management well that's practice and we defined this earlier somebody has to have the responsibility for putting governance in place somebody has to improve how we manage data as a valued asset whether it is a data governance office or a data governance officer somebody has to have that responsibility organizations are setting up data governance offices as the organization still standing okay I got I heard a noise as if I was I was cut off it's very important to have a have a part of the organization that has the responsibility for data governance whether you call it a data governance office whether you call it a data management office whether you call it a chief data officer's office whatever you however you define it for your organization it's very important that somebody has the responsibility for this and whether you call it the data governance office or whatever you call it within your organization I think it's a new role it's a burgeoning role within a lot of organizations questions are perfect I don't know if they're perfect but they're they're hopefully they're good enough for the cause absolutely so we have a a question here I'd like to hear more about the role of data policies in the metadata effort how is it best to develop new data policies by governance and then help enforce them via metadata so policies so we actually did a webinar on this not too long ago on data governance policy so please look it up and see the the old versions of it you know policies can be very important to some organizations if you're a policy driven organization to have a policy around governance that states that we're going to manage data as an asset that we're going to have clearly defined accountability that we're going to follow the rule regulations that are given to us and that we're going to be consistent in the manner that we do it I've seen organizations create policies around that I've seen organizations say that we're going to have a policy around what metadata we collect less of those than I've seen more than I've seen around just data governance policy in general but it really depends on your organization whether or not there is need for a policy and if that's what it takes to get people off the snide and get them moving then it's important to have a policy but otherwise you know I'm only I'm only really an advocate of creating a policy where a policy is necessary to get people to act and if it is then start with a policy if you get your senior management to sign up for it on it you're better off than most because it says they understand it they act you in the steps that you're taking and it comes back to that very first practice they say it's important to sponsor you but do they really understand what the heck it is that we're doing if we can get them to understand better then there's a more like a higher likelihood that we're going to be successful in our organization and for this all that we have time for today but one of the great things about this series is Bob will write up the answers to the questions that we haven't yet to answer so keep them coming in and I'll make sure and then get a copy of those to Bob Bob thank you so much another great presentation from you today and thanks to all the attendees for the for such great interaction we do love all the questions you submitted and quite a few tweets going out today which I just love all the engagement with all of our webinars thank you so much and I hope everyone has a great day very much Shannon and keep the questions rolling in and hope to see you next time thank you