 Hello and welcome my name is Shannon Kemp and I'm the Chief Digital Officer for Data Diversity. We would like to thank you for joining today's Data Diversity webinar, the importance of metadata sponsored today by top quadrant. It is the latest installment in a monthly series called Data Ed Online with Dr. Peter Akin. Just a couple of points to get us started due to the large number of people that attend these sessions you will be muted during the webinar. For questions we will be collecting them by the Q&A section or if you'd like to tweet we encourage highlights or questions via Twitter using hashtag data Ed. And if you'd like to chat with us or with each other we certainly encourage you to do so. And to open an access either the Q&A or the chat panels you may find those icons in the bottom middle of your screen for those features. And just to note the Zoom chat defaults to send to just the panelists who may actually change that to network with everyone. To answer the most commonly asked questions as always we will send a follow-up email to all registrants within two business days containing links to the slides and yes we are recording and will likewise send a link of the recording of this session, as well as any additional information requested throughout the webinar. Now let me turn it over to Jesse for a brief word from our sponsor Top Quadrant. Jesse, hello and welcome. Thanks Shannon and hello everyone. Thank you for joining us today. I'm with Top Quadrant. And Peter's going to talk to you about all things metadata. What I want to do with you is share an approach, a unique approach for modeling, capturing and leveraging this metadata. Semantic graph standards and technology can be used to form what we call knowledge graphs, knowledge graphs of your interconnected, who, what, where, when, how and why. And all of those dimensions will make a lot more sense to you as Peter works through his slides, but the knowledge graph approach using semantic graphs is the unique capability of our top grade edge system. Top Quadrant is an enterprise data management company. We help large enterprises to better use, understand and employ their metadata. We serve some of the largest companies in the world with essential point solutions such as tagging social media and optimizing feeds and even up to very large data fabric deployment that cut across the entire topology of a business. Our success is about getting you live and solving real problems. The principle that we operate under is that enterprise knowledge graphs have a purpose. It's not enough to get our technology in your hands or even into the right hands. It's about using that technology, those knowledge graphs to leverage the metadata to make an impact. And we go on that journey from beginning to end. We call it our go live process. We partner closely with our customers to ensure that we can drive to real business outcomes. It starts with pairing you with experts to see the power of graphs to enable your organization to save money, save time, minimize risk and develop new products quickly and safely. Customer success due to our top rate edge system comes in many forms, but includes increased collaboration because of the workflow engine and the version control capabilities of the platform. Enhanced discoverability and faster reporting times, both of those because of the semantic very rich, meaningful representation of that metadata and being able to query it and even infer extra data from it. And even a decreased in storage costs because in many ways metadata weighs a lot less than data does. With that being said, what metadata do you have that could form rich knowledge graphs representing your enterprise and can you do it in a way that's highly leverageable and governable. That's what our top rate edge system is about. We stand behind semantic web standards and use those through and through. If you want to learn more about our platform or even the semantic web standards, you can always reach us at topquadrant.com. Everyone have a great day and we're back to you, Shannon and Peter. Jesse, thank you so much for this great presentation. You have questions for Jesse and about top quadrant feel free to submit them in the Q&A portion of your screen, as he will be joining us in the Q&A at the end of the presentation today. And now let me introduce to you our speaker for the webinar series Dr. Peter Akin. Peter is an acknowledged data management authority as associate professor at Virginia Commonwealth University, president of Damon International and associate director of MIT International National Society of Chief Data Officers. For more than 40 years, Peter has learned from working with hundreds of data management practices in more than 30 countries, including some of the world's most important. Among his 12 books are the first making the case for data leadership CDOs, the first focusing on data monetization and on modern strategic data thinking and the first objectively specific to first to objectively specify what it means to be data literate. So recognition has resulted from these and a pre-COVID-19 intensive worldwide event schedule. Peter has also hosted the longest running data management webinar series from data diversity, which is very true. We're in our 13th year, Peter, I think, starting before Google, before data was big, and before data science. Peter has founded several organizations that have helped more than 200 organizations leverage data specific savings have been measured at more than 1.5 billion US dollars. His latest is anything awesome. And with that, let me turn everything over to Peter to get his presentation started. Hello and welcome. Hi, Shannon, and I think our time is going a little too fast. We're probably on our 12th year instead of our 13th. It's been an absolute pleasure all along and thanks to everybody for joining us today and taking some time out of your schedule to spend about 90 minutes here learning about the importance of metadata. Right into the program. First of all, overview wise, we're going to talk about what metadata is in the context of data management. The two terms are interoperable and interdependent in that sense. Then there are four specific strategies that I'm encouraging you to adopt as you begin your metadata journey or continue on at the first one is to do not treat the word metadata as a noun. It's not clear. We'll hopefully make it up by the time we get to the end of this. Second is to make sure that your later governance groups are speaking exclusively in metadata. Third to treat repositories or glossaries as a capability, not as a technology. You definitely don't start your metadata process unless you first examined what available building blocks are out there that you can get a hold of notice. I've said many, many, many resources available to that we'll finish up about an hour from now with a little bit of benefits, applications and sources in there and then get to the part where we'll invite Jesse back on to talk about the takeaways Q's and A's and everything else that goes into this and I just want to again say how grateful I am for the community out here. This is a slide that I got back in response from Peter Campbell. Thank you, Peter for sharing this when I sent out the advert for this program which was a set of pile cabinets and Peter said back and just said just as an FYI. This is actually something that was done fairly early on around here where they were trying to classify the world's knowledge and this little repository here that you're looking at is in fact available online. And it's called the mundane, which stands for mundane sort of an interesting choice of words that they used around that but it does represent one of the earliest known and probably most successful efforts to manage metadata on a wide scale basis. Let's start off of course with the term metadata in language when we haven't figured out how to do words. Exactly right we stick a hyphen in between them so it was like well we need something meta and it's got to do with data so we'll call it meta dash data in there over time that hyphen gets lost and so now we have graduated Thank goodness to the word metadata all in one term although you'll see a little bit later on we didn't quite get it into the Dimbock in time in there, and just another little piece. There is actually somebody that owns a copyright on the term, but obviously it cannot be enforced as one of those oopsies that the patent and trade office occasionally makes an error on there. One of the main problems with metadata is that people don't understand the context within which it is used. And that is because they don't understand what it means to talk about data management so first of all the left hand side is the way people often think about it we're just going to go through all these books and reference them you know it's a great thing or if we label everything we'll get it all right or I'm not sure what Microsoft was attempting to show. This is an actual Microsoft advertisement here, where they're marking things up on a restroom door it doesn't say much about us but what I think I think we can all agree on is that data is not broadly or uniformly understood depends sort of where you start from it's like the blind people at the elephant different parts of the elephant come in contact with the blind individuals and so they end up with a different perception of what it is. And similarly in data it happens as well that you think data is largely about the point in which you've screwed to it. It's really, we have as an industry to continue to work on this and this is why I'm so happy we have such a vibrant community somebody will likely send me something here but we used to call data management everything between when data is sourced, and when it is used. That doesn't tell us much about the process so while it's probably a true statement, we need to refine it a little bit and the first thing that we really need to change in here is that it's not just about using data but about reusing data, because if we just use it. One time it's probably not as valuable or as leverageable as we'd like it to make. We're working on a newer definition for it that tries to encompass that sort of elephant in the room here as the idea that data is divided up into some preparation and some exploitation types of activities in here and I've just listed a couple of them across there. In addition to that of course we need to have a formal reuse management system if we don't engineer the data so that it is available to be reused. It will not likely be reused instead it will be used a second time through an entirely much more expensive process. The preparation time means 80% of our time is spent in that preparation the left hand side of this particular piece and 20% in the analysis piece and of course it's all got to be governed by an ethical use framework within that particular context. So I'm still trying to define data management here let's dive in and look at the word meta in particular lots of definitions here the one I'm highlighted in purple here is beyond transcending more comprehensive at a higher state of development and here are our card catalogs that we look at now I know that some of you are young and we have to tell you the card catalogs is in a library. That was a place where you traveled to physically in order to access reference materials be for Google then you had to manually search through this card catalog for subjects by reading abstracts and other information that's printed on these three by five index cards and organized in a fashion. The cards also contained a link to a physical address of where the reference material was in the library and again these reference cards identified what are the books where they located. You can search it by subject by author by title, the catalog shows again relevant information around each of these that allows you allowed us in the old days to determine which books helped us to meet the users requirements and without the catalog. Finding things was difficult time consuming and frustrating and quite frankly impossible at the time as well so we're going to go with our first definition of metadata of data about data. And if you take a particular piece of data you can make lots of metadata about that a good place to look around your own organization is that you have a group likely in your organization that is already maintaining metadata about users who access your network so you can go to your network group and ask who is the person who is responsible for determining which devices are permitted to log on to your network or which individuals or at what access points and when they do access what are they allowed to get. Again, there's probably a named individual in your group somewhere that takes care of this find out and now you have somebody who's doing metadata management within your networking group. And it gives you a little bit of a benchmark. For example, I've worked for a lot of organizations before and I found one that had five people that were in charge of this particular piece it's a pretty big responsibility a large user base in this case. And then I said well if you've got five people determining whether or not people are allowed to get on to your network or not, and you've only got three people keeping track of your data it doesn't seem like the right number does it. And they agreed with me on that particular piece. So now that we have data and metadata what do we want to do with it and the thing we want to do with it is to leverage it. So you'll notice here I have a lever set up in this case. Again, one kilogram ball cannot balance that other one but if I make it larger it can get there so again we can move more with less which is the definition of leverage all the way around. So keep increasing the ball it increases the leverage on that. The way to think about it in terms of data is that you've got a bunch of data on the left hand side of that equation. And you've got some technologies that you'd like to have in the instance that I'm showing here, you're looking at the fulcrum, excuse me, and the lever lever and the fulcrum lever is the triangle the fulcrum is the the ruler like thing that we're looking at there so that's our use of technology here again if you didn't know how to do this, it would be very difficult, even with a set of proper technologies of course we've got people in your organization. I call these all of your knowledge workers supplemented by your data professionals that are in there and there is some sort of a process although likely because most organizations have not adopted uniform ways of describing this, your process that you will be one that is pertaining to work groups as opposed to the organization which means your opportunity for standardization is significantly less again notice the three-legged stool there of people process and technology in order to do this going of course you could do everything with your processes so you use strategy to guide the particular process and we're going to add another word in here to rot. Rot is data that is redundant obsolete or trivial, and if you can reduce the amount of redundant obsolete or trivial data that you have you will increase your data leverage of course the bottom line to this is that metadata is a primary means of applying leverage to data in this case data leverages a multi use concept it permits organizations to manage their data, both within the organization and when they are exchanging data with their various partners in support of the mission leverages again, enabled by metadata, it permits implementation of what we call data centric technologies processes and human skill sets, focusing on the non redundant obsolete and trivial data. The bigger the organization the greater the leverage potential of course exists and treating data more asset like simultaneously reduces it costs and increases knowledge worker productivity. Again, metadata is your key to all this process. So, when we look at this again, a little bit deeper dive in the rot area right now one of the things you have to understand of course is that some of your data is worth more than other parts of your data, it's a separating of the week from the chat. And the first question that comes up is is well organized data worth more will the answer to that question is quite simple before the information age. So we had of course books that had all sorts of metadata in them to help understand the books. Imagine if instead of handing you a well bound book with page numbering and indices and things. I instead handed you a bunch of pages, and without a backing on those pages, the knowledge on those pages disappears pretty quickly. So, clearly there is value to having data well organized. Again, I'll put a little shout out to Abbie covert's wonderful book, how to make sense of any mess a good book about the need for information architecture for non information architects. So, once again, 80% of organizational data is rot redundant obsolete or trivial data. And the question becomes which to eliminate well metadata is required for the identification of those assets. It's required to focus attention on repairing common data assets, those are the ones you want to fix first if they're shared of course, and it permits value to be ascribed to the data at the necessary granular level that occurs so then the question comes up. All right, if we're going to do this, who is best qualified to do this and the answers a combination of your data people coupled with your business users in this particular instance. There's a repetition here in this particular talk metadata yields what do we get well, you can ask the question do we have this specific or this class of data assets yes or no. What is the quality of these well in this case it's been judged not suitable. What will the cost be to improve this class of data assets with the right metadata you can provide an answer that says 35 cents a piece times 6 billion or whatever the number is that you can provide. And can these assets be provided more granularly in this particular instance the answer is no. Now if you haven't seen this diagram before. Welcome to our world. This is the dama demock wheel that we talked about showing 11 practice areas in there and once again you can see I've circled in yellow the metadata practice area, ranked as one of the top areas that we look at notice of course we spelled it wrong but that's a minor detail in this edition on this and from the demock they have a overall diagram here that talks about this I would read this to you if I just wanted to cheaply do nothing else for the rest of this particular piece but I'll leave that for you guys to do and instead show you an example of how metadata is used in an MP3 app. This is the app called used to be called itunes I think on Windows it still is called itunes on the max they now call it music. And let's just see what happens if I take a CD back in the days when we had CDs, and I stuck to CD in my CD player in my computer. Music the app would come back and tell me how many tracks I have 25 tracks on this and it also tells me the length of each track well that's useful information but it's probably not what we're really looking for in terms of managing music. So one of the next things that happens of course Apple and others connected this up to a piece called grace note.com which is a grace note media database and through a fingerprinting operation it runs out there grabs this information and brings it in and labels the songs that are on the CD gives me the artwork that's associated with it sure would be a pain to type in all of this information and of course with limited exceptions we end up with good results from here however occasionally you do end up with a question back from grace note that says is it this album or this album. I can't tell. Now let's take this metadata management example just a step further I've now taken tracks and turn them into songs and times and artists and artwork and things. Now I want to group them I want to group some like things together I want to put all of my miles Davis recordings together in this case. And so to organize it I create a smart list for an artist containing miles Davis however when I did that it actually showed me results that I didn't get what I was desiring I had another miles Davis CD that I had forgotten called live at the film or East. In there and so I need to fine tune that and say I don't want just artists containing miles Davis but artists containing miles Davis an album containing the complete birth of the cool in order to do this and I can move it to a folder or I can do something else with it. But by the way, most of your MP3 players already have this built in so everything that I've showed you here is completely redundant it does this automatically for you. When it brings in the CDs puts it in there in an album with these particular pieces so there would be no point in this but there might be a point in collecting all my miles Davis together if I wanted to make a miles Davis playlist. This same architecture of metadata the arrangement that we have does now work for the interface the processing and the data structures applied to podcasts movies books PDF files and the economies of scale are just enormous again the leverage that we get from this is enormous so now you all know how to take your MP3 app and show somebody else a use of metadata around that particular application. Once again we get back to our metadata yields, it yields valuable information about your musical assets do we have this specific miles Davis recordings but is my most played miles Davis recording it can tell me that how much does it cost to acquire more miles Davis recordings in this case it was $1.29 a piece. And can I listen to the entire album or playlist before dinner, not easily in this case because you can see it is 1.3 hours on this. So I've done a little bit of definitional material let's get into some strategies around how to do this. Again the idea here is. It's critical to understand and to make everybody else understand what it is that you do if others don't understand what you do. You are perceived as a cost whereas if others, others understand what you do you are perceived as a value it's a very simple piece of guidance, but it has served this well and this is what metadata does it helps represent understanding and use. So data in order to be understood and documented as a it has metadata specifications. It needs to be understood by two different people looking at the same specifications, same sheet of paper if you will think of musicians trying to read music, but it's also important to understand that this implies that the same understanding is here is shared by humans and systems as well. Organizations have taken different approaches to metadata I mentioned or Shannon mentioned in the bumper up top that I spent some years with Walmart and when they looked at metadata they said oh fantastic. We can incorporate part of the Walmart logo in there and Walmart said metadata is any combination of any circle and data in the center that unlocks the value of that particular data. So it also should show you that for every piece of data you can have multiple pieces of metadata around it which means there's more metadata than there is data who this could become problematic hold that thought for just a second. And again shout out to Brad Melton out there at Walmart who came up with that particular example it's a very powerful one I believe. So if you look at outlook, however, outlook will do the same kind of thing here's a screenshot for my outlook in the upper right hand corner, and the what information that comes in there give me the subject the priority user ID in my inbox where the why is the body of the letter the when is the sentence received. And I of course just like all of the rest of you use outlook to weed out important stuff. I can find the good stuff among all the junk. I can organize future access with a series of outlook rules, so that I can anytime something comes in from one of my bosses. It goes straight to the top of my inbox in order to do this. Just imagine trying to manage email if outlook did not make use of the metadata. In this case the who from and to it would be very very difficult so metadata is, if you will, everywhere in all of our systems it is data. Excuse me metadata is to data what data is to real life so data reflects real life transactions and metadata reflects data transactions around it. In some ways it's very much like the digital twin concept. My colleague David hey has a wonderful book out that dies into that a lot more detail around this but I'm going to bring up the Gartner definition here that I think is so important. Gartner defines metadata as saying metadata unlocks the value of data, and therefore it requires management attention. Okay, so now let's go to metadata management set of processes that allow us to ensure metadata again at this point in your organization. All of your knowledge workers have learned data on their own is very likely a true statement. And in addition to that they've also then learned metadata management on their own as well, which means they've all learned it differently. That's a problem for most everybody and metadata as you can see here is ubiquitous let's get to the bottom line. What is a jaren then why is it Peter says you can't use it as a noun well the reason is one person's data is another person's metadata, any piece of data can be used as metadata, any piece of metadata can be used as data. So the use of the term metadata is really more of a verb than a noun, the proper term for it is jaren it's a form that's derived from a verb but functions as a noun. Therefore, metadata describes a use of a data, not a type of data very important distinction and the value proposition associated with that is really key to understand because since 80% of your data is redundant obsolete or trivial. There's no point in maintaining metadata about rat data in there and we can now start to focus in and get what we're really trying to do here's an example. Again, this is just if you want to read more about this, but it was the first time that people realize that even looking at new systems could be useful in understanding metadata so here's, for example, the way people soft used to ship. There are people soft modules around administer workforce compensate employees and develop workforce three primary modules of people soft that it had in there, and the people who were encountering it for the first time said well, where do I learn about this and we simply call it the number of sub modules in each of those sections and showing that they were relatively speaking about equal in that but if we looked at the expansion of the administer workforce, and you can see that recruiting workforce and manage maintaining competencies is much more complex than planning successions and managing positions around that so perhaps not perfect metadata but certainly useful metadata here's another use of it here this is the old system that we had here at Virginia Commonwealth University many many moons ago. Anthony Danielson is still out there as a former student say hey to him as well as Chris digs and some of the others. But here's the database that is behind it in other words the metadata that is looking at this. And all I really have to do to help you understand this particular screen is describe this as a series of parent and child relationships the parent is something here in the right hand corner that I've circled called student database master sdbm, and everything else is a child ranging off of that so even though you can't read anything about this metadata database you certainly understand it. And more importantly, you understand it well enough that if I show you that this was the actual proposed system to replace that original system, you can see here that this is an absolute travesty. Whatever was doing this have no idea what they were in fact doing. There is almost a jail story in there but I'll have to wait until Q&A if you want to hear about that. That said, IBM did at one point, put together what they believed was all the metadata that you would ever need into something called the IBM AD cycle and there is a website out there that has this stuff again it goes into our here's a great website that started on some of these particular pieces but the more important piece from this little definitional section is that people when they understand metadata they start running around their organization and looking at stuff and saying is this metadata is this metadata is this metadata the answer to that question is always it could be metadata so the real question to ask is not is this metadata, but would we obtain value. And that would include this particular metadata within the scope of our metadata management practices, because if you just collect metadata about everything and you recognize that 80% of your data is redundant obsolete or trivial. You will be doing a lot of work for not a lot of value in that process. Let's move on to our second strategy then enforce metadata to be the language of data governance. And that is critical let's just talk about what data is for starters, data, if I use the word 42, that can be a piece of data, and I can associate it with a book that I'm very fond of the hitchhiker's guide to the galaxy where 42. In this context means the meaning of life, universe, and everything, some of you will know what we're talking about others are completely confused let's change 42. And that also was Jackie Robinson's Jersey. Again, a very happy set of associations associated with that particular number or here's a third association with that number. It represents my age 21 years ago. Well, the question of whether Peter is allowed to consume adult beverages in the Commonwealth of Virginia where he resides is obviously answerable by that particular piece of data. So data are facts, paired with meanings. Data source shows up in lots of different places so we have to differentiate between just data and useful data data that we actually find useful how do we find it useful. Well when data is turned into information, it is usually done so in response to a request. So if we take some data and find out what data has been requested. We now know we have information as a result of that. And of course it's obvious from this structure so far that you can have data without information but you can't have information without data. So all of the information literacy programs and things that are out there I think are pretty silly, because they're really not focusing on the right issue there's one more level for this as well. How is the information used well, we've called this over the decades that I've been in the business wisdom knowledge and intelligence. I'm going to come up with some new term as we go forward on this but notice I've given you a fairly good metadata structure there data is a combination of facts and meaning information is data that has been provided in response to request these are objective criteria, and then intelligence is information that has been strategically used. Again, this is the most basic form of metadata in order to do this. Let's take it to the next level which is incorporating strategy in here. Organizational strategy, absolutely has to come and be where the data strategy is derived from. What are we doing with data to help the organization achieve its objectives. Similarly, that focus on data strategy has to be the question, what the data assets do to better support strategy is what data governance should be doing. The direction how well is it working gives us some feedback in there as well. We're going to incorporate the stewards in here as well just to make sure everybody follows the business strategy the organization strategy should be expressed in terms of business goals, and the language I've already set of data governance should be metadata. So given those two pieces. This means that what we're really looking at here is the opportunity and I'm so sorry I went too fast on that is the opportunity for everything to go correctly in a way that keeps things focused on specific actions that will help the organization use data to better achieve its strategic objectives. If we're doing anything other than that we may want to ask the question about where our priorities are around all that process so there you go is the entire diagram put up there without my errors in keyboarding over top of the whole process. If we use business goals and metadata to confine the language the vocabulary, the focus of what we're trying to do in both data strategy and data governance, we will much better be able to support the organizational strategy and monitor how effective our support is around that. When we look also at what comprises the data community of course it and systems development are going to be there as supporting infrastructure legs around this we've got a combination of leadership stewards, some subject matter experts and then everybody else that we put in there. Most organizations will draw a line around the left hand side of this diagram, just to talk about that and say this is going to be our data governance operation that we have over here and the purpose of leadership of course is to get resources, obtain feedback from people about decisions, the stewards are then charged with implementing those decisions. This means they need some action to take and some changes that need to happen in order to do this. Of course, we're still going to continue to get data and feedback and ideas all the way in and once again, if we're doing this without talking about metadata, we are introducing more errors into the process instead of less. So again our metadata here yields valuable information about the organizational data governance assets and process do we have a shared understanding of our goals, are we focused on similar goals how effective are we being there may be some cost per data and what kind of metadata do we find most useful right now well this particular organization says the supply chain metadata is of most value to the organization at this point in time in our hypothetical example here. All right, number three strategy treat repositories and glossaries as capabilities, not as technology. In most cases you will see people using the term glossary is where you keep definitions, I cringe when I hear that. The problem is that first of all, all data models all data components are incomplete without definitions and my definitions are good purpose statements are generally better than definitions and I'll show you an example of that here. Of course, all of this then goes back into its all metadata let's look at the specific example. This is from the veterans administration system something that I worked on. When I was at the federal government at one of my exercises in there. The question is, what is the purpose statement well first of all it says bed. It's a principal data entity and this is a substructure within a room that's a substructure with facility that has information about beds within rooms. Where did it come from well there's the reference and where it came from here the attributes that it has now that may be a partial or full list we would like of course to know that it is in fact the full list we would not want to have any ambiguity around that. We also are associating this entity with another entity this of course also is metadata, saying that one room contains zero or many beds in order to do this. And finally, all of our models statements components are in a draft form until they have been validated from that particular process now. Just a quick little story here. One of the things that we looked at when they looked at the definition about this this was a very bad definition here here was a better definition. The primary means to be used to track patients within the facility, each bed will track exactly one patient. Again, it's a terrible terrible statement, because of course the first question that pops up immediately is. Okay, so you're going to use this as the primary means to keep track of this. Are we going to put the bed ID attribute, we need to be able to identify this bed from all the other beds that are possibly out there. And second of all, what room is the hallway, what room is the elevator those two questions that contractor was unable to answer, and consequently unable to identify the entire scheme of keeping track of patients by putting them in beds. Of course they went to a wristband which as you know, turned out to be a much better way of doing this it was kind of an older thing but once again notice that the purpose statement in the metadata, actually save the government a fair amount of headache in this particular interest let me tell you another quick story on glossaries and this is from Nokia which is a wonderful company if you want to read about at least as book here transforming Nokia is just a fabulous one. It really started out as a tire and rubber products company moved into consumer electronics when I had an association with them. They were in mobile phones and interestingly enough 2% of the population of Finland speaks Swedish so the entire Finnish population tends to be bilingual and Nokia as a company wanted to play internationally so they mandated the use of English in all business settings that meant if I went to Finland and had a meeting with Nokia people as I did for about four years. And I was able to speak in English to me which was really nice because of course I don't speak Finnish. In the process of learning this however Nokia discovered lots of words were unknown and culturally, they had to change the focus from saying, don't sit there and pretend you're the only one that doesn't know it in fact it's culturally bad not to ask questions and good to build a common vocabulary so when an unfamiliar term was used in a Nokia meeting, the group access the NTB to see if there was a golden definition. I should have made the book, the book, a golden instead of blue there down in the corner, and the group would vote whether or not to submit it for inclusion consideration for inclusion in the NTB and weekly, the NTB group would sit around and review the submissions, and then publish the versions of the NTB which of course if you haven't figured out at this point stood for the Nokia term bank, one of the best examples I have seen of taking an organization and building a business glossary in a very cost effective way, and in a very culturally sympathetic way in order to do this. Another topic with respect to glossaries is that the conversations between people wanting glossary and the people selling glossary tends to be uneven there's a technology gap that tends to occur because the customers are not knowledgeable and the vendors of course are supremely knowledgeable a lot of that can be made up through do it yourself types of exercises and I'm going to show you a couple of things right now that hopefully will get you started on them. This is a model that we created for something called FTI financial transactions international, not relevant in today's environment but it just happens to contain the data that you would need to keep track of if you wanted to keep track of metadata in your existing systems. There's a blueprint if you will for it again you'll get a copy of the slides for all this, which means you can build your own repository out of something like Microsoft access now I'm not suggesting that Microsoft access will function as the repository for your needs for your organization. If you can work with one for approximately two years or three years, you will be in very good shape because here's an example of one that was built here. Again they're looking at a table, FT underscore T underscore account. You can see that in the tables entropy so you can click on entity domains and return. If you wanted to go back to a different screen. In this case, but instead what we're going to do is take a look at how all this comes together so here. I'm sorry, different table FT underscore T underscore ABDF. No way you would figure out what that was but you can look at the column details and for each column you can look and see whether it occurs in a primary key. You can look at the column right there or as a foreign key. In this case it does not appear as a foreign key. And then what the actual columns of that table are this type of build for you is not an expensive process and you will learn so much from doing this and all of the basic blocks exist out there for you so you don't have to learn this all the way from scratch so once again metadata in this context yields do we have this specific class of assets is this data item used elsewhere well we can answer definitively know it is not in this instance. Again cost to acquire can these assets be shared securely. These are the types of questions that metadata will allow you to answer. And the key here again is to start from existing building blocks and the key here is there are lots and lots of things that have been done already. So let's talk about architecture for just a quick bit architecture is about things. And then what are these two thing one and thing twos and the function of those things. They need to know how they do perform their task individually and what type of data that they use as a part of this and how do these things interact. And as a pretty abstract set of concepts these are necessarily in order to figure this stuff out when we go to a data architecture then we look at architectures and say details are organized into larger components, and then that means that this is going to be fairly intricate because this organization is now going to be inculcated in subsumed into larger architectural components so anything that is intricate there those processes and properties are inherited by the entire models that also then introduced dependencies into the process and architectures, which should introduce purposefulness into the process of course we're going to do that from the data perspective, we've got the attributes for the intricacies, we've got the entities and objects organized into the models for the dependencies and the models organized into architectures in order to do this. And there are lots of examples I've shown a couple here here I've got thing, and you can see the thing ID and the descriptions and things like that that are up there. You can see the dependencies that I've organized into the model that says one must might eventually have many of these things in order to connect them but there's no clear necessity to have that intersecting entity existing within there. And models are then organized into architectures of course, there aren't terribly many examples of architectures because most architectures end up kind of looking like this. I imagine an enterprise architecture for any of these large organization. It's going to be big and confused and of course, it is all metadata in order to look at this. So, keep thinking around the process of building towards your data architecture but you don't need it all at once, you don't need to start it. All at once, it can be built in pieces and phases and notice in this particular architecture there's a green phase, and an orange phase as well as a purple and a red phase, and a teal phase in order to cover all of this up. I just want to tell a quick story on this particular diagram here too. This was created by a very smart CIO, who knew that the organization was buying a ERP and enterprise planning program something think people soft etc etc bond. There are many options that are out there in the market and the management had somehow gotten it in their head that that was going to fix all of their problems they weren't going to need to have any other systems because of course the sales person told them, this system will do everything that you want. So the CIO sat down and maintained this, and this is actually called the quilt diagram by the organization. It was hand done for the first time. And it was so useful they kept reusing it. So they now have it in sconce in somebody's office there's an owner of it just like there's an owner of person letting people on the network and not. We talked a little bit earlier there. This is how you organize and use your metadata around all of this and again, it is all metadata because it's data that's being used to leverage other sets of data. Here's another metadata piece here, particularly for design patterns. Perth Australia where Clive used to live by the way Clive someone that came up with the purpose statements on that. Make sure we give him credit for all of that. You know just a bunch of buildings in Perth Australia well it's not anything particular to Perth that the restrooms are located in the same place on each floor why is that well because generally restrooms depend on gravity and water to take things away and so you want to make sure that they are having the shortest amount of travel time that you possibly can so even in these large buildings that are here you can be sure that the efficiency of the restroom. Is a pattern that they repeated on each floor of the building that they were able to do so because it saves time and money and effort, just the same way as understanding metadata will help your organization understand better its use of the specific systems that you have. Why would we do anything different with electrical wiring HVAC floor plans, etc, etc. The architecture design patterns are the kinds of things that we will put into all of these buildings and organizations and there are a number of really good examples that you can use to go and find out about this again I mentioned David Hayes book and before there's a different book of David Hayes data model patterns, David Marco one called universal metadata patterns, metadata models. I've got one called XML and data management it's got patterns in the Bible if you will have this is a lead Silverstone three volumes at the data model resource book. And if you can find a copy of that that's still on open there do exist you can find them. There's actually Erwin models in the back of this book that you can immediately on use and transfer into your metadata environment right away. Again true story Len and I were at a conference in San Diego a couple of years ago, I happened to see him walking by and one of the questions coming from you all was have you ever found a pattern for a cash register operation in a pharmacy. And so the outline was able to tell me that's book to pet chapter 13. Literally he was just walking by and remember that stuff very, very well. You remember a little while ago I showed you the model of the data model metadata here is a generalized model for maintaining metadata. Again, like to it on my website there that you can go out and take a look to it but this model here will generally allow you to move things from or to whatever it is you are attempting to get them into or out of, in order to do this. So just a very, very useful starting place in order to get this done. There's also some semi structured metadata to people tend to call unstructured I don't like that term because unstructured means you really couldn't convert it to structured so we can take semi structured data and we can make it more structured, or there's still probably we can describe non tabular data as tabular data when we're doing it but there's descriptive structural and administrative metadata around each of these non tabular sets of data that you can use in order to better leverage all of this. Again, you might imagine SharePoint is doing exactly that kind of leveraging when we're looking at it. So these metadata patterns again yield you various comparisons and starting places. You want to build this pharmacy billing system from scratch when you have access to something that can show it to you right there. Will the proposed software fit with our existing systems it is now considered best practice when somebody proposes either a cloud based app or a regular app for you to ask them for a copy of the data model so that you can determine the best fit to your existing infrastructure your existing architecture. The industry best practices exist. Yes, again, they're not only existing but the federal government is making them into laws which is a wonderful thing around that and we can look and say has anybody published a model for implementing GDPR in this context. In this instance, the answer was no but again there may be other instances around all of that. So, let's talk about some benefits and applications and sources around all of this the first one's a bit scary. I'll probably remember President Obama came out at one point and said okay well you know it's just your metadata so let's just see what metadata means metadata means they know that you using your mobile device. You were on the Golden Gate Bridge because of the geo positioning piece, and you, you know, called a suicide hotline but oh the topic of the call remain secret because it's just metadata well as you can see, all of these questions about metadata, allow you to learn about what's actually happening in the organization. In fact it's so valuable that many companies have built an entire business around it here's an example of a net business called in Vera, that I was peripherally involved with but I love their commercial network so this is literally their commercial. And it says something like you know company and company be talked to each other and of course there's third parties. And these are all the types of documents that company and company be might exchange via phone facts mail email etc etc, even EDI of course comes into the picture now of course we've got XML based transactions. All of this allows us to help well, again imagine this between a and B, and then imagine a is also not doing business with just be but also with CD and a number of suppliers. And of course all these folks are talking to each other in the background as well, which means we have a tangled mess of confusion watch this part coming up this is really where the value proposition comes in. The company was called in Vera. I'm not sure exactly what it stood for one of those wonderful Latin words it probably means truth or something around this but you can see what in Vera did was insert themselves in this particular industry as a broker of transactions and things that were in here was a 100% metadata base piece. All they needed to do is connect with either SAP JD Edwards whatever the other companies were in order to do this they were able to pull all of this together. And again this is still coming from their advert. The value here was the in Vera clearinghouse that they were going to put in place, all of this type of data because they'd have it not just about the companies that they had connected up to this but they had to do industry scanning and find out how they're subset of industrial partners, compared to everybody else again, the metadata skill sets values patterns were reusable across a number of different domains. I mentioned before that the federal government was incorporating metadata in it. This is an act that was signed on the 14th of January and 2019. It meant the use of best practices, dictionary, term bank control vocabularies. All of these things are absolutely critical don't worry you're not expected to memorize it as it scrolls by there you at high speed. But here are the highlights all federal data is now open by default, non political CDOs are required, which means they can't be taken out there government employees they're not political employees. So open data and open models are required in policy evaluation. The last bullet point however is what gets everybody the penalties for violating the FIFA act are higher than HIPAA. That's enough to make people pay attention right away. So we've talked for very rapidly at quite a distance here of talking about what metadata is in the context of data management. We have to define data management in order to do that. And when we talk about using metadata we are trying to leverage the data. We may be looking for tags we may be looking for attributes we may be looking for the value of attributes we may be looking for mandatory relationships that exist in here there's lots of specific things that you can look at and I've given you an example to use with you do your iTunes or your music app. If you want to understand how to do this just imagine trying to manage your music collection or movie collection or P podcast collection without metadata of course it doesn't work. I've given you four specific strategies in order to do this. Everybody learns metadata as a noun and they tend to run around pointing to things and saying is this metadata if it's metadata we must manage it. It's not the correct metadata is a use of data, not a type of data and you should only use it when it provides you value. Second strategy, enforce language enforce metadata to be the language of data governance. Again the idea that I've seen so many people sitting around and arguing about this that and the other thing, and they don't even have the terms correctly on here so what they're arguing about is an exercise that ends up being completely futile. Yes, help make sure that your data governance people are speaking the language of data governance which is metadata around that. Look at the idea of starting to pull together your glossary repository thing as a capability, not as a technology. It's a cyclical process you should do a version and then another version and then get more people involved and then make the scope wider and again crawl walk and run your way up to the point where you can have a good conversation. With our wonderful partners in the vendor community who have some really exciting stuff. As you saw Jesse showed at the beginning of the hour up here. And finally, don't start out with a blank sheet of paper. There's nothing that terrifies people more than a blank sheet of paper. There are many, many, many resources to available to help you start to jump start these efforts around this. Again, lots of benefits in order to do this and we're going to do a couple of quick takeaways. And then we'll invite Jesse back here and look to your questions as we get forward on this so again benefits the idea is increasing the value of strategic information. We use metadata to get rid of the 80% of data that is getting in the way of us obtaining value from the remaining 20% that we can reduce training costs by simply showing people and documenting the stuff. And then as a turnover I did a complete exercise of building out an analytics group completely funded by the cost of people not quitting their job this particular organization said it costs them $50,000 to rehire people in a particular job, and they would quit because there was no documentation around about what they did so we put the metadata in place the turnover rate dropped, and we were able to show documented serious positive costs around that. Business analysts will allow them to find out knowledge workers spend almost 80% of their time recreating knowledge that already exists or finding knowledge that exists. If knowledge workers get to the information that they need to faster, it will help overall productivity in our organizations and bridging the gap between business users and it think of them as disparate musicians, trying to sing off the same sheet of paper around this. Again, all of this helps to do time better faster, right in order to do this and reduces risk because we have much better control of the change management process. We're reducing rework inconsistencies, what Tom Redmond calls hidden data factories around this. So a couple of quick takeaways as we round up to the top of the hour. Again, thinking about it as data about data is a good place to start. But it is so incredibly important to make sure that you get beyond that, because if people think of it just as data about data you've already seen that every piece of data has multiple pieces of metadata about it which means it's just simply impossible. If we're merely managing our organizational data, how are we going to manage all of the organizational metadata about our data it's a non starter on this so managing data about data, where it provides value, because metadata does unlock the value of the data and it doesn't require management attention, but it's much less about what and much more about how which means your organization has to start developing these capabilities and don't get me wrong. Every time I come to visit my I've worked with over 1000 companies across my 35 years on this, there are people in your organization who are already doing this and doing it very very well, but they're buried somewhere, and they don't have the visibility to understand it. Again the quilt diagram story that I told you a couple of minutes ago, was somebody doing this and the CIO knew they were doing it and went to them and said, can you enhance the quilt diagram in these certain ways so that I can make this point to management that the new ERP is not going to replace all of our systems that are here. And again it was, it was a very very good way of maintaining that particular piece we'd like it to be more formal. We want to make it to be more recognized in the process. Again, when your data governance processes are going off the rails people are sitting in meetings and they're not sure why they're there or what their purpose is or anything else is going on. This is why you need to be speaking in metadata so that there are real business problems tied to the other end of these data things that happen, so that we can get excited about the data things happening, and talk about it in terms of metadata processes. Again, the idea is in any organizational challenge there is a metadata cause of this. So we need to make sure that we have that in place so we can use it in the way we'd most likely want to do this. It's not really well understood, but a challenge of any sort is going to have a data component. And if we've defined the data component imprecisely or in many instances wrong. There's simply no way that we will ever be able to spat to correctly satisfy the problem on this. That's a different lecture. Let's talk about it perhaps offline around this. And finally, when you do get it right. The question that pops up should pop up on a regular basis for you is, should we include this data item within the scope of our metadata practices. The answer to that is yes, if it provides value for you. I'm going to bring you with a couple of recommended reading pieces. It's a fairly long list from the Denmark that's out there. But again, lots and lots of places to get started on this. I'm going to bring you around to this corner just to make sure that you know we're going to do data quality next month on this and here's the other upcoming pieces and now it's time to go back to Shannon and invite Jesse to come back in and see what sort of questions you guys have. Thank you so much for this great presentation and if you have any questions for Peter for Jesse feel free to put them in the Q&A portion of your screen. And just to answer the most commonly asked questions just a reminder I will send a follow up email to all registrants by the end of Thursday for this webinar with links to the slides links to the recording and anything else requested throughout. So diving in here. You didn't mention quote unquote standards is it necessary to designate metadata standards across the organization. So a great question Jesse did he get back on here. I don't know how you guys end up using them but the wonderful thing about standards and courses that there's so many out there you want to go first Jesse go ahead. I guess there's two kinds of standards whenever I hear the question and I don't you know and I'm trying to figure out exactly what the context is. There's standards in the sense of like requirements and policy level, like we have to do this kind of standards. And then there's actual metadata standards you know whether it's to exchange data or the representation of the data possibly the model of the data. So that that's what came to mind initially Peter. Absolutely. And the key with it is that many people as they understand standards and how they work tend to think of them as perhaps superhuman. So standards can help in order to do this but standards by themselves will not solve things all the way around let me just give you a very brief example of this in the United States, we were unable to describe uniform crime statistics until very recently. In the last couple of decades, when something called name the national information enterprise information model was created and provided the platform and the structure and the standards of course, to report on crimes because some jurisdictions would report one crime one way and some would report another way and we have somewhere in the order of 20,000 individual jurisdictions in the United States reporting crime statistics. And putting all of that together was impossible without standards, and now that we do have standards we can now start to look at things in a slightly different way. Here's another quick example though, ICD stands for international disease codes, and that went from 9 to 40,000 to 10, and there's going to be many more at 11 in order to do this, we do need increasing levels of granularity so that we can understand the difference between, and I'm not a doctor right lung infections versus left lung infections, you know, or something like that. But if we don't have the granularity and the data we won't be able to understand that so it's a gradual process of moving things through. And standards are very important to this but it's also important to understand the limitations of standards as well. We get this question a lot Peter as you know, and we're doing a webinar on it next year I'm very excited. How can we implement and enforce a data strategy if we don't have C level sponsorship. It is harder to get people to understand this I can imagine most people listening to an hour of this stuff with the two of us talking would be not excited. Certainly, you know, nothing makes me want to put my, my code on and go out and walk the dog or whatever around this, but it is important. And I guess the first thing is that you always are going to have to be managing upwards in one form or another in your organization. We haven't incorporated data education at literally anywhere in the curriculum. Think about it for a minute the only thing that we teach undergraduate students about data is how to build a new Oracle database in the cloud at least now but it's still a new Oracle database. If there's a skill we do not need any more of on planet Earth it is how to build a new database around this. So, people are not aware that they don't know. And so it's a really good idea to find a way of relating to them in some sort of an elevator pitch type story and I've got several organizations and I've worked with where we have these sort of pet stories that come up I've got one organization. I'm going to sit down and start talking to them about metadata issues and somebody will eventually stop they wouldn't Peter. Are you going to tell us the chocolate story again. It's like no I don't have to because you all remember it good for you right they finally learned that particular lesson, but it's, I think, rarer at the top levels to really get that buy in and the only thing that they really understand is the bottom line, or if it's a mission oriented again part of the government or something. It's the impact to the bottom line so you have to translate that metadata into it. And again, Jesse I'll go one more example here and then kick it over to you guys to see how you do it but one of the things we've all come out of the other side of COVID hopefully okay on this. There's a good set of stories out of the United Kingdom, where they had people taking, keeping track of COVID cases on using Microsoft Excel. Now why would a person who's a healthcare professional need to understand which version of Excel that they're on and the answer turns out to be that if you're using Excel with an Excel file, it simply drops rose after 16,000. Now how would a healthcare professional know that any goes if they were entering at the top of the spreadsheet we're causing rose at the bottom of the spreadsheet to simply fall off. So that's a little piece of trivia and metadata that they wouldn't necessarily have known, but in this case did cause the United Kingdom to make some decisions that they perhaps wouldn't have made otherwise if they had had the correct data, because the metadata was incorrect about that Jesse how do you guys look at selling to metadata when you're trying to work with organizations to help them solve the problems that they're facing. And, you know, bringing it to the C level right working up in your slides Peter you mentioned governance and C level people CDOs they want to hear about governance and tying that into your message is really important because metadata is what is governable. And I can't recall exactly how you talk to it and slides Peter but that was important. And so when you take metadata up the chain, bring things like governance and how and why with you right they want to hear use cases, applicability, and the fact that you can do something like governance around it. Also is important to help them with the value proposition there. Again one of the other things that we talk about in data governance in particular and you're right executives are more concerned about data governance, because they've all been exposed to corporate governance. And so any executive that hasn't had a set of education around corporate governance is probably going to make an error pretty quickly in today's environment to do this. So, when we talk about what we're trying to govern. One of the things we have to have is a list of those assets. So if you were trying to keep data about something. What are the list of data items that you are taking care of. And even if that list exists in somebody's head I've got a great picture of a guy named Ron and one of my organizations that I worked with and he's got to be standing in front of a projector that has the word data on his head. It's obviously making fun of him for being sort of data but actually the way they operated in that organization instead of a glossary. They went to Ron and said Ron where's this thing connected to these things and where's this and Ron had it all in his head it was great. And as long as Ron doesn't win the lottery and disappear from that organization they'll probably be just fine. But as you can imagine there's some people in corporate risk that are saying, I think that's the way we're going to go into it. But it is always the case that you're going to have to manage upwards we just haven't done a good job as a society of telling people how to do this well we're saying data governance is managing data with guidance for what type of guidance should we have. Well, if we're reporting crime statistics, then we should be using name as the standard for reporting crime statistics or our crime statistics will be not integratable within the larger picture around all of that. And Jesse, let me ask you about sorry go ahead please go ahead. I was just going to say Peter you know it takes us back to the first question. What about standards and you brought it all together there. So, Jesse when somebody's looking at buying some of your technology. How far up do you have to go in order to the techies look at your stuff and go oh that's fantastic we've got budget here we go let's do it, or do you have to sell to their bosses occasionally. Both. One or the other may be coming at us, they might be coming at us very well organized together with a purpose. But a lot of times it might be something like you know the techies come in their interest and they're like this is great. Can you even help us bring the message up the chain. And we'll do that we'll collaborate with them help them with their use cases and define you know what we would call like a you know a go live charter. What would it take to get you to success and then that's a that's a good communication piece to work up the chain with and management at some level is going to be concerned with what's on my plate and what's on somebody else's plate and if this is on my plate and it's going to help my numbers. Yeah, and you know if you can show that you can make their numbers better without them having to do much because what you're trying to propose is bringing solutions that other people are putting in place, and everyone gets to benefit from it that's a that's a really important part of the story. Fantastic. Super. Great question thank you. So, what should companies do with their rot data. From a risk perspective, the answer is they should get rid of it. Again, I don't know if you're doing business with the federal government these days but the federal government is actually sending out things from various agencies that are now saying this email is subject to a seven year retention policy. I've seen that happen in several cases in order to do that. Yeah, it's it's a it's a tough process in order to just you want to jump in. It's, I like the get rid of the rot analogy right so it's the truth but that's what a lot of times entire efforts have to be dedicated to. Maybe we're at a point to where it's like easy to be monitoring and aware of our retention policies, but then how do you even get rid of the rot, and then there's standards to follow and policies to follow to to to get rid of it so it. The life cycle of data and metadata includes rot. That's just something that you have to face up front. And Jesse, I'm going to say I'm a couple years older than you. And I have boxes around the house from the last time I moved, which was almost 20 years ago that have never been unsealed. So that's how I'm going to get rid of my rot. I said if I haven't needed it in 20 years I'm probably not going to need it in the next 20 years. Yeah, it is tough and I'm sure that's not what the question was. I think the question was probably focusing in more on that how can we use metadata to actually identify rot and that's probably a more interesting question because when you discover for example well let's just toss a couple of statistics out the average organization has customer data stored in 13 different places around the organization. So there's a number for you all if you have a fewer than 13 you are above average. If you have more than 13 you are now below average well how would you use this well you might want to look at some of the technologies and Jesse maybe you can give an example of this. I don't know if you're familiar with the knowledge graph that you were talking about before, but to say you know let's go out there and do some scanning and find out oh look at this I got exactly 13 instances of where customer is. Then the question comes up, which is the one that's getting updated. Right. So how do we go into that. How would you guys approach something like that. Well, you know, the knowledge graph allows us to bring all of the different assets together. Right. The knowledge graph are the who what why when in and wears the assets themselves can have metadata attached to them. So whenever you get something like, you know, I've got x and I found x in 13 different places. Well, we need metadata about those 13 different places for that kind of data which one of those 13 is supposed to be the source of truth that alone will get you a step in the right direction. Let alone starting to flag the other ones as, you know, necessary rot cleanup attaching, you know, and indicating, you know, because of this policy and regulation right here, you know, section 1.1.1.1.12 says that you're not even supposed to be storing that kind of information anymore. So now you're down to 12. And it's it's an incremental path. New organizations. How wonderful would it be to be able to be aware of all of this from the get go. Don't get yourself in the situation of having 13 places storing the same information, but most organizations are working in the in the reverse pattern. They're going to have more than 13 places storing the same information. And we need metadata about the metadata about the metadata and then we can make decisions on that right and that's the power. Our system allows us to query that and visualize it and bring it up and build reports so that you can be made aware of that. But whenever it comes down to it, it's what metadata did you go and collect not just on the where the stuff is but the information about the source where it is actually. I great example Jesse on that I like to say that one of the responsibilities of the data steward community that I'm showing on the screen here right now is not just to understand what happens to the data when it's in front of them. But they also have to know where it comes from and where it's going. And if you have that more broader perspective, then you will be able to say, oh, we're all drawing from the same place so we should all use this as the golden source and try to get rid of these other sources and as you said you may go from 13 to 12 but that's still moving you above average and there's only 12 11 more that you need to knock out in order to go and get the rest of everyone. One is speaking the language these days of lineage and provenance and I've even been hearing things about pedigree. You know, more and more which you know was probably more of a topic 10 years ago but it's like it's coming back so, especially when you're starting to talk to the sea level individuals, when you're talking up the chain, describe lineage describe provenance and even pedigree of your data. And that helps tell the story and frames this idea of why do why do we need to get rid of the 13 or why do we want to at least be aware of the 13 that it all starts to fall in that language of lineage and provenance. Again just because I've been around for a while I've seen many times too many times where two people will be at a meeting in front of the sea level executive and they're arguing about whether sales are up or down. And you know they can't both be right sales are either higher this year or lower this year than last year but that they can't agree among themselves on that because they're not using metadata practices correctly. You certainly don't want that information going up the chain any further. Well boss we're not sure whether we're making more or less money this year than last year. Yep. Yes. So, I love this next question, a little bit of snark so you know you're talking about the dictionary metadata being a number of Gerand right is the dictionary says Gerand's and an ING so should it be metadating then. Metadating. Metadating, you know, yeah, it gets in there and we're going to have to work on it and, you know, again, we don't teach people what metadata is as part of college if you're in an IT program. So there is an opportunity here for us to redesign the vocabulary and properly get it straight so if we want to conform to the rules that maybe we ought to make it into metadata. So this thing to define this metadata, the word. Jesse, are you an English major. And no, no comment on this one. I am not an English major so I definitely can plead the ignorance on that one but seems to work in general but yes, the ING piece is an important component. You know, what are the, what tools do you need for metadata management? Jesse, I'll let you do that one right away, right? Go to topquadrant.com, joking, joking, right. Tools. You know, without without the bias of saying, you know, a knowledge graph approach, it's, it's anything, right, people excel rules the world right now. Excel is where metadata is. We, you know, it was funny, Peter mentioned access. We still extract data from access and bring it into 2022 representation and make it interactable with other information. So whatever tool you have at hand, maybe that's what Peter was alluding to in his slides. Get started. That's the most important thing and use tools and capabilities that you have. Draw pictures in PowerPoint if you have to, to start messaging the information. But then as you mature and you want to start using collaboration tools, and you want to start having controlled workflows, getting people to do the right metadata at the right time in the right place. That's when more of the advanced kind of tooling will come in and really help you. But that doesn't mean everyone needs it. You know what I mean, just being completely open and honest, not everyone needs that advanced level yet, but you will get there eventually. And that's where, you know, maybe a system like top rate edge is exactly what you need so that it can help you with all of that more advanced social side management side of everything. But yeah, it doesn't mean that you need that on day one, use anything you've got on day one. I have seen many organizations I know that several people on the delegates list at this point have built their own repositories in order to do this and yes the use of any tool at all is better than the use of no tools. And once you've graduated from that thing you have to say all right now what tools should we use so I'll tell you guys a little metadata story that was one of the big banks that I worked with. They had done a scan of their systems and even though Jesse said it can do with access and nothing wrong access is a perfectly good tool, but access has been sunset by Microsoft and so consequently this organization looked around and said, oh, we have 400,000 production access databases. Now that's a pretty scary 400,000 access databases in production. And they said, we need to fix that we need to get rid of each and every one of them and it took them 10 years. But they did it exactly the way Jesse described they started out with spreadsheets they moved it into yet another access database that was ironic on their part. Then they decided that they'd actually put the thing in Oracle because they had their metadata environment tied into Oracle so that they could actually operate their metadata out of this particular technology and it took them 10 years that they got rid of each and every one of those 100 excuse me 400,000 production access database sets, and they are absolutely certain there are no production access database. Database is running production in their environment today. Now that's an accomplishment. It was a big area of corporate risk. I'm not sure there's a dollar value tied to that in there, but I didn't know that they assessed it from a risk that it was critically important for their organization. So yes, absolutely you can use this to try and get in there and again remember metadata is use of data in a different format. And so that's all you're doing is sort of thinking a little bit differently about how you're using your existing data and how can I use that to better leverage the overall data in support of your data strategy which is supporting your organization strategy Jesse want to add something. Nope, nope, you covered it well there Peter. All right. Hey, well I that does bring us to the end of the questions I'll let you cover them all head out of the park Jesse. All right. Everyone's little quiet out there today while they're actively chatting I was right I was looking through the chat see if there's anything else that we want to cover but it's been pretty, you know we've covered the major discussions here I think. Yeah, so, Jesse, thank you so much for joining us again and thanks to talk quadrant for helping sponsor again and making these webinars happen Peter thanks as always for another great webinar. I see you've got the schedule up there and we're putting together 2023 agenda which very excited about. As always, which will be our 13th year in protection next year. And just thanks to all of our attendees for being so engaged in everything we do we just love the chat and the networking going on as always and just reminder I will send a follow up email by end of the Thursday to y'all with links to the slides and links to the recording. Hope y'all have a great day. Jesse thanks for joining us today Shannon thank you as always. Yes, thank you Peter and Shannon. Thanks guys.