 Hello and welcome, my name is Shannon Kemp and I'm the Chief Digital Manager for Data Diversity. We would like to thank you for joining today's Data Diversity Webinar, Data Management Best Practices, sponsored today by Trifacta. It is the latest installment in a monthly webinar series called Data Ed Online with Dr. Peter Akin, brought to you in partnership with Data Blueprint. 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. You can click the chat icon in the bottom middle of your screen for that feature. And for questions, we'll be collecting them via the Q&A in the bottom right hand corner of your screen. Or if you'd like to tweet, we encourage you to share highlights or questions via Twitter using hashtag Data Ed. And 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 to the recording of this session, as well as any additional information requested throughout the webinar. And if you'd like to continue the conversation and networking after the webinar, you may go to community.natediversity.net. Now let me turn it over to Matt for a brief word from our sponsor for today, Trifacta. Matt, hello and welcome, and thanks to Trifacta for sponsoring today. Hi. Thank you, Shannon. Just to make sure, can you see my screen? Yeah, looks good. Okay, perfect. All right. Hi, everyone. My name is Matt Durden. I'm a senior manager of customer and product marketing at Trifacta. We're very excited to be a part of this data education webinar series, as we feel the alignment of people, process, and technology as Peter will discuss is critical for organization success in the modern data context. Right now, we are seeing a big shift in data management and analytics. There's a shift from IT-leading data transformation to more of a collaborative process or coexistence with the business. We're also seeing a shift in the content and context of data from transactions to interactions and behaviors. We're also seeing a platform shift from mostly on-premise to either cloud hybrid or multi-cloud deployment. And finally, we're seeing a process shift from being top-down to iterative and collaborative while still maintaining a high level of governance and control. Even with all of these big important shifts, many organizations are still relying on technologies and interfaces that were first developed some 30 years ago to do data transformation and data quality. These technologies don't mesh well with shifts in data and, sorry, shifts with people and process that we have seen. This is what Trifacta has laid their focus on and apologies with the shelter in place here in Illinois where I'm based. It's hard to have childcare. So apologies if we hear a screaming child in the background. It's also well understood that around 80% of any analytics project is dedicated to preparing the data for analytics. This involves accessing the data, understanding its content, structuring, cleaning, blending and validating that data to ensure the final output meets the quality needed for your analytics project. The reason this process is so time consuming is that the technology being used is not a fit for this purpose. Either IT teams own data preparation using ETL tools which can create a massive bottleneck between IT and business or business teams are using spreadsheets which lack the needed scale, can be very error prone and have poor lineage. Or it is done with code which is often too technical for analysts and is time consuming to visualize the data, understand its content and then create a script to clean the data for those that want to use it. Not to mention the frustration of debugging whenever something goes wrong and the difficulty within the transferring those scripts to others. Trifacta fills this gap by providing the best of both worlds. Combining visual machine learning driven guidance, Trifacta makes it easy for non-technical users to discover clean structure and blend their data. Each step is stored in a recipe which can then be compiled and run on infrastructure in the cloud to transform data at any scale. These recipes can then be orchestrated through the data pipelines to get continuous value from your analytics project. IT teams can then spend their time managing the infrastructure, security and governance. While the business teams then gain the agility and autonomy to answer the changing business context quickly and effectively. And at Trifacta, we have a lot of customers that will speak to this. We have some of the largest organizations in the world using our technology, as well as startups that are looking to get a competitive advantage in the marketplace. And in keeping with the spirit of education, we also just released a new series called The Data School with Joe Hellerstein. Joe Hellerstein is a computer science professor at UC Berkeley. And you can check out this video series on our blog and our YouTube channel. We currently have two videos that are part of the series and we're releasing a third one shortly. And they're very short and consumable, so definitely check that out. Check that out. That's gonna be an ongoing educational resource for professionals who work with data, people who work with data systems, and managers who define data strategies. And finally, if you'd like to learn more about Trifacta, just please visit our website at Trifacta.com. And with that, I will hand it back over to Shannon and Peter. Matt, thank you so much. And we are a family community, so we welcome guest speakers anytime. So no worries there. And if you have any questions for Matt, and Matt will be joining us for the Q&A portion of the webinar at the end. So feel free to submit questions in the Q&A for him. Now let me introduce to you our series speaker, Dr. Peter Akin. Peter is an internationally recognized data management thought leader. Many of you already know him or have seen him at conferences worldwide. He has more than 30 years of experience and has received many awards for his outstanding contributions to the profession. Peter is also the founding director of Data Blueprint. He has written dozens of articles and 11 books. The most recent is your data strategy and I hear he's working on yet, his 12th book. So, and Peter has experienced with more than 500 data management practices in 20 countries and consistently named as a top data management expert. Some of the most important and largest organizations in the world have sought out his and Data Blueprints expertise. Peter has spent multi-year immersions with groups of diverse US Department of Defense, Deutsche Bank, Nokia Wells Fargo, the Commonwealth of Virginia, and Walmart. And with that, let me turn everything over to Peter to get today's webinar started. Hello and welcome, Peter. Welcome, Shannon. Matt, thank you so much for the intro and for setting up, I think, a really complimentary talk. Just want to say myself, I'm a very happy trifecta user and have been extremely grateful for their support over the years working towards this. So, the presentation today is titled Data Management Best Practices. But what we're really talking about here is slightly different and probably better said as practicing data management better. That's really the way we need to think about this process because it is so absolutely new when we compare it to other things that have been going on in the even IT space, much less the business space. So, I'd like to start out with my current data truths. Yes, it is absolutely true that data volume is increasing faster than we are able to process it. That data interchange and overhead costs of poor data practices are measurably sapping organizations and individual resources, therefore productivity. That reliance on existing technology-based approaches alone do not and have not material address the gap. And finally, there is an industry whose sole purpose is to extract data from citizens. I don't want to sound too much conspiracy theory here, but it's real and we need to pay attention to it. So, let's see what we're going to look at today over the next 50 minutes or so, we'll talk about motivation. Really key is that most are frustrated and unsatisfied with the current data exploitation and we're not making progress when we look at measurables on that. How did we get here? A little bit of a history on this because it is kind of interesting and important to acknowledge. We'll talk specifically about the data maturity model and how that interfaces in with the dim box. So, the real key is to apply and understand them together. We'll talk just a touch on strategy. Again, as Matt said earlier, one leg of the three-legged stool is technology that the two are people in processes. We'll answer the question, of course, how does one get to Carnegie Hall with a little bit of music, and then we're to next, and of course, the fun part. When Matt will join me at the end for some more Q and A. So, let's get started. First of all, a question, how literate are we as a society? Turns out the government's been measuring this for us. So, whether you like the measure or not, it does exist. This is the National Assessment of Adult American Literacy, and it's been rebranded recently as PIAAC, which doesn't make any more sense here. Key is that the scale is from 1 to 500. And so if we look at it from a literacy perspective, there was a measurement in 12, 14, and then one in 17. There was a numeracy also score in terms of literacy. It was also unchanged, and in particular relevance to us, digital problem solving is really key here. Again, all three of these are statistically unchanged, which is not the way to think about it when you think about what is happening to the data. Some other measures that come from an interesting source called the National Literacy Project. You'll hear more about them in some upcoming things we're going to do. 14% of people admit to having a good understanding of how to use business data. 21% of those aged 16 to 25 classify themselves as being data lurid. Future employees then are critically underprepared for data-driven workplaces. 8% of companies have made changes in the way data is used, but 90% feel data is changing them. Business decision makers, only one third feel they can confidently understand, analyze, and argue with data. 32% said they were able to create measurable value from data. 27% said their data, this goes on and on and on. None of this is good. Again, these are business decision makers that are making these statements to us. This is not a good state of affairs. So companies also are falling behind in their efforts to become data-driven. The key around this, again, the first measurement up there on the top of the screen is that the number of people who've been, sorry, number of organizations who've been claiming themselves to be data-driven is declining. So the last three years, 2017 through 2019, it's dropped on that. Gorgeous data culture, generally no. Creative data-driven organization, generally no. Treating business, data as a business asset, no. Competing, again, half, right? So conclusion of the survey, I love the survey, by the way, is the firms need to become more serious about this process. There was one little interesting piece that occurred from 19. It was 5% technology, 95% people in process. 2020, the number shifted slightly to 10% technology, 95% people in processes. So we're not doing well. Now, the question is, how did we get here, and frankly, where are we? First of all, a question that I'm always asked is, can you take our data management practices to the next level? And the answer is, well, if you don't know what level you are, then I can magically, it's not my fingers and you're at the next level. But it doesn't work that way. We really do need to understand this. And most importantly, we need to coalesce around a single measurement of this because it's time to be put aside all of the various individual private ones. I'll show you a specific reason for that in a bit. So I had a title in the early late 90s, sorry, late 80s, early 90s, called US Department of Defense Reverse Engineering Program Manager. And we sponsored research at some place called the CMMIFEI, which is Carnegie Mellon University's Capability Maturity Model at the Software Engineering Institute. No reason you should have to remember what those initials are, but it was a real research project asking the question, how can we measure the performance of DOD and our various partners that we have in here? Interesting side note, I also was told to go check to see what the Navy was up to, and that's where I ran into two fellows who were luminaries in our field, Clive Finkelstein and John Zachman, who were doing some very interesting stuff that is the subject of a completely another talk that I'll give you some other time. The SEI, interestingly enough, however, responded with an integrated process and data model that made sense. They were sensible people, it just made sense. There was literally a COTAR, however, at DOD who said, no, I can't have that stuff in there because your name has nothing to do with data, so take all that data stuff out of there. And DOD was required SEI to remove the data points of all of this. Very, very interesting stuff. That grew into what we're going to talk about today. I literally was up there at one point, picked it up, met this fellow at the MITRE Institute in Berk Parker, who ran an internal maturity measurement project at MITRE. We did a good article out of this, published a great deal of data, took a lot of different sample sizes around the whole thing. From that, it evolved back into the CMMI, and I'll introduce you in a minute to a colleague who's been very insightful and moving forward, but let me tell you why it's important that we take a look at this from this perspective of sort of focusing towards one of these models. The answer is that, first of all, from a research perspective, most other ways of doing it are worse and measurably worse, and that cannot be said about most people's assessment model. So good research evidence shows why we should do this. And the reason we can is because this friend and colleague, Melanie, has put together this wonderful piece of work that you're going to see next, which was to take these wonderful ideas and research papers and put it into something that was actually useful. So Melanie was the primary author on this. She usually joins us on these. I'm not sure what happened on this particular one, probably a schedule conflict. Anyway, we'll catch up with her if you get a chance to hear her talk on this subject. She's excellent as well. The first version of this was released in August 2014. It is a way of describing how one should take a look at each of the bits that comprise data management. I say bits because nobody has much more of a conceptual concept around them. So again, we've got a structure for each of these bits that divides it into different levels. You can be assessed at different levels. And the key aspect of this is that it describes what an organization does. And that's such a fundamental importance because if you describe what you do, you get much better results than if you describe what you think you do, or what maybe you ought to think you do, or it gets very confusing right away and goes completely off the rails. So this focus of you are what you do is a critical piece throughout the entire process. To take a look at it, we start off with a base level that everybody gets one point for having a pulse. It's a very low standard. It means that your practices around that are informal and ad hoc. However, if they are defined and documented to some level, you get to a concept called managed. You get two points for this. You get three points if you actually are standard and consistently use these by defining them organization-wide. But if you use this data that you have standardly and consistently defined and measured to manage the business, you can now start to improve what happens. So you get five points if you actually use all of this data to help improve the way you go about this. This model is not unique, but it was invented by Carnegie Mellon. And it forms the basis for TDQM and ISO 9000 and other types of processes just to take a two-second deviation to ISO 9000. If you get to ISO 9000 certification, that would have meant you were defined at level three. So that's not a very tough standard. That says that you do what you say you do, which is good. People would like to have that, but we really want to get a little beyond that. So there's a bunch of certification around all of these pieces. Again, best person to reach out to is Melanie on that to give you an idea of where that goes. This gives us a solid piece of research to use. Now let's take it a step further. This DMM gives us these five practice areas that we use. The first one is called strategic management. It doesn't make any difference if you're working hard, if you don't know where you're going. So we're gonna manage the data coherently. We wanna govern these assets with professional fields. We now have enough balance behind us, enough past history, enough experience that we can say we have a class of data governance professionals that exist. That data operations now can be conducted according to a standardized data management task. That it can be maintained fit for purpose and using the right platform and the right processes around all of this. These are the five practice areas in data management. Now you need some supporting processes here as well. And we're gonna make sure that everybody acknowledges that. But there's another aspect of this description and that is that the data foundation is only as strong as the weakest link. So what that means is that if I have the scale that I've described to you before, initial, managed, defined, measured, and optimized. And I assign those to the various processes, the practice areas around here. I'm gonna give them all three by the way, just to make it very easy for us. So platform and architecture, data operations, data quality, data governance, all get a three. The platform here, all the efforts in this organization cannot be any stronger than a one because they don't have a strategy that will take them in a single direction and allow them to focus their efforts on it and make it an object of practice. We'll get into that a little bit more. But in practical terms, this means literally, and I have seen it happen, organizations could put hundreds of millions of dollars into something that they call data quality and it won't help because the foundation is only as strong as the weakest link. Now, this is actually the place I'm talking to you from, my house in Montpelier, Virginia. Amazing, we have internet, Shannon and I were laughing about that a little bit earlier. The key, I just wanna show you that when I was building this particular, oh, it shouldn't take me, I didn't build it. My wife designed it and I paid for part of it with a bank loan. Interesting thing about the bank loan is that the bank gave us exactly this much money to put in the foundation. And then they wanted to stop and see the foundation. They did not wanna be to build on top of a core foundation before any further construction could proceed. Now, this is a very smart way of doing business put into the bank and there is no IT equivalent unfortunately for it. So not having these foundational practices, again, having it only as strong as the weakest link gives us a real problem, but it does give us some ways in which we can measure how we're doing collectively. So we have on the left side here the five data management practice areas from the DMM and the five levels that they can be performed at. And when we put them together, you get sort of heat maps as a sample. Heat map doesn't represent anybody's particular areas obviously much more detailed. Sometimes organizations like to compare themselves to other parts of the benchmark and see how they're doing to compare it to the banking industry or something or here's one done for the insurance industry. Again, they can be done because you have this kind of standard. It can't be done if you don't have a common standard to put them in place. I also always end up asked for this. So when I'm working with a group of executives, this is a airline. And yeah, I'm looking at this and going, oh, ones and twos, I don't understand this. Okay, we'll just wait. What's your competition? And they go, oh, we're one and they're two. All right, now we need to know what that means and why that's the case. And you put the third piece up, compare them against everybody else and you can see they're behind competition. And again, before you can take any of the twos and make them into a three, you have to take all the ones and make them into a two. So there's some very definitive challenges that can be objectively identified, measured as you apply additional resources to it to try and bring up the performance around that. I noticed they didn't tell you which airline or any of the other companies here, but the International Finance Corporation did tell us we could use theirs. They had asked us to come in and look at their treasury group, which was a pretty straight no, their information systems group, but their IFC, the business portion of it was actually world-class leadership. So one of the other measurements around this, one of the other takeaways is that you should really walk down the hall and find out what your colleagues are doing. Sometimes they don't need to be an IT. And this literally when I said world-class, they were better than the imperative averages as well as the overall averages around all of that. So this comparison of data management maturity, you haven't done a whole lot of this type of stuff, but it just isn't changing. We're not getting smarter about data and yet data is becoming more ubiquitous to say an Holly easy thing. So again, I'm making a plea here while all improvement processes begin with an obligatory phrase. This one, the MMI, the MM is the only framework that is proven to give the benefit of literally decades of practice and benchmark. Organizations not using this one run the risk of not being able to compare the results meaningfully and not being able to contribute to the overall takeaway with all of this for the organization. Now, when we talk about the strategy around this, I like to shift the discussion into something called the theory of constraints. Many of you have read a book called The Goal many years ago that incorporates this into it. And what I want people to think of from a data management practice perspective is they haven't been doing most of these practices to the degree to the refinement that they'd like to in here. So this is a way of practicing that process. First identifying the component that is limiting goal realization. What in your entire environment, it was one thing that if you eliminated it would make things better. He is to focus on that only, on that one constraint, making quick improvements to the process so that you can really focus in on that aspect of data management and subordinate all the other projects that you're doing to that one. Because if the problem still persists, you have to repeat the process until it is removed. Always going to be this fashion. And what we're really trying to do is to apply that process in a manner that we can get better at. And I like to talk about the knowledge of a better data sandwich. Because most organizations have sort of in their inaccurate literacy and even supplies of data, not real use of standards around this. And so our job as data professionals is to come forward and standardize these, put them together in a way that will make them work smoothly. And this cannot happen without engineering and architecture concepts that are simply not taught as part of technology work. I had to go to a tea farm in India to see the stemming quote where it said quality engineering and architecture work products do not happen accidentally. No, they don't. And of course we had the word data into all of these. We understand now why it's so important for all of us in the data world to take a perspective on this. So let's see how this works with the DIMBOK here. DIMBOK one or two, both of them work, right? I'll allow us to look at things from again this perspective of dividing up 12 areas. So the best way to think about it from that perspective is to look at it as organizations attempting to get better with one or more aspects of something on the wheel here. Whether you use this wheel or the five that were paired piece, the key is by not focusing on everything else, you're going to be able to make measurable progress in that area. And so I like to tell a story around this of a person that had trouble kind of getting this and the story goes like this. Why aren't my data probably isn't aren't while it is good? Why are our data problems solved by a data warehouse? So somebody told me a data warehouse would fix these problems. I got the data warehouse that didn't fix it. But what happened? Well, in this case, just with most data warehousing problems, they got to the very end of the project. They ran out of money. They treated it like a technology project. There's no money left over to do any kind of architecting of the data. We instead forklift the data into the warehouse. There are a couple of problems with worklifting the data. One, there's no basis for making decisions about including or excluding stuff. There's no inclusion of architecture engineering concepts in the discussion. There's no idea that these things are even missing from the problem. And 80% of all the data in the organization is wrought. Why would you want to put it another version of it in the data warehouse? So I like to instead say there's three rules of warehousing data that really applies to the transformation that should occur in this case. Transformed data in the cloud should be less in volume because of the rot. The 80% of your data that is redundant, obsolete, or trivial should be filtered out on the way into the cloud rather than simply forklifted in there. It will make your exposure to everything a lot less and it'll be a lot cleaner. Now, data in the cloud should be cleaner than data outside of the cloud because the opposite can't be true, right? You wouldn't put up with data in the cloud being less clean. And so you need to go through and clean the data on the way in as well as transforming it into less and then making it more shareable on top of that. Now, the reason this particular approach to data management didn't work, unfortunately, was while this individual was trying to do this correctly by applying data warehousing a better approach based on an analysis of the failure here would have been to try and get a little better with data quality, data warehousing, and data governance as a part of it. And those three pieces form the first version of their data strategy. That gives you the ability then to focus in on one specific area by three things. Four is probably too many and two is probably not enough. I hate to do that. The second version of the same thing maybe we discovered that warehousing is good but the next thing we needed to do was look at metadata. And now the nice part about this is that we've gone through data warehousing and it governs twice. So we're twice as good at those areas but the first time we're attempting into metadata areas. So we'll get some practice around that. For three, again, that third time through data warehousing we're getting good data governance. Also we get a three in that cycle as well. So now we're getting our first exposure to reference and master data management there. Not that these would solve the problem. And of course you see one of the things that wasn't included in any of this was the idea that we should be looking at data governance. So we like to use the term lighthouse project to describe this. The idea would be that from a lighthouse you've got a lot of things that can help you in terms of strategy. You've got a lot of data that's in use in the business that needs to be improved and you've got a lot of needed data skills that your organization wants to find. Let's find one piece that focuses right in on that section that gives you the best of all three worlds. Allows you to exercise three different parts that takes your measured performance in that area on the DMM scale from a start to a better type of a process. A place where you're not going to be absolutely at the very beginning of it. This will be guided by an organizational strategy. Very, very key to understand. Most organizations think of the strategy as the product. The strategy instead must be thought of as the process. Thinking of it as the product puts too much weight into it. A strategy that sits on the shelf. I can't tell you the number of data strategies I've looked at that number, excuse me, that have hundreds of pages that are associated with them. We need to take a step back and say, what do we mean by a data strategy? And the idea is we're improving our data management practices, the data strategy ought to be the thing that focuses our attention on the thing that's blocking us the most that will allow us to proceed and use data in support of organizational strategy in the best, most effective, most leverageable way that can and should evolve over time. If it doesn't, you're in a very boring industry. I've never seen it not change over time. It's just been crazy. So here's a quick example around this. The good guys were up on the top of the hill and the bad guys are down at the bottom of the hill. So we'll have a different strategy than if we had the tables turned. The bad guys were at the top of the hill and the good guys were at the bottom of the hill. So that ability to do something from a strategy perspective is absolutely crucial. And the key is that Eisenhower had a wonderful saying, in preparing for battle, I've always found that plans are useless but planning is indispensable. That process of approaching what you need to solve from a strategic perspective is absolutely critical. The definition that I prefer to use for strategy is a pattern and a stream of decisions. Now, I'm gonna describe to you three different business strategies at this point that will give you an idea of how to pull this into a little bit more usable perspective. So if my organizational strategy is to provide the lowest price to the customer, that's the way I want to be known, that's our messaging, everything that I'm doing around that piece. That will allow me then to take aspects of my data and see what data could give me better customer information that would give me better prices for the customers. Usually in a context like this, you're changing the cadence of the reporting. In some cases, the reporting is monthly and you go through it daily. In some places, you over-correct and go to an hourly which overloads your systems. But some aspect of this becomes the focus. And so the strategy becomes around trying to take things that you can do that will get that data to that cadence that you're attempting to get it to. So that would be a process that you could use to help focus what your activities are around the data strategy. That would be perhaps the lighthouse project that we talked about before. Given that data strategy then become this pattern in the stream of decisions, what we need to do is to back away from the idea that we're going to create the next five-year plan. And again, I've seen this happen over and over and over in an organization. Do this, think about where your organization was three months ago. I don't know about you guys, but I was telling my undergrad to a class that they were facing the best job market that they could have faced in my lifetime, at least at this point, it looked terrific for them. And all of a sudden they had 30 million people in front of them online. These plans are much more about how we adapt than we are about what we're going to do given a set of circumstances on this. So when we look at what's happening in our data environment, what we're attempting to do is to figure out what projects that we can do under the guise of a program. This is the other part that has been so extremely difficult to implement. Your organization needs to understand that these activities that you're going to do comprise a program in the same way as your organization has an HR, the Human Resources Program. The Human Resources Program is not something that we would have absolutely started on as a place to go and emulate, but it turns out to have a very similar history about 70 years ago. Thank you to John Lathley for drawing the story to my attention on this. HR was done by each work group. And while that was an interesting approach, it wasn't standardized. And with the increasing amounts of legislation that organizations were facing, it made sense to look at it from a standardized perspective. Data management should be thought of in exactly the same way that you need to develop a set of capabilities in your organization, such as you have people who have expertise in these areas and can go in and solve and improve data problems in a way that is meaningful to the organization just the same way as if Army has a group of soldiers that it can use to go achieve military objectives given that particular piece. So the idea here is to find something in your organization and try to develop light size projects that you can use to go in. And this then becomes a pattern in a stream of decisions where you're able to take your expertise and say, I know how to fix that because I've run into that situation before, but you can't do that from the start. So the way you start is to get practices. Let's also talk about the piece that Matt added as well here. Again, I appreciate the comments earlier because technology is part of a three-legged stool and it's an important part. I said earlier that you can't do what we're attempting to do as professionals without understanding architecture and engineering concepts. And these architecture and engineering concepts are not in our existing curriculum. Your people don't know that they don't know they need to do with them. We need to work on the people in the process sites because actually the technology side works pretty darn well and that's terrific news but it's wonderful to have partners like trifecta being able to contribute in this environment. And again, let me just go back to what Matt said. The idea that you're paying people who are very expensive to do things that shouldn't take them a long time is probably not where you want to spend your money. So looking at that 80% of the time that each one of them will tell you to a person that's the way they spend their time is 80% of their time munging and 20% of their time actually doing the data science work that you wanted them to do. You're able to take that and reduce that by any amount at all. Even let's say that you make them from 80% totally unproductive to 60% unproductive. You've doubled their productivity. We've also though not done nearly enough around the people in the processes. And I'll just come from a strategic perspective from the data management perspective. I had so many organizations come to me and talk to me about whether we should introduce like 14 levels of data strategy, excuse me, data steward into the organization. And my advice is always know just the opposite. First you should introduce the concept of stewardship and say these are data stewards. Their role is to advocate for the best use of the data in their purview to understand its sources and its uses in the context. And that will help everybody understand better the organization's data because it's too complex to be understood by one person. So a great component. We're doing more with that. We haven't done nearly as much with processes but this is where the DMM comes to play. If we do them according to the DMM do the five categories that we're talking about. We will be able to get and we have lots of organizations that have made progress just in these areas by improving their practices around this following this standard method. And gradually according to each, we've aired, if you will, on the side of many organizations checking a box by saying I bought a data warehouse. Well, that's not good enough. One of my favorite concepts is the concept of garbage in, garbage out. And the concept of garbage in, garbage out is pretty straightforward no matter what you've got. If you're feeding it with rubbish, you're still gonna get rubbish on the other end of it. And it's a simple lesson to learn but unfortunately it's one of the most basic. And part of the process should be that we should get the data in better shape before we should invest significant amounts of organizational capital in a technology that is potentially problematic at best in these days and days. So again here, the relic at school is what can I do to look at this from a balanced perspective? And I'm gonna add one with a point to the people part which is the Chief Data Officer role. And we do need to have somebody who's in charge the organization as an HR organization wouldn't function at all with somebody not in charge. There's gotta be one person up there, one person's noose, neck in the noose is another way to think of it. But it is important to have that leadership without it. It's very unlikely that the organization will be entirely successful around that. So given these three pieces pulled together here, how do we get organizations to change? That really is hard because we're talking about absolutely very, very fundamentally held changes. Things that we have taught the students incorrectly for years and years and years. Imagine the only thing we've taught them in the community is you need data people to build a new database. And what did they learn? I don't need data people to do anything except building a database. So we have published papers that have shown that data people used to report to the top of organizations. And now they tend to report three levels lower than they did 20 years ago. Is that a trend? I can't tell you yet because we haven't had funded research in those areas because it doesn't seem to be important for some people. All of these kinds of things that have gone on. And we do have some measurements here. But again, I started off the session here by showing you that we have so many people who have been trying to call out for this and to say, we really gotta make it changing it. We're not changing as a society. Data is increasing and our ability to process it is not increasing at the same rate. We need to understand better what we can do in order to do this. And again, just as HR was not considered critical at that point, now you cannot imagine the organization saying, well, I think we're done with HR. We hired all the people we're gonna hire. Maybe that's true, hope not. We'll pull through the economic piece here as fast as we can on this. But goodness, nobody's gonna let you get rid of your HR group. We'll no longer need HR when you go out of business and you will no longer need your data management capabilities in your organization when you go out of business. So what does change management leadership do? Well, it allows us to come in and take the business, the behavior that we have in the past and change it to the behavior that we wanna have in the future. And these are good people who really do understand this. It's a wonderful piece and many organizations have internal capabilities that are able to do it. If not, you can certainly rent them on a very good basis. And it's absolutely something that you should be consulting. And they are also, by the way, will help you with the messaging component of this as well because the messaging component is absolutely critical from this perspective as well. So when we look back here and say, how are we going to get organizations to actually make this change? I found this scale to be very, very helpful in the past. This was developed by Mary Lippert in 1987. And when I walk into an organization and I see vision and I see incentive, I see resources, I see action plan and I see anxiety. I know the thing that's missing are the skills that they haven't been given the proper training so that they have comfort that they will know have the abilities to be able to do what they're attempting to do. Similarly, when I walk into an organization and I see skills, vision, incentive, action plan and I still see frustration, I see really good people who know what they need to do and have the skills and knowledge to do it, but they don't have the resources to do this. And this diagnostic technique has just been so useful over the years, but it turns out to work on exactly the same plan that we were talking about as well. In other words, it's only as strong as the foundation and according to Dr. Lippert here, change only happens when you have vision and skills and the incentive and resources and the action plan. Do you actually get change? You miss any one of those ingredients just like a key in a lock and you do not get success. And culture is absolutely the biggest impediment to shifting organizational thinking about data because they don't know what they don't know and frankly, we've got to do better as far as communicating with them. So I've written a bit on this and I wanna just draw your attention to a case study. It was kind of an interesting one. I've had 12 organizations come to me so far and say, well, you really captured us perfectly. Figure that's a compliment. So it doesn't cost you anything. We paid for it so everybody can have the downloads here and we've gotten some very good feedback. Hope you guys enjoy it as well. It's not again the topic we're doing here. What we're gonna do is say now one final thing about strategy. And that's the idea that when one looks at strategy one shouldn't in fact be tied to the idea that it's a working plan much more represents these capabilities that we're trying to do. So I'm gonna tell a little story here. If you can't tell already, that's a picture of Bruce Springsteen and has a wonderful video on YouTube. I'll give you the link there. Please enjoy it because it's just a wonderful piece of music. Were we able to join each other's live? I would ask you guys to raise your hands when you recognize the song. Now the interesting part about this is a little bit of history that happened over here or somebody was telling the story of this and Bruce Springsteen is playing Australia for the first time in a number of years. He's excited to come down. He wants to honor the Australians who are there by doing something that is an Australian song. And he thinks to himself what other song would be more Australian than a song I'm gonna play for you? And he turns to his band who probably never played this song in their entire lives likely when they're on the airplane and said, can you guys work this up for me tomorrow? This is the result. So just to take a minute here and watch what Bruce Springsteen does. Hi everyone, I'm Bruce Springsteen. I'm here to hold my shoes on. I'm just gonna hold my shoes on. It's okay. Part of this post is really a recluse that you need. Yes, stay in alive. Just to give you a clue to my age, 1970 to 70 year I graduated high school. That was my most hated song in the entire world because I was a rock and roll person and that was disco. And of course I was young and stupid. So here we have now the greatest musician in the world playing some of the greatest music in the world. And I never would have thought that combination was enjoyable, but I would be dancing around the room if I wasn't doing the webinar right now. Why am I telling you this? Because Bruce Springsteen didn't write a strategy that said in 19, gosh, I forget what year this was, 2018, I'm gonna play Australia and I'm gonna be able to play Stan alive for no. What Bruce Springsteen said was, I wanna have a band of musicians that I could snap my fingers and they can do any song that they need to do. And of course that's exactly what he has put together in order to do this. Which do you think would get him the furthest and fastest? And the answer of course is this. So having the music is great, but trying to write your data strategy the very first time is kind of like trying to write the world's best pop song on your very first try. It doesn't work that way, but only with practice, practice and practice you actually get the results that you're looking for or some really great music around that. So let's take a little bit of a step back here and sort of see where we've been. The idea was, are we happy with the way we're currently exploiting data? Hopefully nobody is. I know very few organizations that would raise their hand and say, yes, we're totally happy with it. And by the way, I wouldn't be on the webinar today, would they? The other question is, are we making progress? Yes, technologies are getting faster, but you've seen the technologies are only part of what we're attempting to do here and they can only fill a certain part. We need to take and change our concept of data literacy. There's no longer just data that needs to be data literate, but all the way down to what I call mobile data spreaders. These are young people who are just really, really good at surfing the web with their mobile devices. They need to be data literate in a way that we haven't considered before in order to do this. We can't make progress if we don't know how to measure it. So we have a scale that has been around for two plus decades, has hundreds of research papers that have been focused on it and has been refined by some of the very fine minds that are in this area. Let's use that scale and let's continue to build on that scale. And the vendors come to you and say, yes, I'd very much like to give you, but you have to use our scale because ours is else the one that's ours, right? I was like, yeah, you actually do need to step back and say, let's compare ourselves to the way the rest of the world does this. Why wouldn't you use the organizational measurement assessment that has been around, has hundreds and hundreds of papers written on it? Repeat myself, sorry. So we take these two ingredients, the data maturity model and the measurement part of the model is a fantastic way to evaluate how you apply the different aspects of the body of knowledge. The other part about the body of knowledge is you probably are not gonna be able to apply them individually in isolation and trying to do everything simultaneously doesn't work either. So it seems the groups of three seem to be the best in terms of allowing us to take a focus in with a set of concentration efforts here that allows to do more of this than other things. We're not saying exclusively or anything else along those lines, but very definitely focusing in on it and practicing. And I outlined a very simple straightforward process for looking at three aspects of the phase of a data warehouse and looking first of all with different focuses on the warehousing and the governance pieces. They got to be practiced three different times in there. So hopefully the organization by the time it's done a third cycle through either data governance or warehousing was getting an idea of whether it was doing it well or not. Finally, the thing that provides us gives us the direction is understanding what the organizational strategy is. If we do not understand and cannot figure out how to support the organizational strategy, we're in the wrong job. That is one of our primary functions as data professionals. So taking the organizational strategy and improving how data is used to help the organization achieve this strategy is a key piece and it's not about planning. I have experienced many, many organizations attempting to do this. And did anybody plan for where we are today? No, of course not. Just use one example, but it is absolutely useful to develop these capabilities. So even if you don't wanna buy into all of my raving on that one, at least build up some capabilities, give people some practice. The best way to think about it is that the capability of data management is focused internally. And when it's looking around its environment, it's trying to identify, diagnose, figure out how well certain things are done, where it could be performing. And then to suggest investments that would allow the organization to invest X and get back something beyond X6, beyond, that would show a positive return on that investment and help the organization achieve its strategic objectives. As we move into that, we've got to be able to say, how do we figure out the focus there? Now these things in groups of three, understanding that strategy is a capability, not a plan. And that this is part of a stool, but definitely not the entire stool. Because if we only focus in on the technology leg and check off, we probably will not have a completed set of balanced requirements that we need to understand. And again, these are organizations that we're reporting themselves at the beginning, saying we are not making progress in our attempt to be data-driven along these lines. So how do you get there? You practice, and you keep practicing, but you do it in an organized fashion. I hate to sort of tell the story, but it's a true story. I was in a grade school at one point and I wasn't going to pass the one grade into another unless I learned this song on the recorder, how to play the recorder. And I did not fancy it. I was a bass guitarist. I wanted to play rock and roll. I had no interest in this classical stuff. And finally I realized that if I didn't do this, I wasn't going to pass. Like, okay, got my attention. So what do I need to do? Well, I got up an extra hour early in the morning and I went out into my mom's sewing room and I would practice there with a recorder. Ugh, didn't like it, but went from last in the class to best in the class. Just a matter of practicing over and over again. Now I won't pick up a recorder today and try and play anything for anybody, but the practicing was absolutely critical. And hopefully it helps me when I play bass as well. Well, let's talk about where we need to go next because it is a little bit consequential. Big changes need to occur in our organizations from a data management practicing perspective. We need to show them that in our organization, the data volume is increasing and it's increasing faster than we are able to process it. You need numbers. You need to be able to sit down with management and say it was here last week, it's here this week or whatever it is. A great source for these types of presentations is a group called DOMODOMO.com. They do a wonderful job at the end of each year. They describe how many minutes people were watching Netflix during the year. I think it's something like 700 hours a minute or something like that. I'm sorry, I can't remember those right off, but they do a great job of doing it. And you can bring those numbers home and see how they are relevant to what you're doing in your organization. Data interchange and overhead and other poor data practices are now measurably sapping organizational and individual resources. And that hurts productivity. Okay, people will buy that if you put down and show them some examples. When we're doing data modeling, we always say, if somebody suggests a new term, give me three examples, please. And we say one, two, and three, and then, okay, we've got it. That type of thing generally works. If you have three examples that you can describe to people about saying, well, look, somebody has to redo that report almost every month. Each, if we take it as only half the time, it still adds up to a waste of $200 here. Little small things like that, they can add up, they will add up over time. And organizationally, we have to wait until we get the data right before we can release it, which means that we're not serving the customers of the data in order to do this. Very big change. Technology-based approaches solely are not working. And we have to change our educational methods as well. So these are another set of constructs that we need to put in place and think about how we're going to change that. Again, I'm giving you a look ahead already. Our approach is going to be, we need to start educating mobile data spreaders. And finally, there is an industry whose sole purpose is to extract data from citizens and use it for to make money. It's not against the law, so they're going to continue to do it. You've probably heard about some of them, that scrape pictures and license plates and all sorts of things that are out there that you can find and try to monetize on. So the questions we have is let's talk, first of all, and say, as we're trying to get better at data management, process is more important than results. We literally have to practice, like getting in the water the first time, whatever type of analogy that you'd like to think of as well, make it an analogy that everybody in your organization can understand and attain to around there. Second, failure is itself a lesson. No, you don't want to be Thomas Edison and fail making the electric bulb 9,999 times and get it right on the 10,000th time. Yes, he was optimistic and said I've learned 9,999 times how not to make an electric bulb, right, your management's not going to pay attention to that. But when you do make a failure, you're not any worse than IT. IT delivers one third of the projects on time with full functionality. In fact, if you partner with IT correctly, you can actually help each other. They will say we can't do it without the data people and the data people say we can't do it without the technology people. Failures in and of themselves are very definitely lessons in these areas. And again, the people in the process aspects are not receiving enough attention. I guess one of the reasons I was excited to see the trifacta sponsoring a data science course online that they can talk to us about how all of these technologies can be used because one of the things people say is trifacta. How do I actually, is that a cookie? I don't want to have trifacta. My wife considers this a success when she's out working the dog, so I'll come back and say trifacta. Anyway, move on. So we just spent the last hour talking quite a bit about motivation. Frustration's here, not making progress. There's a good base of science in this. The ingredients are a combination of the dimbox and a combination of the maturity model, checking out how to apply them together in those areas. So take us back to the top of the hour and just a quick advert on the other events we've got coming up in June. We'll be doing data governance strategically. And in July we'll be doing data management and data strategy for our ability between the two. And hopefully we'll be back in person at EDW in Chicago on 22, 23 October. That's just when my sessions are, the entire conference will be the week. And now Shannon, I turn it back over to you. We get our Q and A. Peter, thank you so much for another great presentation. And indeed, hopefully we can rejoin together again in person sooner than later. But I love this, just to answer the most commonly asked questions, just a reminder, I will send a follow up email by end of day Thursday to everybody with links to the slides and links to the recording of the session along with anything else requested throughout. And you have questions for Peter and Matt. Feel free to submit them in the Q and A in the bottom right hand corner. So diving in here, for organizations starting their data management journey, which foundational capabilities proved to be the most valuable to focus early on for the business side, not IT side. The really interesting part is that we've been taught and still teach how to do things technically and to clear they are successful. What that leads us to is a situation where we can say these data things happened as a result of my responding to your request. Unfortunately, we need to stretch ourselves a little bit further. And maybe Matt has some examples of how he's seen this occur out in his wilds on there. We need to make these data things occur, then very clearly have business things that occur as a result. If we don't have that tie between data things occurred and business things occurred, it becomes very difficult for people to understand exactly what happens when they push the buttons on these various operations that we're doing. Matt, give you any setup? Yeah, and actually, prior to coming to try it back that I was an analyst myself, and I think one of the key things as a part of this is as data continues to grow, and there's just more data than ever before. I mean, no matter what industry you're in, you've got phones that collect data constantly, you've got watches, it might even be embedded in your shoes. There's just more and more data by the minute and second than we've ever seen before. So having IT be dependent upon building everything when a lot of times the business user, like that was my case, like I was the expert in the data. IT necessarily wasn't the expert in what was in the data, but or the value of the data, but what the data was itself as far as schema and everything else goes. But allowing me to have the ability to report out that data and tell the story and make sense. And this is something that happens across a lot of our customer base is that that allows IT to focus on other priorities which helps then drive the goal of the organization as a whole forward and then allows the business to continue to innovate and be proactive versus reactive. And it's just this coexistence that's starting to become more and more the norm across most organizations we work with. I can remember back when I was young and naive, Matt, I used to think I might want to be a CIO and then I realized that the CIO's job involved saying no, no, no, but we're doing less with more each year. So it doesn't sound like a prescription for success but I appreciate how they're able to succeed in that field. How do you deal with industry benchmarks where companies are using different maturity models? Is there a value in comparing apples and oranges? In some ways, many of the, if you will, competing pieces that are out there are very similar in terms of their measurements. So if you're able to look at them and see that, say yes, the first phase is all about whether we can repeat the thing, whether we can do it twice the same way. I have a wonderful friend, Norman, who I've played guitar with for more than 40 years. And Norman, I call him a Keith Richards guitar, he cannot play the same song the same way twice. It sounds brilliant every time he plays it and it's so exciting to play with him because each time is an adventure through it but it frustrates him. And so he will not be that person who will ever get to a number two on that scale. On the other hand, he's one of the best performers you'll ever see. I'm not sure I went off in that direction, but I guess the idea was that if we don't have an idea how we're trying to do this, then it makes no sense. So organizations who are doing this with other scales, yes, there are a bunch of them. Some of them are quite good and put a lot of effort into it. And if you look, you'll see they are somewhat similar. But other organizations are simply not helpful in here. They may be helpful internally, but it's just not, I hate to be so pedantic about this, but it really is incumbent on us to say, we need to be able to make these comparisons. You guys wanna know, how am I doing? How should I be doing? What is realistic? By the way, my numbers are one in 10 organizations tries to do this well and one in 10 of those succeeds. So if you're having any success at all in these areas, consider yourself in a position to be able to take a sustained competitive advantage. And that should be the conversations that you guys are having. Matt, I'm not sure I left you anything there. You wanna add about comparing the apples to oranges and things. Now I think that's more of a, I think you summed it up pretty well and that's probably following your expertise better than mine. Melinda, it is a source of frustration because we should at this point now say, having this data is so valuable to us. I mean, again, there's lots of other things we have worried about in the world today, but this is really good stuff that we can look and learn and learn how that, and what they did is they made a general improvement model. So why not use a general improvement model to improve? I mean, gosh, it was reasonable and then trying to do proprietaryness around this does not make any sense when you consider where everybody is and how much we have to do. Sorry, I'm on the soapbox, I'll get off. I love it. So next question here is, we are currently working on putting in place a data sharing framework for data in our HR system to various parties within the company. Obviously there's a lot of sensitive data in an HR system but we are trying to enable the business with their data needs. Right now we have to jump through a lot of hoops and get a set of unique approvals based on each request before sharing the data and we'd like to streamline that while managing risk. Do you have any advice? But it surprised you to know that you're not alone in that particular set of frustrations. Interestingly, something worth perhaps Googling or maybe we can send the Shannon, I'll remember to send it to you, but. The federal government recently passed a law saying that that type of behavior was counterproductive. Now, you're probably not governed by the federal law but gosh, if the federal law says it's the right way to do it, maybe there's some guidance there that's the right thing. What FIFA describes is an acknowledgement that doing this and requiring one-offs makes nobody rich but the lawyers. I'm not beating up on the poor lawyers but let's just think about what happened. An agency or a part of your organization wants to get data to another place and they have to go through some rigmarole to do it. Most people will say, I'd rather not. Then they hear things about data agreements and they go, well, I don't know anything about data agreements and I'm not qualified so I should stay away from it. Yes, the whole purpose in FIFA was one of the titles, title three, was specifically constructed to say that three agencies and I think it was the Bureau of Labor Statistics, Treasury and Finance maybe or OMB. I forget, don't quote me on that but they operated so intently with each other and were governed by such burdeness restrictions and requirements to do things that the law explicitly says you guys will be better off forming one agency to serve all three of these groups. Now that's interesting but it also is recognition that this is a better way to do this and that we should not be looking at one-off types of situations very similar to who we should not be doing one-off security type functions either. That's a different question we can go off in those areas. Again, Matt, I'm sorry, I'm not sure I left you anything there but given I guess, well, I don't know, just go ahead. Let me guide you. No, I mean, yeah, I think you covered it pretty well. I'm trying to think of anything that we could add to that but like that doesn't come off too commercially, one of the ways. Let me go ahead and be very practical here. One of the things that people don't know when they get a hold of data is what am I looking at? And one of the things that profiling technologies in general of which trust fact is a really good example will let you quickly and easily look at things and verify and determine whether you have the right or the wrong or a version of things. So that's just one way in which that that can be very helpful around that. Right, yeah, I'm glad you said that because I didn't, if I would have said it, it would have come across as more of an advertisement but like, yeah, that's one of our key components to our platform is making sense of the data initially and visually and then also I think for any product that's out there, I mean, there's so many technologies to wade through when you're considering what to make a part of your architecture stack and what to use. And I think any of us need to be able to adhere to security and governance and that's a huge focus for us to is, you know, if somebody's using trifacta that they don't have access to data that they shouldn't have access to and we integrate natively with whatever security is already put in place. So that's one of the key things that we really focus on as a company. Cool. Sure, earlier it seems like you were stating is that rather than building the foundational structure needed for data management, what we have done across time is simply buy the latest tech hoping it will solve all current and historical issues. Is that true? Somebody's setting me up for something that we've over relied on tech, yes. And more importantly, we've taught people that tech is the only answer and that's really not correct. The biggest answer is that we need to unleash knowledge worker productivity. And if we can start to take not just the data management profession but all professions and make them more productive, we have generally good things that are gonna happen all the way around. We'll get into prognosticating too far. Matt, anything you wanna add on to that or comment on that? No, I think that's more a call from the Peters wheelhouse. I'm new to the Bureau of Land Management and that's one of the only data administrators in the state. My first task out of the gate was to create a state data management plan. I attempt to speak up and share my preferred approach of working with one, with just one program. I didn't gain any traction. They want a written plan. Got any references that I can cite. My plan is to exceed the state and hopefully the focus. Any references or assistance you have or recommendations for her? What I think we wanna do is put them in touch with a group called, I forget what their latest name is. It was just a group of state CDOs that we can put them in touch with that will probably be very willing to share information because government to government, there should be very little need to replicate this stuff and to redo it. And that's exactly the type of savings that we need to do is to share best practices. Shannon, that's one of the reasons you have such a community around this is because people are very interested in contributing and sharing. And I know that whoever set me up for that technology question is trying to fill in the agenda on that. But there are so many things that people are just unaware of and that we haven't done a good job of making them aware of so far they don't know that they don't know. And that's a sad thing. I mean, again, just very practical here. How many times have any of you looked at a data set and said, oh gosh, the date's in the wrong format. It's MMDD, YY, and I need to make it YYMMDD to conform to name standards or something in there. By the way, maybe the wrong information by doing that. You sit down at the spreadsheet, you start coding the stuff, you can just obviously with in the office that you can go and ask a person to get that done. Some tools that will make this much, much easier and much, again, just the idea of having to get up and go walk someplace inhibits a lot of people from actually having to do it. Just taking any actions along those lines there. So if they know they can do these tools themselves, they will actually make these improvements and figure out not spending all their time fixing them. They'll figure out what the cause of these things are and go upstream and prevent them from occurring so they don't have to do the rework. So they do it, but they can't do that if they're swamped underneath all those. Many of us are reverse engineering data quality into systems that were originally developed with no consideration of data quality. What do you recommend to those of us in this position? How do we justify the resources that we need to accomplish this? I sat on a panel last year at one of the events that was dedicated to the proposition that it was impossible to put a price on quality. And I stringiously disagree with that, which makes me a popular panel person because I'm vociferous in my views around this. If you are causing data quality problems in your organization, you have what our friend Tom Redmond calls hidden data factories that are going on in those organizations. And those hidden data factories are sapping productivity of your knowledge workers. They are doing things that you don't know about. They're undocumented. If they leave their jobs, somebody else has to do them. Probably people across the cube are doing the same thing, not coordinating the results of those bits and pieces. It is a huge, huge problem. And that is where we need to focus our reduction in non-value-added work is to look for tools and technologies and processes and training that can help organizations figure out what their people are spending their time doing and use appropriate technologies to go in and train them. Again, once you've been trained on these types of bits and pieces, you know better how to use them. You know better how to prepare your organization for it. It's not required that everybody get there, but they are easy enough that anybody can get there. So finding these business cases, building them out slowly and surely and saying, look, I'm going to invest. I can tell you the story of this one. That is a company that did direct mail sales. They were getting very bad sales lists. 30% of the names they were getting on these sales lists were incorrect. They didn't even have a serious level agreement in there. They were just simply supposed to provide data, which was unhelpful. So we started analyzing the data that was coming in and looking at how that data could be used in the organization, found other uses of it for some aspect of it. And more importantly, had a lot of evidence that the organization wasn't providing what it said it was going to provide. And that was done at a set fee. And so there was a clear value proposition that was able to create it on that piece. I think the reason this seems hard many times is because we haven't practiced it. I think practicing it just a little bit will help you get towards that. That story is in the book Monetizing Data Management. So that's on Amazon plug there. See, Matt, I can do it too. I can do a little look. Anyway, go ahead, you want to jump in? Yeah, I actually had kind of a quick example of that too. So back when I was an analyst, I worked at Pepsi for a long time. And I think a little bit of it too goes into play of it's just a huge amount of effort to redo what's been done, even if it's inaccurate or incorrect. And I remember this scenario, there was a bunch of access databases that had been built and they'd been around for years. And it's almost like people knew some of the reporting was inaccurate, but didn't really want to dive into it because they would have to fix it. And I remember there was one time where we actually did look into some of the quality of the data that we were receiving. And then as it turns out, there's like whole product lines that would be missing from like Tropicana data from one of the big customers or vendors or anywhere that we were receiving data from and essentially paying for that. And so you're paying for all this data and you have it not be accurate or reflected in the downstream analytics was a massive impact on the way you, the business decisions you made. So there's all, I mean, there's a million different examples of that out there which can kind of showcase the importance of some of these aspects. And like you said, in your presentation, if you have a weak point anywhere, it's gonna be reflected at some point, it's gonna bring the whole thing down. So that's gonna be strong all around. Or else, you're not gonna have easy analytics to be reporting on. And I made it very practical in the example, in my household, if I build a poor barn on top of a, sorry, a good barn on top of a poor foundation, I will spend money on vet bills that doesn't go to the bank and that will make everybody unhappy. We're getting too far off topic Shannon. That was good. I love that discussion, it makes it great. And a very important topic too. Do you think the definition of roles and responsibilities required to build a solid data management is key to success and where can we find it in this type cycle? I think it's two different questions, but we as a community ought to be doing a lot of sharing and I think we are doing it, I see evidence of it. I can now evaluate an organization's data management capability by the types of jobs that it permits. Again, I mentioned earlier, sometimes it's not good to permit too much complexity around stewardship early on because people spend their time tripping over things and figuring out boundaries rather than concentrating on the main mission, which is advocating for better use of the data and supportive strategy. The idea that an organization has the other job data job categories is a very good indicator that they've been thinking about it because the HR people tend not to approve the job categories unless there is competition for those spots. So it kind of talks about the vibrancy of the community that's in there. Nevertheless, if just be very stark with the numbers, if I have a group of five data people and they cost $100,000 fully loaded, that's a half a million dollars a year, in order to keep that program going, you are gonna have to show some kind of value at some point and showing a million dollars value after two years can be very daunting to organizations. I think only because we haven't practiced it and if we can start to share this and this is the second half of the question, where is this stuff written down in a place where we can go look it up and get copies of it. That's a great function for a professional organization. I can think of a couple that would be ripe with just exactly that opportunity, but I'm not aware that anybody's done things in those areas that said within the federal government and the state governments, which when the federal and state governments were seeing a lot of this exchange of best data practices going. In fact, a lot of this material here has been done at the state level in a couple of states very successfully at this point. Matt, I'm not sure I left you anything on that one or not. That's all right. Well, here's another great question for you guys. Is state of governance the glue to data management or is it vice versa? I'm gonna take that one, Matt. That's a really good question. I mean, I'm not an IT professional. I'm just a marketing guy, but I think both can make a case for both. And I think ultimately, you want to have processes in place that make your analytics pipelines work. And one of the downsides that can come around of this, and I think people who are in business and I know I myself have been a part of this as well, is just the redundancy of work that ends up happening without that. And there's just a lot of time wasted, which equates to money and dollars loss from just redoing the same analytics projects over and over again or using the same data in different ways. And people just overlapping is, I think that's a huge drain on any business, especially large organizations where there's a lot of data being worked on and you've seen almost like satellite data warehouses is built on people's desktops. And so I think it's a process that just needs to be in place and is critical. And it kind of does work for everybody at the end of the day once you've boiled it down. I can hear the pain in your voice from having been there, Matt. Yeah. It's absolutely true. Think about it. We haven't educated people about data. How do they learn individually on their own as their journey through IT and whatever's going on in the business takes them that way? That's not the most efficient way to do it. It certainly doesn't promote standard practices. There's certainly nobody out there who's saying, what we really need to do, in fact, let me, my problem with it, I agree with everything that you said, my problem with it is that organizations don't realize that the skill that they need to get is not taking the data and delivering it in the dashboard or alert or whatever it is that they're working towards, but that the skill is in becoming more facile with the process and learning more about it and getting the end users to where they are able to fish on their own without having to be handed the fish to go to the biblical analogy on that. I think that so many organizations, again, just have no idea at the untapped productivity. Matt, if you wrote a story on that one or two, I'm pretty sure somebody would get mad at you at some point. We've written several. That's fine. Maybe there's good things to include in the stuff we're gonna send out. If I could send this to you. Yeah, we've got a few. We actually did a webinar not too long ago with Bank of America who talked about some of these processes and where they were just wasting a lot of resources on regulatory processes. And no matter what industry you're in, there's gonna be regulatory restrictions that you have to adhere to. And so there's just a lot of resources that go into that in general. So if you can find a way to cut those down, that's key. But I think a lot of organizations already understand that and that's part of there. A lot of initiatives that we see going forward is reducing some of those wasted time or redundancies or things like that. Not to say that they're not of value, they are, but I think a lot of organizations are trying to figure out a way to cut down on some of those processes. And Matt, let me ask you to go out on a limb here. Do you think the amount of regulation in the future is gonna be more or less? We've already had projects that have kicked off because of more. You're so right, you're so right. So it would seem for some organizations that it might become good to get good at responding to regulations. Sounds like a skill might be the other thing. I've not seen a lot of education go into the IT audit staff. And that's another place where they discover tools like this and go, ah, wow, this would make China my job really easy. Yeah, exactly. That's a good example. Who's gonna do all that workforce? Ah, so much of it. Shannon, more questions. D, lots of great questions coming in. We could probably spend another good hour on these. It's really good. So I've got a couple of different questions here from two different questioners, but they're similarly related. If we separate structured data management, relational database management, is the data management maturity status better? And then on the flip side, is there more specific guidance for those entering a no secret world? So far I have not seen anybody attempt to measure things in those areas. I think if it has been, let's bring it out and expose it. And if not, we need to get somebody somewhere at a university that needs tenure to study that as a field and publish some papers on it. But it's not an area that I know has received a lot of attention and it should because it's absolutely key. A regulatory compliance perspective in which data privacy and protection regulations are increasingly global? Any thoughts on any common missteps in impacting data management focus and strategy, particularly with regard to executive leadership? But you may have seen more on that than I have. Yeah, I'm trying to think of anything. I mean, a lot of times we're adherent to policies that have been put in place. I think, I mean, we were kind of like shocked, I think, at the adoption of cloud technology given, and I think a lot of people probably had the same feelings like how would you think that a bank or a financial institution or a healthcare company would be going into cloud technology? Just thinking about that in general, things almost, you know, like there's no way that would ever happen, seems too risky. But it works, it's great. I think that because of the governance policies and securities that are set up in place, and this goes back to some of the stuff we talked to you earlier, that they can leverage those technologies which drive their organizations forward because they know how to do it. They set those in place. They put them in practice and it works. And I think we've seen, you know, pretty successful implementations across a lot of organizations and agencies we work with because of the protocols that they have in place on something that, you know, three, four, five years ago we weren't even thought was possible. So I think that's just, it's a great sign of how these things can work. We are making progress. I'm talking to you from Virginia Commonwealth University which is home of the Fourth largest medical research facility in the country. And our Gmail and Google Doc suites are HIPAA compliant which gives us yet another level of protection around this. And just real quick, because somebody gets specific on this. I think I'm seeing some much more around GDPR because everybody understands that much more. CCPA is coming along very quickly and I think we're gonna see actually quite a bit of activity around HIPAA as this continues to roll out. So those would be three initiatives I'd keep my eye on from regulatory compliance perspective. Well, I think we have time to slip in. One more question here. So many great questions but I'm just gonna go down the line. You talk about educating mobile data spreaders but they are the easy targets and are usually lower in an organization. How do you best approach educating those that have been in an organization for 30 years to drive cultural change to data governance? The best way to make connection with people who haven't been familiar with data is to personalize it. We're seeing that not in just the COVID situation where people say, oh, well, that's not a healthcare facility I care about. But all of a sudden they realize that that's part of their zip code and it does get personal very quickly. So we're becoming better at learning how to educate all the way around. I'm personally in favor of advocating a set of skills that are prerequisites to a number of categories here. Shannon may we'll do a webinar on that sometime. So we get that up and running. I don't have time at this point to get into it but it's gonna become more and more a problem. Absolutely. Matt, anything you wanna slip in there to? Yeah, I think what you said, Peter, is exactly correct. And I think for us we try to make it as easy as possible to understand your data. So I think that's key too. And there's so many different personas and people who use data on a daily basis that they've got just a various different level of skill sets. You think of somebody who is mostly like an Excel user and then you put something in front of like a star scheme or something like that or JSON or an array. That's gonna be difficult to understand. I think that power then becomes being able to visualize and see your data and understand it and communicate the value out of it. And I think that's where a lot of that change can be driven. That was trying to switch screens here real quick and show another representation of something similar that you were talking about, but Sharon's gonna wrap this up, right? I am. Unfortunately, that is all the time that we have for. And thank you both so much for these great presentations and great conversation. Just a reminder to everybody, I will send a follow-up email by end of day Thursday to everyone with links to the slides, links to the recording. Matt, there's a request in here for a link to the trifecta blog. So I'll make sure that that's in there as well. And any additional information I'll scour through, make sure we get that all to you. And again, I hope to see you all next month and thanks to trifecta for sponsoring today. Hope you all stay safe out there and have a great day. Thanks everybody. Thank you. Thank you so much. Thank you, appreciate it.