 Hello and welcome my name is Shannon Kemp and I'm the Chief Digital Officer of Data Diversity. We'd like to thank you for joining the current installment of the Monthly Data Diversity Webinar Series Real World Data Governance with Bob Siner. Today Bob will discuss how to govern glossaries, dictionaries, and data catalogs sponsored today by Elation. 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. If you'd like to chat with us or with each other, we certainly encourage you to do so. And just to note, the Zoom defaults to the chat to send you just the panelists, but you may absolutely switch that to network with everyone. For questions, we will be collecting them by the Q&A section and we encourage you to share highlights by your favorite social media platform using the hashtag RWDG. And to find the chat and the Q&A panels, you may click those icons at the bottom of your screen to activate those features. And as always, we will send a follow-up email within two business days containing links to the slides, the recording of the session, and any additional information requested throughout the webinar. Now, let me turn it over to Michael for a brief word from our sponsor, Elation. Michael, hello and welcome. Hello, Shan. Thanks for having me. Just a quick check. Can you see my screen? OK. Yeah, looks good. You sound good. OK, great. Well, thanks for having us, Elation, here today. So my name is Michael Meyer, and I'm a technical product marketing manager for Elation. And so today, I just wanted to kind of quickly kind of tee up the subject a little bit about, you know, every every company's goal really is to become data driven. And we all think about the, you know, the three key pillars that people process and technology. And, you know, for those that are already out there and have maybe you might be in the process of looking for a data catalog or maybe already purchased. And so one of the things that's it's always really interesting to me is that that technology piece can sometimes be the easiest part. And what we start to find out is when we start to talk with that with people that are embarking on this journey, the questions come up of what do I do? You know, where do I start? How do I get to success? How do I measure that success? And how do I build to get to that next milestone? So in Elation, we came up with what we call our active data governance methodology. And really it's an autonomous continuous improvement methodology guiding organizations to help create that data driven culture through data governance. So quickly kind of show you a little bit about what it looks like. And so basically, we will often refer to this as our governance wheel. And you know, the first part is really having a group established around a governance group established to get going. And in order to get started, one of the things you always like to to mention is that it's best if you pick a use case that you know that we'll have positive business outcomes with it, something the business is looking for instead of trying to go out and just grab all of your your data and start trying to catalog it within your organization. So as you start to take that use case through that you've actually picked out, the first thing is really about populating that catalog. And this starts out with ingestion technical metadata. From there, it's about being able to empower data stewards. So recognize folks in your organizations that have knowledge around the data and really tell them how important it is for them to be able to share so that others can understand the data too. From there, it becomes about curating assets such as doing things such as documenting and providing better descriptions of this information and also things like creating glossaries in terms so people can understand. So when you think about it, really about coming data driven, it's about having people be able to find, understand, and trust their data. So if we look at this first side of the wheel, it's really about that finding and understanding we've built out. Now, in order to trust the data, you really want to be able to have your governance policies and controls in place and guide users to the right and proper use of data. From there, then it's about driving community collaboration, going out and promoting with others in your organization and showing them about the trusted data that they can now use from the catalog. And again, being able to leverage, this is an opportunity to meet others too that may have knowledge that they want to share within the catalog. Last but not least is monitor and measure. And really what this boils down to is take time to promote the successes and the positive business outcomes that came through by using this information that you gathered. From there, then also typically you're gonna look at maybe some dashboards you have set up to see how overall how the program's doing. Last but not least, do a retrospective of kind of what went well and what didn't go so well and make adjustments and then get to that next business case to work on. So what does this look like in terms of kind of in action out in the field? So as you as an organization have databases, data lakes, file systems, BI systems. And so that first step is really being able to have a system that can automatically go out and extract and keep up to date that physical metadata coming into for a catalog. The next thing is really important is to have some human or to have basically human guided machine learning going on. What do I mean by that? So really you're looking for AI and ML to be working in the backgrounds to do things like auto suggest titles and glossary terms, but you really want a human to be able to basically approve those suggestions. This way you get a basically a model that gets better and better over time is specific to your organization's needs. And so from there, then in the last part really through the technical metadata has been able to go through those relational databases query logs and to be able to comb through and find out who are my top users of data in my organization because those could be definitely key candidates for stewards. What items are most popular? Where are my popular tables and schemas that people are using? And that can help with being able to go through and prioritize from a stewards perspective of what they need to work on. And then other things about more details about queries such as joins and filters, understanding the relationships of other data that it's being used with. And last but not least is really being able to see that full flow of data from the source systems all the way up through into your dashboards or something like that, having that full lineage. So now that we have this process going and we have this technical metadata coming in, now it's about going in and really being able to have the curation process go. So the stewards that are assigned to this can start curating. Again, that curation is the process that helps people to understand the data. And so now data consumers can start to be able to find it and then also to use this in a self-service manner to be able to find the data they're looking for. And then last but not least, again, through this process as we show is about our governance team then continually monitoring and then providing the next business cases to keep going. So these are just some things to think about in terms of also about your people and process too when embarking on a data-driven journey. So with that, I'll turn it back over to Shannon. Michael, thank you so much for this great presentation and for kicking us off. And thanks to Alation for sponsoring today's webinar and helping make these webinars happen. If you have any questions for Michael, feel free to submit them in the Q&A panel as he will likewise be joining us for the Q&A portion of the webinar at the end. Now let me introduce to you our speaker for the series, Bob Sinner. Bob is the president and principal of KIK Consulting and Educational Services and the publisher of the data administration newsletter T-Dan.com. Bob specializes in non-invasive data governance, data stewardship, and metadata management solutions. And with that, I will give the floor to Bob to start his presentation. Hello and welcome. Hi, Shannon. Thank you, Shannon. Thank you, Mike, for a great presentation. A lot of things that you talked about really play very well into what we're gonna talk about today because all those phases, all to become data-driven to be able to utilize the tool, to utilize the process. All those things need to be governed. They're not gonna govern themselves. So again, thank you everybody for taking time out of your schedule. Whether you're seeing this live or you're looking at the recording of this, good to have you here. So again, my name is Bob Sinner. I'm gonna be talking to you about how to govern glossaries, dictionaries, and data catalogs. A little bit about myself before I get started here. As you know, we do this monthly webinar. We've been doing it for many years. On the third Thursday of the month, next month we'll be talking about one data governance for all. And that includes master data. I'm gonna be speaking at the Enterprise Data World event coming up in Anaheim in September. And I think there's some deadlines that are coming up for early bird registration for that. So you might wanna take a look at that. Also, I've written a couple of books now on non-invasive data governance. The first book, Non-Invasive Data Governance, The Path of Least Resistance and Greatest Success came out in 2014. And since it is now 2023, I decided I would create a second book that has really lessons learned and perspective games since the first book came out. So the second book is called Non-Invasive Data Governance Strikes Again. I have online learning plans that are available through Dataiversity on non-invasive data governance, non-invasive metadata governance, glossaries, dictionaries, and catalogs, similar subject to what we're talking about today. My consulting business is KIK Consulting. Shannon talked about TVN.com. And when I'm not doing all of these things when I have time, I'm also a faculty member at Carnegie Mellon University here in my hometown of Pittsburgh, Pennsylvania in their chief data officer executive education program. So what we're gonna talk about today, I've broken it into five categories. We're gonna talk specifically about how to govern the glossaries, how to govern the dictionaries and the data catalogs. I've broken it into these topics because I think it's a really logical and in a well-ordered way to be able to start talking about this subject. So first we're gonna talk about populating what it really takes to populate the business glossary, the data dictionary, the data catalog, and why it is so important that there's actually stewards and there's people in the organization that have accountability for the metadata and for the tools themselves. So we'll talk about populating the glossary, dictionary, and catalog. We'll talk about what does it mean to actually govern a tool and then to govern the metadata. Everybody talks about data governance. They don't talk as much about metadata governance. I think we need to change that a little bit because as you'll probably hear me say several times in this session, the metadata is not gonna govern itself. The tools aren't gonna govern itself. Talk about formalizing accountability for metadata. What is the impact of having a tool that's ungoverned and metadata that's ungoverned and the people process and technology that Mike talked about being ungoverned, gaining consistent value from the glossary, dictionary, and catalog is how we'll wrap it up today. So before I typically get started, I like to begin with a set of definitions that I use all the time. So if this is repeat for you, I'm sorry, but I think it's really pertinent to today's conversation, especially when we get to the metadata definitions. So I work my definition of data governance very strongly. The execution and enforcement of authority scares the heck out of some people. You might wanna take them that down, but the truth is at the end of the day, if you're gonna govern your data, if you're gonna govern your metadata or your tools, you're gonna need to execute and enforce authority over those things. Stewardship I refer to as formalized accountability for data, so people already have relationships to data. If they're being held formally accountable for those relationships, they're a steward. It's not something to opt into and to opt out of. You're a data steward if you are taking action with data and if you are being held accountable for it. In terms of metadata, yeah, it's data about data, but it's typically that data that improves both business and technical understanding of the data. And yes, there is such a thing as metadata governance and metadata stewardship. We need to execute and enforce authority over metadata, over our metadata tools. That's what we're talking about today, but we also have to recognize that there's people in the organization who are metadata stewards, people that need to be held. Again, the metadata is not going to define, produce and use itself. We need to help to get people to be held formally accountable for their relationship to the metadata in the same way that we're getting people to be held accountable for their relationship to the data. So the first topic I really wanna address is let's talk about populating these tools, populating the glossary, populating dictionary, the catalog. And we gotta ask ourselves a couple of questions. We gotta dive into this a little bit deeper. Where are we populating these tools from? Does that metadata already exist somewhere? Is there somebody who has the responsibility for selecting the appropriate metadata to put into the tool? The metadata that's actually gonna be used by people and get value out of using the tool. We'll talk about how to go about selecting the appropriate metadata. We'll talk about sourcing the metadata and the quality of the metadata that's in the source. I know I talk a lot about cheeseburger definitions. I'll talk about it again when we talk about the source, metadata quality, getting the metadata into the tool. And one of the things that's really important and you can use the tools to do that and that is to rationalize the metadata within the tool. And I'll explain to you what I mean by rationalizing the metadata. So when we first start talking about populating metadata into our glossary, into our dictionary and our tool, we gotta realize or we gotta take a look at where is there existing metadata within the organization. So we've gotta know where our metadata is going to come from if it's going to be entered directly into the tool or actually if it's going to be a byproduct of some other discipline that we're working on, a data modeling discipline or a data integration discipline. Oftentimes the metadata about those things it's inherent within the tools and the products that you use in your environment. So if you're doing modeling, you've got modeling metadata. If you're doing data movement, you've got data movement metadata. The question becomes, is that important to somebody? Who is it important to? And what is it going to take to get it from the native source that it's in now into a place where people can actually make use of it? So like I said, there's metadata that's a byproduct of things that you're presently doing, but then there's metadata as an intentional effort. And what I mean by that is, I don't know how many organizations start their business glossaries and their data dictionaries within a data catalog tool. They may do that, that's a good place to start if you've already got one of those tools. But in the meantime, you're probably keeping this in a spreadsheet or in a document. So it's an intentional effort that you have business stewards and technical stewards and data stewards who are providing definitions. So it becomes a very intentional effort to make certain that we're documenting our definitions, documenting our business roles. So we need to, again, take a look at, what are the native sources of this metadata? Is it a tool that we're just creating and playing around with ourselves on our desktop or is it just part of another process or another aspect of the data management discipline? And then what's it going to take to move that metadata from one place to another? So I know before you jump all over these diagrams and say that they're too small, I will blow them up here in one second. But I wanna talk to you about some of the things that are really important about selecting that appropriate metadata to govern. And actually that process of looking at all the metadata that you have in your organization and narrowing it down to the metadata that's going to add the most value to people, that's a process in itself and that needs to be governed. You know, we're very used to people saying, give me all the data and I'll do with it what I need. No, no, no, don't give me all the metadata, give me the metadata that's gonna add the most value to me and to my business community and to business outcomes within my organization. And then so we need to select, what's the appropriate metadata to cover for a business glossary? Go ask generative AI or go ask, do a search on Google to what are the appropriate pieces of data to be collected within a business glossary or a data dictionary or even within a data catalog in general and then narrow that down, work with your business community, work with your technical community to see what metadata is going to be used and then focus specifically on the quality, the availability of that metadata that you know people are going to get some use from. So look to see what metadata specifically needs to be governed within the glossary itself, the dictionary, the catalog. There's other metadata, there's lineage metadata, there's confidence metadata, there's visualization metadata. Just it's a process that it's not gonna take place on its own, somebody has to have the accountability for selecting the appropriate metadata to govern. And typically I talk about there being three levels of metadata by looks at the diagram when I share it with you in a minute, you'll think that there's actually four levels. Well, let me share it with you right now so you can see. The top level is not really what we're talking about in today's session. So the three levels of metadata that I share with my clients as being a very good place to start with especially within your catalog is the metadata that's associated with your glossary is the metadata that's associated with your dictionary and your catalog. And if you look at the verbiage along the lines that connect data dictionary to business glossary and data catalog to data dictionary, you can see that for every term that you have in your organization, there's gonna be a lot of different data that is associated with that term. And the funny thing is that that data is gonna reside in multiple systems. And if you're gonna make data discoverable for people, they're gonna most likely come in through that top level, the domains, find the terms that they're looking for, the data that's related to it. And then the next step is to actually get them into the data resources themselves. So again, the glossary, the dictionary and the catalog, when we're selecting the appropriate metadata to govern, we should be looking at those tools that we're gonna be providing to our business community and see what's the true value? Where are they gonna get the most value from this metadata? So when I talk about rationalization, that's basically making sense of your data, connecting the dots between your data. So you have a column named ABC column name in one system and XYZ column name in another system or maybe they're named something that doesn't obviously connect them together. You wanna know how your data is connected together. So is this rationalization metadata, the metadata that lets people say, oh, we call it this over here, what do we call it over there? How are they the same? How are they different? When we run a report and we get different numbers because we're using different data that is supposedly the same thing, people wanna know why. And that's where the rationalization of the metadata comes in. So when we're talking about selecting the appropriate metadata to govern, that's gonna be a piece of it, the rationalization metadata. Now, I could spend a lot of time talking about this diagram, but I'm not going to do that. It's just a way of taking that vertical and that horizontal diagram that I just shared with you and again, rationalizing the metadata so that somebody who is a business user can come in and see what the standard name for data is, see what it's called in system one, system two, what it's called on the different reports that you have in your organization. So when you're setting up, this is not gonna happen on its own. Again, when we're talking about governing glossaries, dictionaries and catalogs, the first thing we need to do is understand what metadata do we need and who is gonna have responsibility for that metadata. Okay, so now let's talk about the source metadata quality and some of the issues pertaining to it. Well, right now, most likely, at least within most organizations, there's not a lot of accountability. There's not a lot of formalized accountability for data quality, let alone metadata quality. So if you're going to govern the metadata and the tools, you're gonna have to have some level of formal accountability for metadata quality. And that includes, like I was saying before, the definition of what metadata we're going to capture, but also what's the definition of the metadata that we're capturing? When something says field name and that's the label we give to a piece of metadata, what does that mean? So again, there's definition of what metadata are we collecting and then there's the definition of that metadata. I think you'll all agree that somebody has to be responsible for producing the metadata. If it's not metadata that's inherent within another tool, somebody's gonna have to have responsibility for producing it. Obviously you want somebody to use it. I talked about cheeseburger definitions. Again, my definition of cheeseburger definitions, the definition of cheeseburger is a burger with cheese. The definition of a student account number is an account number for a student. You're just using the names, using the words of the term that you're describing in your definition. So if you want more descriptive definitions, somebody's gonna need to provide those. And as I say a lot, the metadata will not govern itself and the challenges that you have around metadata, they're not gonna self-correct unless somebody has the responsibility for making certain that somebody is tending to the goods, meaning the goods that are in the metadata. So getting the metadata into the tool, I can tell you from being a repository administrator for one of the larger health insurance companies in the US, earlier in my career that I spent a lot of my time building connectors and getting tools to be connected and getting them to be to stay connected. That's a big part of the governance of the catalog tour of the metadata repository to itself is building those connectors, keeping those connectors working. Let's say you're connecting to a data modeling tool and the data modeling tool comes out with a new release that you wanna go to. Well, is that connector gonna break? Somebody needs to be looking at that. So that's why I say that's the life of the repository administrator. And then there's when you buy tools from a single suite, there's an expectation that all these tools are gonna talk to each other. It doesn't always work out that way. So again, somebody needs to govern to make certain that even tools within a single suite are talking to each other. In a lot of organizations, they'll start out by having business glossaries and data dictionaries like I mentioned before in a spreadsheet or in a word processing document. Oftentimes the business glossaries and the data dictionaries stand alone. And in fact, I know a lot of organizations that do inventories of what data dictionaries and data business glossaries they have, they're in a ton of different formats, they're in different phases of disrepair or repair, so to speak, they're not consistent. So if you're gonna keep your business glossaries and your data dictionaries as important metadata tools that are not yet in your catalog, my suggestion is that you wanna be consistent in the way that you're collecting them, because at some point, you're gonna wanna get that metadata into your data dictionary tool. And it's gonna be a lot easier if you don't have 30 different formats to try to bring that metadata into your data dictionary tool. Again, somebody needs to govern that process and make certain that these things are being built in the first place. Somebody needs to govern the process to make certain that we can get that metadata out of these standalone tools, if that's where it is, into the hands of people that can use it. I talked about rationalizing the metadata. I did a search on some good rationalization cartoons and that's the one that, comics and that's one that came up for me. It's great because I have two daughters, one has a dog and one has a cat. So I think it's very relevant. But what is rationalizing the metadata? It's making sense of the metadata, making sense of the data, making sense of the business. I had a friend of mine in a presentation they gave recently, talk about the CCDO role. And the CCDO stands for the chief connector of the dots. And that's really what a data, a CDO does is connecting the dots. When you're rationalizing your metadata, you're connecting the dots between one system and another system between your glossary and your dictionary, between your glossary and the domain. And so leverage the tools, like the tools, the things that Elation has built into it and other tools, use their artificial intelligence in a machine learning capabilities to help you to rationalize your metadata. And of course you wanna automate it as much or wherever you can, just realize that it's not always 100% something that you can automate. There's going to be people that are gonna be accountable for that rationalization. Again, you have to govern the glossary, the dictionary and the catalog. So now let's talk about what it means to govern the tool and to govern the metadata. So if you remember back to, I think it was slide three in this webinar, I gave you the definition, maybe slide four or five, the definitions of some key terms, data governance, the execution and the enforcement of authority. Think about the execution and enforcement of authority in terms of your metadata. Think of it in terms of your tool. Think about formalizing accountability, which was my definition for data stewardship. There needs to be formalized accountability for data governance, for data, for the metadata, for the use of the tool. So we're gonna look at it from the tool perspective, from the data perspective. And again, as I always say, the metadata it's not gonna magically govern itself. It requires that people are accountable for making certain that you have good quality metadata that's specific to what the business community needs and that's gonna address their concerns. And the tool will not govern itself as well. So what do I mean by execution and enforcement of authority? As I said, that's my definition of data governance. It means that somebody, it means that there's rules, that there's processes, that there's guidelines and that somebody is responsible for that. So when I talk about executing and enforcing authority, that authority is in those guidelines and in those processes and in those things that you're defining. They're also in the production. I mean, we need to execute and enforce authority that people don't hand you a data model or a data dictionary with a whole bunch of cheeseburger definitions. Now that you know what cheeseburger definition is, so make certain that you execute and enforce authority and you recognize and you reward those people that are doing a good job of effectively producing the metadata for the organization. Executing and enforcing authority for metadata and tool usage. Now making certain that this metadata that we're spending a lot of time getting into chip shot, tip top shape is gonna be utilized by people that they understand how it will help them to do their job more efficiently and more effectively. And again, you're probably gonna get tired of hearing me say it, but I said the metadata and the tool will not govern themselves. So I'll share with you examples, especially about the tool. If the tool, what happens when you take a tool and you don't govern it? What state of disrepair does that become in? I'll bet you Mike could talk about that too as well. We could talk about that in the Q&A at the end where there needs to be governance applied to the catalog tool. There also needs to be formalized accountability. And that is you need to recognize who your metadata stewards are. Who are the people that have formal accountability for defining what metadata you're gonna collect, defining the metadata itself, producing the metadata and using the metadata. So again, the idea of being non-invasive in your approach, and I'm not talking a whole lot about how to be non-invasive in this session here, but to be non-invasive instead of assigning people to be metadata stewards, look to see if there's people that already have the responsibility for that data documentation and help to formalize that accountability. Again, recognize people that are already doing it rather than assigning people. It's a lot more invasive. I know I feel like it's something that's over and above what I'm presently doing if I'm assigned something new. So recognize your metadata stewards. Evaluate your metadata stewards based on how well they execute the plan, how well they execute what they're supposed to do in terms of providing metadata in enforcing authority. So build it into people's jobs or build it into their job descriptions, gamify it, do whatever you need to do to formalize accountability for the tool, formalize accountability for the metadata. And notice I said recognize. I don't use the word assign. I don't like the word identify. Recognize has a positive connotation that comes along with it. So if we can recognize stewards and help them to do a better job, we still need to make certain that we're governing the tool, governing the definition of the tool, even for those organizations that don't have a tool yet, you have to govern the process of evaluating what are your requirements, looking at the appropriate tools in the environment, somebody has to manage that process. For the production of the tool, all the technical aspects of the tool, for the usage of the tool, for the support of the tool, without these things, it's gonna be very difficult to sustain your data catalog and your metadata environment within your organization. So then there's the metadata that I talked about before, the metadata for the accountability for defining the metadata, producing it, using it and supporting the metadata that we're putting in our tool. And as I said, the metadata will not govern itself. I say it a lot. That's because it's true. The tools won't govern themselves. In reality, the people within your organization, they're probably right now not used to having metadata and tools. So this is gonna be new to them. So we need to explain to them how it's gonna be useful, make it easy for them to get to the metadata and recognize that any challenges that you're presently having around the tools and around your data, they're not gonna correct themselves. People need to be guided. And I was just talking to applying about this earlier today, the more that we can build the metadata, the data documentation into people's project plans and take the burden off of the data governance and the metadata team and put more of that burden for, as I was saying, blocking and tackling in terms of metadata and the tools into the hands of the projects, the more likely we're gonna be successful within our organization, especially if we can build it into our project management methodology. So now I say the tool's not gonna govern itself and that includes the glossaries and the dictionaries. Somebody needs to find the requirements for those tools, assess the current landscape. Do we already have these tools, manage the evaluation? Again, I don't wanna read through the entire list. I've already read through most of it, but look at the last bullet there, because that's extremely important. Metadata tools in your environment, if you are not prepared with somebody who is going to install the tool, configure it, load it, monitor it, maintain it, support it, and I'm talking about the tool, not the metadata itself. The chances are you're gonna end up having problems with the tools. So there's a lot of responsibility that goes along with governing the tool itself beyond just the metadata that's going to go into the tool. So let's talk about formalizing accountability for metadata. What is meant by formalizing accountability? What is meant by formalizing accountability for metadata definition, production and usage, and then for the governance of the tool set itself? So what does it mean? And I get asked this question a lot. What does it mean to put formalized accountability into place? And this is a good definition of taking those words and expressing what you mean by formalized accountability. It's having systems, it's having processes. It's basically like Mike was talking about, people process and technology for holding people responsible for their actions. So that's what formalized accountability means. And in order to do that, we need to define the roles and responsibilities. We need to define expectations. We need to also monitor how well they're following those formalized accountabilities. Oftentimes we'll do that through the development of our framework or policies or procedures that you're going to use to guide decision-making and just make certain that you get other people who that you're working with to participate in these activities, to provide feedback, promote transparency. So where if something's not working, give them the ability to come to you and let you know it's not working. You know, promote integrity of the metadata. You can tell your stories about how metadata has not been kept up to date. And it's difficult enough to get somebody to come to your data catalog the first time. But if the metadata and the tools not up to date and the integrity of that metadata isn't high, it's going to be very difficult to get them to come back the next time. Ensure answerability, issue identification, improved performance, all of these things are ways to define what I mean by formalized accountability. So when it comes to formalized accountability for metadata definition, these are people that have the responsibility for doing these things that you see on the screen for ensuring clear and consistent understanding, the data integration, data standardization, regulatory adherence. So the people that are formally responsible for the metadata formally accountable or who are the stewards of metadata definition, this is a list of typical activities that they would be responsible for. The fact is it's not really a new job for a lot of these people, the ones that are responsible for creating your glossaries and your dictionaries, they want these things to be true. So they're going to, you would hope that they would actively play a role as the stewards of the metadata definition. And then there's accountability for the metadata production. And I don't want to go through these in too much detail, but you could see accountability for metadata usage. You know, if people want to interpret the data, right? If people want to collaborate around the data, they want to enhance their ability to analyze the data or make certain that you're complying to all of the regulatory needs of your organization, somebody has to be accountable for, they're obviously accountable for the data usage. If we can help them to do it better, do it more efficiently and effectively through providing them valuable metadata, somebody has to be responsible for getting that metadata into their hands when they're doing their data interpretation, when they're doing collaboration analysis, all of these types of things. So again, formalizing accountability for the metadata itself through the definition production and usage of the metadata is a big part of governing glossaries, dictionaries and catalogs. Because again, the metadata is not going to improve on its own. It's not going to govern itself. You need to have people who have those responsibilities. And the last one is accountability for the governance of the metadata tool set. So if you're going to follow a centralized metadata management or you want to improve data discoverability or enhance data lineage, I hope some of these sound like your organizations because that's where organizations are looking to get value from their data catalogs. Again, expect the metadata that's going to be necessary to improve the data discoverability. Again, remember those three levels or actually four levels, including the domain that will help to make the data more discoverable to people. Somebody has to govern that metadata. Enhancing the data lineage traceability, it may be inherent in another tool, but people may not have access to that tool. And in fact, if we bring it over into our data catalog, are they going to know how to find it in the tool? Again, enhancing your capabilities around these areas, these all go into the accountability that's necessary for governance of a metadata tool set itself. So now let's talk about the impact of ungoverned tools and metadata. I haven't found ways to, I want to have a sizzling sound and a big boom as the impact of the asteroid coming down. But I want to talk about what does ungoverned metadata look like? What does ungoverned management of the tool look like? How do we build out confidence from building better understanding and context for our data? How can we get people to use the tool? What are the criteria of people who use the tool if these three things are four things or five things? Let's talk about these things. So what does ungoverned metadata look like? Well, it looks like there's inconsistency in the naming of the data or of the metadata. There's inconsistency in the definitions. There's lack of standardization. Again, don't want to read through this whole list. I hope you'll go back to this webinar and to these slides and consider looking at these when you're saying, well, what does, if we're actually going to govern our data, how is it gonna make our metadata or govern our metadata? How is it gonna actually look different from ungoverned metadata? I know of a lot of organizations that are asking that question, what does ungoverned data look like? A lot of times it ends up being their present state. And then what does governed metadata look like or governed data look like? That's the future state. That's the ideal state that they're looking to attain. So use these kind of quick descriptions of what ungoverned metadata looks like. It's insufficient documentation and lineage, inconsistent data security and access controls, lack of ownership. Well, you know, it's great to record the ownership, but if people don't know who owns the data or who owns the process, that metadata is not really adding value to a lot of people. Here's a good slide too. I love this slide. I love the image that goes with it. What does ungoverned management of a tool look like? And this could apply to your glossary, to your dictionary, to your catalog. It means you've got inconsistent categorization of the data and the metadata within your tools, lack of standardization. It means that some of the metadata might be missing within the tool or there's limited capabilities, whether those would be search capabilities or data discoverability capabilities. I think data discoverability in the data marketplace. I know I've heard Elation and other vendors talk a lot about creating that data marketplace. What is a marketplace? It's a place where we can go in and we can find things that we need. Well, it's a marketplace for data. You want to make your metadata, you want to make your data more discoverable and how are you going to do that through the providing data documentation and metadata? So what does ungoverned management of the tool look like? Insufficient security on that tool, absence of ownership and accountability for keeping the tool up to date and keeping the tool active and training new people on how to use it, all of those things. That's what ungoverned use of a data tool or a metadata tool looks like. And when it comes to the confidence from understanding the tool, this is where a lot of people are having problems with the data. They see the data as being ambiguous. They don't have trust in the accuracy of the data. Well, you want them to gain that same level of confidence in the metadata that you're providing to them. And again, that's not going to happen on its own. There need to be stewards and people that have this specific responsibility for the metadata or it's going to become a bunch of cheeseburger definitions and it might be there or it might not be there depending on what metadata people are looking for. So if you want to improve people's confidence in the data the confidence comes from the understanding and the context that's built into the data. And we want to make certain that as we are looking to govern our tool and our metadata that we're focusing on eliminating the ambiguity and improving the data accuracy, improving the way people have their access to be able to discover and explore the data of the organization and those types of things. So again, these tools are not going to govern themselves. We need to make certain that we are focusing on formalized accountability and executing and enforcing authority over the metadata and over the tools themselves. And that's the glossary, the dictionary and the data catalog. So people are going to use your tool if it, and here's just a handful of things that I've experienced. If it's you, if there's a user friendly interface I had a friend talking to me about a tool that they were using kind of an old fashioned tool where the company is never really focused on the UI the user interface. And it's not really usable to individuals or not really searchable in the way it needs to be that people are going to use your tool if it has a user friendly interface. If the information, as I mentioned earlier in the webinar is relevant and it's contextual the way that it's going to help them in business outcomes and in their job. So asking them what metadata is important in focusing on that specific metadata and the quality of that metadata and the governance and the stewardship of that metadata might be a good place to start because you're looking to get the relevant contextual information into their hands. Robust search and discover capabilities, integration with existing tools, training and support. That's people are going to use the tool if they understand it, if they've seen it if they've lived through it if they have the training and the education to use the tool and understand the impact and the value it could have to them in their part of the organization. And here I'm not going to spend too much time on this but again, I say the metadata and the tool will not govern themselves. So if you're taking one thing out of this session it's that we need to find people who are accountable for these things within our organization. So the metadata, the metadata tools require active governance. It's essential for organizing and maintaining effective metadata without governance as I showed you on some of the previous slides the metadata and the tool itself can become somewhat chaotic and unmanageable. So last thing I want to talk to you before kicking it back to Shannon is how can we take the governed glossaries, dictionaries and data catalogs and how can we start gaining consistent value across the organization by implementing tools that now have metadata in them that is well-defined as well-produced as well-used people are trained on it. Well, here's some other ways that you can do it by standardizing terminology and understanding improving quality. I'm going to walk through each of these five before I kick it back to Shannon here real quick. So by standardizing the terminology and the understanding the way to be able to gain that consistent value from these tools is through communication and collaboration. Don't try to enter in all the metadata yourself collaborate with other people within your environment activate them, activate some metadata stewards who are not in your data management group or not in your data governance group activate the stewards themselves and get them collaborating and communicating your data integration data interoperability all of these things you can standardize your terminology and understanding if you focus on taking care of these items and that's really important is standardizing your terminology and understanding because people are going to be looking to the metadata once they have access to that as a resource to help them to understand and get more value from the data by improving the data quality and the data accuracy. Again, you're going to gain consistent value from these tools if these tools by setting standards and putting those standards into your data dictionary now you have something to compare the quality of data to to do profiling against to do cleansing against validation standardization. Again, there's a lot of ways to gain consistent value from your glossary dictionary and catalog tool just a handful of ways to put things for you to consider as you start moving your governance of your tools forward and your governance of your data forward by providing efficient and effective data discovery and access we talked about how important that is to implement the catalog provide search and filtering capabilities with appropriate access controls user friendly is really important. Self-service analytics is a direction that many organizations are heading towards by strengthening compliance and risk management Again, another use of the catalog start documenting these things within your tool and get people to help you to do that get other people within your organization to become the stewards and the active governors of the metadata within your organization by increasing collaboration and data literacy again, these are just consistent ways or these are ways for you to gain consistent value from your business glossary or data dictionary in your data catalog tool. So for the last 40 minutes or so I've been talking about these things with the idea of focusing on governing glossaries dictionaries and catalogs we talked about populating the glossary the dictionary, the catalog what it means to actually govern the metadata that's going into those tools and the tool itself what it means to formalize accountability for metadata what happens if we don't govern our metadata or govern our metadata tools and then I shared a couple of ideas with you for how you can gain consistent value from your glossary dictionary and catalog and with that Shannon, I am going to kick it back to you to see if we have any questions. Thank you so much for another great presentation and just to answer the most commonly asked question just a reminder, I will send a follow-up email by end of day Monday for this webinar with links to the slides, links to the recording and many of the things that Bob showed throughout and anything else you all asked for so many great questions coming in like they just right out of the gate we had lots of questions and I want to start with a couple of questions that came in, Michael when you were presenting once some tech metadata has been ingested but nothing has been curated yet can for an inhalation can those assets be hidden from general view until some curation has been done? Yeah, absolutely so there's a process in which you can go about making things private until you're ready to make them public so to speak so that you can be working on them in the background to do that. Perfect and what's the impact of having a business glossary with no standards? You know, one of the things that I always think about is like having a metrics like definitions or things like that inside of a business glossary and if it's not being governed then it gets to be to the point where I can remember a couple in an organization I was working with where we had a couple of different ones where people had different opinions and so you really need to work that out as an organization to make sure you have a good definition for that that everybody can align to otherwise it just continues to make chaos with things like that. And I was gonna add one thing to that too if you don't have standards within your glossary you may be collecting different information different people may be collecting different information the way that they're collecting that information may be different so yeah, it makes sense for all the things that you just said, Michael and somebody has to be looking at those things. You know, and this is starting to come up a lot so how will generative AI impact metadata management? And Elation are you gonna be enhancing your catalog to incorporate generative AI? Is that in the pipeline? So I don't know, Bob do you want me to take that one or do you wanna go with that one? I mean, I can't answer the question as to whether or not Elation is addressing generative AI but yeah, you could answer that part. But I'll just real quickly say that it's gonna impact everything that everybody does because it is, you know, organizations are gonna probably want to adopt it faster than they're gonna be able to adopt it. And when they do adopt it they're gonna have to make darn sure that the metadata that's going to be available through large language models and generative AI that it's governed. I wrote about it in my new book that there's the data governance challenges associated with large language models. So yeah, we need to be aware of it and we need to incorporate it into things that we do. So now I'll shut up and let you answer, Mike. Yeah, so from the perspective of Elation, you know, we've actually had AI and ML in the product since it's been out for 10 years but with the new generative stuff I think the premise around it is being able to make suggestions but yet human approved for a while because you don't want something automatically taking over with nobody ever approving it. And that's where I think you can probably run into some problems of not having the right context around things like descriptions of information inside of your catalog. And what the usage of generative AI has to be governed. Absolutely. Because again, you know, I have used the tool for different things but it doesn't give you the answer and it may get you somewhere where you're trying to get quicker than you would have without the help but it has, you're exactly right. There has to be, the human element has to be included in it. And that's a lot of, again, what we talked about today with the metadata and the catalog and the glossary and the dictionary, you know, they have to, the process it has itself of using a generative AI in the organization has to be governed. I love it. It's such a hot topic right now. So moving on though, can you talk about how you have handled the challenges inherent in governing language? For example, we have established governance of our data glossary. This does not prevent executives from, you know, using business terms incorrectly. This, it doesn't prevent users from creating new terms that are vague and undefined. You know, how do you handle that? Well, it doesn't happen overnight, number one. And number two is you do it incrementally. I mean, if people are, people need to, if there is, if I'm going by what Chuck Knoll, the old coach of the Pittsburgh Steelers said, the standard is the standard, right? So if you have a standard that's been approved in your organization and people continue to go outside that standard, and that could be a standard glossary term or business term or anything, they're, they have to be educated on why we're trying to avoid that. So like I said, it's not gonna happen quickly. It's gonna happen through governance. But unless there's some, because the executives, you know, why can they do it? Why can they use terms that are on in our glossary? Well, the answer to that question is because they can. Because it's something that's within their capability to do and they're gonna use language that's most specific to that's expressing what they're trying to express. Question I might have is, why is their language different than the language that you have in your glossary? I don't know, Michael, what do you think of that? No, I think that's spot on. And you know, and the thing about that is, is again, this becomes an item in your organization just even having a little bit of light workflow around and having some approvers of that and being able to collaborate and make sure that you are getting definitions that resonate with everybody on terms. Because again, that's the goal is that everybody in the organization understands and can speak to that same language in terms of your glossary and items like that. And common business language will always be a challenge within organizations unless they govern how they use it. Perfect. So, we've got about seven minutes left. I'm gonna try and slip in as many questions as we have here, but keep them coming. Bob will write up any answers to questions we don't have time to get to here. I mean, so many great questions here. So do you have specific suggestions for measures, metrics, KPIs for how to demonstrate the value in concrete terms for metadata management? Michael, do you wanna take that one first so I could think about it? Yeah, so what was the last part of that for something of data management? I kinda lost the last part. Yeah, for metadata management. Oh, for metadata management. So a lot of the things that I've done in the past is I've actually, as you basically make the metadata available, I start to look at it from the perspective of who's consuming it. So if I can actually see that people are actually finding it and consuming it and driving value from that, that's one of the measurement things that I typically look at. It's the kind of putting that business value to it of how it's being used. I guess that's a good starting point in my thought. Yeah, you know what? I think that aligns with the whole idea of definition, production and usage. You could put quality measures and metrics in place around each of those three actions. So how well is, what metadata have we defined the appropriate metadata for the organization? You can actually measure that. Have we started to collect that information and produce that metadata? Is the, how many systems or what portion of the organization, how many critical data elements do we now have definitions for? There's a lot of different ways. A lot of it really depends on specifically how you're using the tool. But I think that the consumption is a very good way because that's the ultimate end game, right? We want people to use the tool. And if we can measure how well they're using the tool and show that it's on an upturn, that's gonna demonstrate value. And just to take one little more side note on top of that. And if you think about it too, it's like how is that used in other things? So like if your catalog has articles or documents in which you're even further explaining information and you're actually linking those into it, again, it just provides another way of looking at the, again, elevated value of having that metadata. Yep, yep. Perfect. So why is there so much confusion between terminologies, data governance and data management? And I know Bob, I think we've done a whole webinar on this. It is confusing. How much time do you have? Actually, I would refer people back to the May webinar because I thought that there was a lot of great feedback from that webinar on where data governance and data management overlap. And in fact, I had a conversation with a client specifically about that with people in their office of data management and in their office of data governance which they don't really have yet. And so the way I view it is that data governance is more people and behavioral focused while data management is more operational and technical focused. Actually, the technical focus of it might be more IT oriented. But that's how I typically view it. And I'd be curious to see what people say in the chat how they define, how they define them differently. But the way I use it is basically operational, data management, people and behavior, data governance. And then they obviously then overlap. So Michael, I'd be curious as to what you think. Yeah, that resonates a lot with me too. I mean, I think about in terms of like, from data governance, I think of like you said it's more from the people side and the things of providing information and curating in those business context when you have that metadata around so people can understand it. And definitely on the data management side, it's more from that IT perspective of providing all of the facilities that you need around things like all your data structures whether it's privacy, MDM, all the different things aligned to there. So I think that's a good, the two really are married together. I think it's when you think about it, but yeah, I'd say business for the governance and more data management is more IT driven. And if you look at the demo wheel as compared to the elation wheel that you shared earlier, the demo wheel incorporates a whole bunch of different disciplines associated with data management, but they also incorporate data governance smack in the middle of that wheel, right? So they're incorporating data governance as being, they're the ones who have really defined data governance as being a part of data management. It doesn't always work out that way in organizations, but at least that's one way to look at it. Data governance is the people in the behavioral aspect of managing your data. Yeah, as being a, I was a data management data engineer or engineer for a while, and I would agree with that 100%. It was like governance kind of helped us and gave us what they were looking for. And we helped to make sure that we're implementing the things that needed to in order to, you know, from not from a compliance, but also for people to be able to find and use the data. Yeah, yeah. All right, we've got less than two minutes here. So as the elevator pitches on this next question, see if we can slip it in as fast as possible. If we need to, if we decide to take one Power BI report as a high value use case for my first implementation, what from that am I supposed to catalog? Yeah, so from that, first of all, you would want to be able to catalog that asset. The second thing you'd want to be able to have is the underlying data sources. So what databases, tables and columns make that up because at the end of being able to extract at least the technical metadata, you should be able to start to see lineage flowing from those assets, those table assets all the way to that reporting asset. So that gives you the A starting point. I say those are the great two places to start but then also the data that's on the report itself, the definitions of that data, the terminology of the data, how the data was calculated. There's a lot of specific metadata about any given report. And I think that starting with it as you have to document it as an asset. I love that you said that Michael because I think that's very important in itself that that report itself even exists. And then there's the source and then there's the data on it, on the report. It's a great way to focus a use case. Yeah, and if possible, find integrations to where you can make that definition once, hopefully within your catalog and then it can surface through your, whatever your BI tool is. I mean, that will save tons of time. Well, thank you both so much for this great presentation. And thanks for Elation for sponsoring today's webinar and helping make these webinars happen. Michael, it's been a pleasure as always to have you with us. And thanks to everybody who's been so engaged in everything we do. We just really appreciate all the questions. Again, we'll get answers to the remaining questions that have been coming in. And I will send a follow-up email by end of day Monday with links to the slides, links to the recording and all the other things requested throughout. I'm so excited. I hope to see so many people. As Cynthia, I can't wait to see you and Bob. I'll see you at EDW in Anaheim in September. Super early bird ends tomorrow. So don't forget to sign up early. Okay. Thanks, Shannon. Thanks, Shannon. Thanks everybody.