 Hello and welcome. My name is Shannon Kemp and I'm the executive editor for Data Diversity. Thank you for joining the latest in the Monthly Webinar series, Lessons in Data Modeling with Donna Burbank. Today, Donna will discuss how data modeling fits into an overall enterprise architecture. 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. And we very much encourage you to chat with us and with each other throughout the webinar. To do so, click the chat icon on the top right-hand corner of that screen to activate that feature. And for questions, we will be collecting them by the Q&A section. Or if you'd like to tweet, we encourage you to share highlights or questions via Twitter using hashtag lessons, data modeling. As always, we will send a follow-up email within two business days containing links to the recording of the session and additional information requested throughout the webinar. Now, let me introduce to you our speaker for today, Donna Burbank. She is a recognized industry expert in information management with over 20 years of experience in data management, metadata management, and enterprise architecture. She is currently managing director of Global Data Strategy, an international data management consulting company. Her background is multifaceted across consulting, product development, product management, brand strategy, marketing, and business leadership. She has worked with dozens of Fortune 500 companies worldwide in the Americas, Europe, Asia, and Africa, and speaks regularly at industry conferences, including data-first city conferences. Donna, hello, and welcome. Hello. Thank you. So, I am going to jump in. So, today's topic, as Shannon mentioned, is on enterprise architecture and specifically how data modeling fits in with something kind of the broader view of the enterprise, which many of you know as enterprise architecture. Shannon introduced me already and won't go into great detail. There's a few things. If you are a Twitter fan, you can follow me at Donna Burbank, and there's a hashtag for today's session, hashtag Lessons DM. I am not the best multi-tasker in the world, so I'm not one of these presenters that can both speak and answer you at the same time, but I will try to get back to you after the fact, if you do. A couple more just highlights that relate particularly to today's session on my background that might be interesting to you. So, many of you know me from some of my previous roles, way back with Platinum Technologies where I kind of worked on their information management and metadata management technologies and did some consulting across the globe there. With ERStudio, I was the director of product management that actually built some of the EA features they have in that tool. So, it was on the BPMN, OMGB, Business Process Modeling Notation Committee, which was very, very helpful and insightful and to kind of actually being a stakeholder and actually developing that notation. So, felt honored to do that. And any of the features in, I guess, well, it was back then called EAStudio where we kind of put some business process modeling and some EA modeling. All the good features are for me, anything bad was somebody else's decision to put that out there. Spent some time with Irwin doing data modeling there as well. I've kind of jumped back and forth my career both in developing products when I feel like building stuff, which I think is really fun. And then I get edgy and actually want to work on site with real clients building stuff in terms of projects. So, kind of gone back and forth my career working with consulting and with products. Currently, I'm running a consulting company, as Shannon mentioned, Global Data Strategy where we sort of do this in practice. Previous to that, I work with a company called Enterprise Architects, which by its name was an enterprise architecture business consulting company. We've got a lot of great insights. I've been doing enterprise architecture for a long time, but some particular insights, that was their main focus and less on the data. So, was able to offer some of the data to them and get learned a bit more deeply from them on enterprise architecture. This is probably unconventional, but I did want to give a call out to a longtime friend and colleague, Ian Coward, who taught me a lot that I knew from enterprise architecture. And unfortunately, he passed away earlier this week in a tragic scuba diving accident. So, sort of a surprise to all of us, but I almost could not mention his name, could not mention his name because some of these slides and case studies I did alongside him, so it was sort of a timely, timely event. So, if anyone, I know a lot of the names on the call, so if anyone knew Ian, I want to let everyone know there. So, on a happier note, there is a series of data modeling series this year. You'll see it's really varied. So, this month is on enterprise architecture, and we try to break it up. What you won't see, and hopefully by the end of the year, you'll see some sort of detailed training on data modeling that we'll be putting together with dataversity. But we wanted to go broader because, and again, hopefully this isn't controversial, but data modeling in and of itself doesn't mean anything. It shouldn't be an academic exercise. The only reason data modeling is useful and interesting is when it's applied to an actual thing or to thread value in the organization. So, as part of an enterprise architecture or building a business intelligence, so you'll see, or for your career, you'll see some of these topics are, you know, how can that help you as a data architect? Usually in the titles we see a lot of data architects on the call, which is always great, and how can some of these things help you? So we try to kind of mix it up. Shannon and I each year kind of look through some of the trends in the industry and data modeling and metadata are always a big topic, so we try to apply it to that. So, hopefully you can join us throughout the year. They're all out there live. If you want to register ahead, it'd be great to see you. So, today we're going to talk a bit about, and if you heard me present before, I always put data in the context of business transformation where I can because that is honestly why I'm still in the business. My first degree was in economics, and I'm very interested in business, and when I work with companies that's what I always love doing, of not just solving data, which can be fun, you know, I can nerd out on data like anybody else, but it's only interesting when you're doing something with it, right? And I think this is seeing a lot of the growth of data in the industry is that company, you know, non-data people are finally getting it, but this can really transform their business. So how can enterprise architecture and data modeling specifically within that help with that? And then I'll go through a few tools and components that are part of enterprise architecture. And some case studies where we've done this in actual practice that should maybe relate to what you're doing. So as I mentioned, data is driving business transformation, and I know that sounds like a marketing headline, you'll see most of the vendors kind of who are anyway related to this calling this out, but it's true. I mean, when you think of, you know, the calling the digital transformation or data-driven transformation or just business transformation, so much of that is really driven by the availability, volume, and interconnectedness of data, and it really is transforming the way organizations do business, just think of something even like an Amazon, you know, their recommendation engine which drives a lot of their business is all based on data, right? They have data on the purchasing patterns of millions of customers around the world and can use that to help you and help them, you know, buy more related things. And so many companies that I'm working with, not all, unfortunately, sometimes they've come up with this funny idea before I even came on site, but, you know, they'll have the poster on the wall, we are a data company, and this could be anything from a telco company to an insurance company to a product company. Folks are getting that the real value is the data that's generated from their doing business and how they can take that data and really leverage it to business advantage. And when you're thinking of how can I leverage that to business advantage, there's kind of subtleties to that, so I like to break it out into two different ways. So one is, I would call business optimization there on the left. How do I become a data-driven company? And simply that's, how do we do what we do better? Right, so if I had better information on my customers, could I put together better marketing campaigns? I could have that elusive as everyone wants to have, you know, 360 view of customer. Could I use better products? You know, a lot of a product that is a, you know, from a telco company or an online store or an online game or an online, you know, anything where you can track the usage of that product, you can actually see what works and what doesn't. So you can actually link that into your product development. You know, customer support. A lot of the companies I'm working with are doing a very good job of this. Can I actually see what my customers are saying through tech analytics and support logs? You know, some really interesting things there. You know, and the standard, if we manage our data better or our company better through better data, you can lower the cost. So we've been doing this for a while. That's fine. I think we can do it a whole lot better now with some of the new technologies that are available, which is exciting and driving a lot of the innovation in the industry. I think the one on the right is a bit new. Again, you could argue we've always been doing this, but I think the tools that are available now, the volume of data that's being generated is really driving this idea of becoming a data company. So either literally monetizing the information you have, selling it to other groups, for example, I worked with a telco, telcom, whatever, everyone has a different word for that, and they were actually selling their anonymized, of course, data to some city planners to see if you think everyone going everywhere nowadays has their cell phone, they could use that to track travel patterns throughout the city, and how can I optimize rail networks or road building based on the data of something completely different, but used for a different purpose. They had worked with a retail company when this particular retail company didn't have a loyalty program, one of the benefits to a company when you're a loyalty program, they can see what you're buying and kind of customize things for you, but they didn't have that, so when people are in the store they had cell phones, they could literally see footfall analytics of where people were going, what they were buying, what ads in store were working best, that sort of thing. Social media, kind of seeing purchasing patterns, ad placement is sort of a big one. Energy, we worked with an energy company earlier this year where again, energy is one of those commodities that they're actually trying to sell less of, they're actually actively helping their customers be more energy efficient, which when you think about it is sort of weird. You're actually trying to help your customer buy less of your stuff, so they were saying how can we use this to actually have a brand new product and with the idea of smart readerings they could actually have services to help manage, shut off your heat from your cell phone while you're at work, some really interesting next generation things, they can only do that because they have the data, so it's kind of exciting and they were one of the folks that did have on their wall, we are now a data company and we're really seeing the value of the data itself. So some exciting things out there in the industry and I think what's exciting for us as data professionals is that, well I guess we could say finally, we have a seat at the table and we are getting the voice. I think I've worked with a lot of business executives where they would love to have someone really explain, I know we have this data, what could we do with it? So in the past we'd get the business requirements from the business and we implement a product to meet those and that still happens and that's great, but I think we can now often come, think of the classic data scientist, I saw a pattern in your data, you realized you could do this, that and this and we get together and get this new insight and I think that's a great opportunity for us and that does link enterprise architecture which really links to your business opportunity and you can really support things like organizational change. So I think it's an exciting time to be in data which is why I'm still here and not out skiing because I think it's sort of fun and I hope you do too and hope by the end you'll see some specific ways you might be able to help your organization which kind of leads to the point at hand of how data modeling fits with the idea of this, what is enterprise architecture? As a data person we always have to start with that definition so I'll do that. So this is a lot, again, they say the cobbler's children always have no shoes so I think us as an industry given that part of our main job is creating definitions are pretty horrible at doing that ourselves. So usually there's six definitions for something. So similarly there's a lot of different definitions of enterprise architecture out there but this one I liked it was from Gartner a few years back because I think it really hits on what I'm trying to show today if you sort of broke it up into different paragraphs just for readability. But enterprise architecture is a discipline for responding to the part I bolded, disruptive forces by identifying and analyzing the execution of change toward desired business vision and outcomes. So you think about that's really what I was just talking about that innovation that really disruption of the entire business processes by analyzing it and looking at the big picture but at the bottom you can read through the rest at your leisure they made sure to point out that this is also driving the evolution of the future state architecture so that's really the foundation to do that and they shouldn't be mutually exclusive activities and I'll touch this later in the presentation. I think enterprise architecture probably just like data modeling probably just like a lot of things that smart people do folks seem to think oh it's too academic it's going to take too much time it's heading the clouds it doesn't mean anything and one of the reasons I like this particular quote is it really hit on both you know Ocontrair this is actually helping you analytically look at your business and help innovate with direction but at the same time you need to have the architecture that supports that so even I'll go into more case studies at the end but even the few I mentioned you can't do all these great next generation things with your data unless you have the architecture for that data to build it with so I think a lot of us on the call understand that but sometimes making the connection to folks that your data and architecture and that whole concept is new to them I should have liked this quote because it kind of balances both of them together so if you're a data architect on the call sort of a definition that might work for you is you've never heard of enterprise architecture if you think of just how you have to model the data in an organization I think if you're modeling the organization itself right so to step back and a lot of us who are modelers tend to do that I'm always drawing something with a back of a piece of paper and that you know we tend to be visual people how do I plan it out how do I draw it how do I make it simple I think of something else we like to do with architects so I think what are the motivations what are the goals what are the business capabilities and processes that we can put together that run the enterprise as well as naturally on the business side as well as I mentioned before you need an architecture that supports that so if we have our business goals and the capabilities and processes that are either built or we want to build how do you have the application the data the networks you know the stuff on the back room that needs to support the organization to make that happen which is one of the beautiful things about an enterprise architecture because it looks at all of those so here's again we are sort of focusing on the data aspect more on this but there's a lot of facets of enterprise architecture here's one way to look at it and some of the artifacts below that within an enterprise architecture so if you think of the business view you know what are we trying to do with the business what are we what are the business processes for the process view the data view and then mapping the data to the process so we'll show some examples of these but with the business you know we actually have an artifact in our practice called a motivation model actually writing down your motivations I was one consultant and he actually did this is kind of a personal development you know we're doing it for you know a project but you could do it for anything why are we doing this anyway we all just step back sometimes that's the most important question of all we can build the best architecture but if no one's ever going to look at it right and what business areas are we supporting one of the drivers and then how does that map to data and then specifically you know what are the business processes that data is supporting is that the front end online sales process is that the back end supply chain all of the above and often I'll talk later one of the case studies I'm sure many of you on the call have done things like data governance and data quality you know especially with data quality that's so much ties to process and people because you think you know what's the classic analogy if your data is the pond hesitate to use word lake but you know and you clean the pond if you're not cleaning the rivers feeding the pond and the data is still coming in badly you need to fix that so sometimes it's something as simple as a data entry problem that the person wasn't aware and so really looking at that process and see how data is affected and created throughout that process is pretty key to really using data or fixing data so when we talk about enterprise architecture itself there's a number of different frameworks again like definitions we always like to have several to choose from in our industry and I show you a couple of this more than this even but a couple of the popular ones that I've used or have been out there I'm a fan of the Zachman framework in its simplicity and I think that's why it gains so much popularity I'll call it to John Zachman if people don't know he was a lovely gentleman and he's generally at the diversity and damoconferences I think he'll be at EDW this year and generally he's a very nice gentleman if you can go up and talk to him but basically it's very simple if you've heard me present a metadata I use similar categories for metadata it's basically the who, what, where, why and when so if you think of enterprise architecture as the who, what, where, why and when of the organization so what is going to be your data and this one's a bit hard to read he's upgraded everything his older one was more simple but a little bit easier to read for us old folks but the bottom if you think at the bottom it's more of the actual architecture the systems etc that are driving the organization as you move up almost think of this as kind of the logical physical logical and conceptual model so he has kind of the enterprise vision perspective the business perspective the architect perspective then the actual engineer building the systems so in my mind that's very similar to the levels of data modeling but he does that with everything so it's the who you think of the organization structure the where I think of your networks the why in terms of motivation it's a very nice kind of holistic look just to see am I catching everything in the organization and again I'll talk about this later in the presentation as well this can seem overwhelming and we don't want it to generally an organization starts either broad or deep you can do a high level across all of the high who what where why when a very high level view of everything and then maybe go deep into one maybe the data is where I want to drill down into some organizations go and complete complete depth and breadth across all of these but other folks kind of pick and that priorities to really focus on one area and we'll talk a little bit about how you might want to focus as well another framework that is out there that I've used as well we think of a toga they have an ea framework from the open group and there's a lot of documentation out there partly because of that documentation it can seem and the surface is being a little more complicated but complicated as always in the eye of the beholder right it's you know whatever you're used to and they all have this strength and weaknesses and you'll see here is some of the same idea we're looking at you know the business the data the applications the technology as well as the capabilities and vision across the organization so similar idea you're kind of looking broad in terms of what do I have what do we want to do with it and then you can go deep into you know actually creating your network diet we're not going to get into those things today but often when I was doing a detailed enterprise architecture say I were for example to go deep into the data there's probably an application or a network or you know each group could go very deep into some of those and we all work together at that high level was kind of the place where we met in the middle and my advice strong advice here is really to find that balance and the thing that helps you find that balance is the business value one of the as I mentioned already enterprise architecture can have a bad rap is being this overly burdensome academic thing that's just a waste of our time and we have stuff to actually do right I'm seeing much of that again a lot of these things we've been doing in the industry for a long time are seeing a bit of resurgence things like governance enterprise architecture where I think for a while it was I just go build things kind of this Wild West in the lower right I think there's a balance I think it's very true I've been on enterprise architecture programs that were too academic and it was just let's build everything at the end level of detail and think philosophically for hours about every bit you know fill out every box of the Zaken framework nothing's really going to get done if you go that deeply I'm a big fan of this picture the thinker with a laptop but at the same time and I wish more people said this because I think some of us on the call can be frustrated by this as well if you're too Wild West and just say oh we're going to skip that modeling phase altogether nothing gets done there either because you're in chaos so my voice just got all agitated right you can see that bothers me when you're on a project oh we don't have time to plan let's just get going you know I'm a big hiker fan you know like free time when I haven't and just think you're lost I'm not going to look at the map I'm just going to start running well you can get really tired really fast maybe a lot better to quickly look at the map you're going to study the map and get every atlas known to mankind you're going to quickly look at the map find where you need to go and go with it and I guess that's why I'm trying to say when done correctly enterprise architecture actually helps improve the efficiency and better align with the business priority so I think every project should take something from the enterprise architecture even if it's quickly and I say this to a lot of my clients who might be you know we'll put together a proposal and they'll see wow we have to do this motivation model and this architecture model and all these things that seems like a lot of time and I say if we start a couple hours in an afternoon and that's enough that's fine you might not even need to go past that so give it a chance and sometimes that one to two hour session bashing out a motivation model really changed the direction of the project and we were glad we just had time to take a breath and think before the project so that was sort of a long winded way of explaining the slide but I think it's really key in what we're going to be covering so back in the actual things are the tools that we can use to build this because I think a lot of the folks in this call are implementers and you want to actually get stuff done I'll kind of go through some I'll start with the data modeling piece wouldn't necessarily normally start here in an enterprise architecture project but given that a lot of the folks in the call come from here that's why I'm starting here we'll go back to the bigger picture in a bit so when we think of data modeling I always show the almost always show the slide in my presentations and every time I don't show it I have to bring it up or I wish I had brought it up so here it is this idea of the different levels of data models and I think one of the reasons is good to show is because people can we can start talking in circles so many people think of a data model as just this physical side I'm going to reverse engineer and I'm going to get my structures from Oracle that's great and there's definitely a place for that enterprise architecture I think often when we're thinking of the business side we're more up here where the yellow is so if you're not familiar with that that's your conceptual data model what are the high level business concepts if I'm doing a business transformation what is the data I need to worry about is it my products my customers my vendors my suppliers you know really kind of that high level business domain level and you know the data modelers in the call can get a chuckle out of this one they'll know you know what do I mean by customer which you know again it can seem so simple but if you don't have those basic definitions at the beginning are we talking about human customers or be the be customers the actual enterprises and then when you start fleshing at the logical model you might see the difference okay let's get customer gender well we're talking about organizations you know we wouldn't have that as more your you know your company ID number or your tax code number or something so it's good to start again it could be a very simple whiteboarding session could be a very detailed model but do it in either case in a logical such to flesh out some of those business rules if you think of that's acting framework you're kind of going down the framework into the depth here that kind of gets the clarification of your business rules starts to think about data structures we could argue back and forth and whether you need to or not I'm a big fan of that because it does start to get some of the business rules can a customer have more than one account if they have to have an account to be considered a customer all that kind of stuff which is key to the business and then the physical again some folks think of enterprise architecture as just up here in the top because often enterprise architecture is just that broad very high level models I'm also a big fan of kind of looking at the reality kind of the top down bottom up sort of leaves me this too so sometimes I think always you should do that top down what are we worrying about it with our business what are the main things we have as a business that we that our data assets our products our customers our linkages between things and then also do a bottom up sometimes you'll find stuff you didn't know oh I didn't know we had this much information about the purchasing patterns or I didn't realize we had product information or just to find out where it is and so also to get the complexity of how hard this is to manage because the devil's in the details we have this great theoretical enterprise architecture but at some point we have to build something and in my practice we always do some sort of analysis of business benefit well I'm sorry everybody does this to some level you know business benefit versus complexity to implement so it could be you know the best thing we could get is linking our customers purchasing patterns with their other friends they have online and kind of marketing to them well we don't have that information right or we have that customer information but it's in 26 different databases but look at our customer or product data is all in one nice clean database maybe we start here you know again you have to kind of do that iterative back and forth between what you have and you know what the conceptual business value is and what the physical reality is so I'm a big fan of kind of looking at both and really kind of going back and forth and seeing you might need to change your physical structures based on some of the ideas you had up top so just quickly if you haven't seen a conceptual data model if anyone was at Enterprise Data Governance online we showed one of these and there's a lot of discussion on this couldn't you just do this with sticky notes yes we have this slide but this is really your brainstorming slide just another in full disclosure this really is a logical data model with full attributes and data types and all of that sort of thing and we obviously just hid that and this particular tool I'm using you can actually just show the definitions as part of the model and I've used this a lot and kind of you can whiteboard and you can either take the results of your whiteboard and put it here or sometimes you know depending on your skill and savviness and how easy the tool is we can use the tool itself in your whiteboard because the beauty of that is the metadata is already in there and then you want to go and link that to your physical structure you can do that easily and someone's asking which tool and I never say tool I just think that's I try to be vendor neutral but it's one of the top three that we tend to all look at the benefit of this is you see the definitions right there so it could say you know a customer is a person or organization who's run as a movie within the past year or you could say well we don't sell the people at all we just sell the organization you know sometimes just showing a simple definition just like this can really pull out some interesting things you can also highlight if folks are having you know kind of those definitions the customer is a customer we've all seen those they're not very helpful and sometimes just brainstorming like this you can show them then going to the logical model and again that's where you start adding attributes so what are the key things we worry about customer first name last name sometimes going here too can start to flesh things out I gave that example you know here's our customer what's their first name well it's an organization right we don't have is it a person that works for an organization you start to get all these different you know business rules you know do you have different roles a student or a tutor is a different type of role so there's a lot of information there that you start to flesh out cardinality that kind of thing so yes a lot of times you're thinking of a physical structure but really the focus here should be on a business the business rules around it kind of customer more than want to count that kind of thing and now I often just started the broad brush again depending on any can go as detailed as you wish this idea of whiteboarding I'm a big fan you know especially if we're working with business users and plus if anyone has heard me present or work with me or see me I can't sit still so I really like the physical I'm always one of the volunteer I'll put the stickiness on the board the actual physicality of having conversation and physically sticking something on the board I think it's nice I think some business people have been traumatized we just want to have a data modeling requirement session and you walk I have been in these rooms where you have the enterprise logical model that takes up each wall and spans a thousand entities if you don't mind we're just going to take your afternoon go through all of these definitions you know funny they run away and never came back right so focus what you just need from them you know we're just trying to get some these the main things you worry about in your company you know regions and policies and agents and account is there a difference between an agent and a seller do you have external agents internal sometimes just an hour half hour with the business unit with a business person with some of these key definitions can go a long way you know maybe you need to bring them in later for some of the different definitions but it's not often an easy way to start just for the for clarity I am not anyone to quote on says you should do your entire enterprise architecture with a sticky note no it's just it's a brainstorming and that should be then translated to something more permanent with metadata behind it like an EA tool or a data modeling tool or metadata repository or even just something like a confluence or a slide share or something to make sure that you have the metadata and it rolls behind this. Okay so that's a bit on data and I started with data because we're data folks I generally when we think of an enterprise architecture would start more like something like this we call it the motivation model I've seen a lot of different displays of a motivation model and I'll show you on the next slide that we use but we don't have to be married to that the idea is this is sort of that formal definition of why we're doing this again doesn't have to take weeks it could be an afternoon it could be 15 minutes of folks are saying why are we doing this because we're trying to increase sales why are we trying to increase sales because we're getting a lot of competition in the market and everyone else is online and we're not you know something very simple because I've been in projects where at the end of the day everyone had a different goal and it wasn't seen as success not because the technology wasn't great because people it didn't tie to their day job it's sort of the so what so one place to start is generally the corporate mission and I guess a lot of us who might be in the weeds of data do you even know your company's corporate vision so it might be printed out on the walls of your organization and HR and executive management put it up there have you taken the time to read it do you understand it and you know sometimes these can seem silly in their simplicity but I'm a big fan of if it's simple and easy to understand someone took a long time to make it that simple right so if you look at this we're saying to be the most comprehensive customer driven online shopping experience in the market well there's a lot of things there you're trying to be comprehensive but you're all trying to be online maybe we weren't online before so you know a lot of things there were not maybe focusing on brick and mortar you know so a lot of things in that small mission that we can say the vision is kind of where we want to head so we're here at the mission today what are we trying to do differently in this case again this is the hypothetical organization but we're trying to transform the way customers purchase goods through social media driven connections which really is more getting on to the idea of social media and and communities and that kind of thing which might be a completely different way of working for this particular company so externally why are we doing this you know maybe because everyone else is doing it everyone's buying online we're not going to be the one to put out a lemonade stand in the front yard because no one's going to come and then what are we trying to do internally well maybe part of this is that we want to get that integrated view of customer but there's disparate systems across the organization and we might need something like an mdm or warehouse or something to aggregate that and understand it and then for your project you're doing a different project in your enterprise architecture project what are we trying to do well maybe we're trying to specifically improve customer satisfaction or you know whatever and there should be several goals not just one but actual have a goal so when you meet it people can say oh that's great I want to kind of show some success and then some specific objectives so how are we going to meet those goals well here we can link purchase history with social media activity that's a specific thing we're going to do with our data or their enterprise architecture that we can kind of do here's an example of one and I generally like to put it in some sort of infographic type thing I think you'll just sort of consume that better so this is really kind of a fictitional art supplies right well you'll see basically what you just saw on the previous slide with the mission with the vision is internal external drivers and then this was for more of a data governance type thing what are the main kind of marketing headlines right we can always sort of roll our eyes at marketing but why does marketing work because it drives to what we care about in a simple way so what are we doing we're trying to drive accountability quality and culture or you know put some graphics in there really really think about how you might do that but the most important thing is we're all driving in the same direction this is the why of why we're going to do any enterprise architecture program so again whether this takes several weeks of iteration talking to different stakeholders I'm not going to go deep into that here but you know a big part of this is to make sure you're not just talking to yourself so you know look at the organization all these things we're talking to do we talk to marketing right did we talk to the brand strategy group do we talk to the customer service group all these different people that might be involved in this make sure they are involved in our stakeholders as well another big thing I like to do is really look for these levers or levers or how you say it from wherever you're from and really that's you know finding that quick win so and this sort of gets back to what I was saying before so you know first of all identify the areas that are going to have the highest business value what is the biggest thing you know and some of these things are the best thought of by walking around the building taking a walk looking around over just thinking what is the biggest thing we're talking about this is a huge big new campaign for big product can we align ourselves with that that's great you know again it could be a smaller technical project but bigger bigger business value you're going to get a lot more kudos than trying to do something big and fancy and techie that nobody cares about or second second bullet you're rearranging the deck chairs in the Titanic right so you know we're trying to polish the counters of the restaurant and the company is going to online ordering we're not even going to be using those code you know counters so why are we doing this anyway as with anything that has value it makes sense to model it so let's look at these eight ways that you have value of the business and I kind of like this business value I really see that as where data can be the fulcrum right so there's a lot of effort that needs to be done there's a lot of load but what you know what's the value where data can really provide something transformative to the business so you know we often go and just prioritize so what would be the biggest new thing that we're going to worry about in this case it is the launch of a new product we have this big new marketing campaign so what are the big drivers we just talked about and then you can actually in the data world start filtering what are those data elements that we care about and focus on those because as you know there's thousands of that's where this conceptual model come in there's thousands of different areas of data across the business you can't manage everything effectively so pick the stuff people care about that relates to this with a marketing campaign you probably need customers and products and partners to sell to and vendors to sell it regions sell within make sure you get those right and show the value for that another thing that's part of an enterprise architecture that I'm a big fan of is a business capability model so this outlines the capability of the organization this is a very simplified version of one but you kind of get the idea just note the note there this is not an organizational chart and sometimes that's a hard switch for people to get so we kind of think okay so what are the big functions of my business you almost start thinking of the org chart live HR and I have development I've engineering and I've sales it's helpful to step back from that and think what are the core capabilities that we do as a company especially when you're thinking of business transformation this can often be where some of these aha moments you know can be and this might be while one of our main things is we have a great source of data how can we manage that or it could be that you know our product isn't really that differentiated but we've got a great distribution system or whatever it kind of just stepping back and saying what are we really good at or what do we have a lot of functions that or it could be this is the business vision we want and we want to sell everything online you know we don't even have a web team so how can we sell online so we need to add a capability so again this is kind of your conceptual to get conceptual data model this is conceptual business capability model to really get at that so again what are the core areas of the business this is our fictitious art for art supplies again we have research and development we have branding and go to market and we have sales and distribution probably pretty simplistic there's probably a lot more and then we have things like back end kind of your core business functions and then we have kind of the shared services like human resources you know legal services and that kind of thing and then with each of these okay within product development we have is kind of a functional decomposition type of thing here we have R and D we have product manufacturing packaging and that kind of thing this is helpful in and of itself what I find helpful is then you know based on these highlight priority business data elements we or business domains we created can we overlay that into some of these business capabilities so if we have customer data what are the capabilities using that or leveraging that or needing that you know we're going to go to that level of detail but or owning that etc you can kind of create a heat map so wow you know looks like and you know sales procurement and quoting using a lot of different data types whereas you know recruitment really isn't using any data at all should they should they be looking at you know employee data yes but that's not one of the things we're worried about so kind of just a helpful way to do that heat map of what do we do with the business let's look broad term of how we need to change that and then specifically how does data support these key capabilities going kind of a next that further is the idea of a process model huge fan of these as well so I have seen companies again do this on a whiteboard you can do it was sticking out if you wish with lines you can do this in something like a Vizio or I've had one customer that it was an engineering type customer and they had each business engineering process to the nth level of detail which was very useful to their business it was actually a lot automated that particular customer actually and they linked it carefully to their data they actually had a quality issue that made it into the news and was obviously not a situation they wanted but because they had their processes and the data map to that they were able within 24 hours to find the problem resolve it and actually have a press release to show that they were fixing it and this particular case one of the reasons I had come in they had done this again before me the IT team got a whole lot more budget they were actually seen as heroes of this scenario so you know for the folks that said oh this seems so theoretical so architectural and do we really need it well when crisis came they were really glad they had it but I've also worked with customers were really it's just a you know quick and dirty you know who's using this data and how they can really help you understand the business usage so here if you're not familiar this is kind of a bpmn style I usually don't get too dogmatic about getting every line in box correctly it's more how can we use this but a lot of notations have this idea of a swim lane so who your actors are the people so I product development I supply chain accounting and marketing this might be for a launch of a product and this is kind of my own version of this but so I have you know your start event we're going to create launch the product I have the development cycle and then within each type of data what do we look at so we have a product we've kind of cut the code name of that product because we're just kind of we don't want to make the real name yet we're just using something we kind of have the product components the names that component that's a kind of data that's touched in that process then they might pass it out to supply chain accounting is this even feasible to build maybe it's too expensive maybe those parts you had in that product development are made of gold and that's way too expensive they might do the pricing and costing that kind of thing then they might if it's feasible they might pass it to marketing and they might create the real name of the product so that's the case we're the same data kind of being updated by two different groups so kind of seeing the flow of that they might create a price that overwrites the you know this could be the kind of theoretical price of this is how much based on the cost and they might look at the market and conditions and kind of change the thing so what's nice about that it kind of shows the well a the importance of data who's using it who's updating it creating it deleting it that sort of thing when you think of kind of fixing the data or leveraging the data and then another sort of sister brother tool of this which wins the award for the absolute worst name ever and I should think of something clever and branded and start using it but I haven't can't so anyone we got someone to pry for creating a better name for this lovingly known as the crud matrix which just sounds horrible sounds like something gross you have stuck to your shoe right crud but and often when I introduce this to somebody who hasn't seen it they kind of look at me funny like no no skip the name it's actually really helpful it's basically a quick and dirty way again of showing how data is created read updated and deleted thus the crud so it could be again these are all the types of things that are data this could be at whatever level could be at the domain level it could be at the entity type level the attribute maybe just saying the product name but this might be helpful so initially product development creates that name and then but it's really marketing that updates the name they're also the one that deletes it you might say you know we don't have branding rights for this no don't use this name anymore so really when you're thinking of data lifecycle this is a huge part of it to really understand the usage of data create update read and delete so again you can get really detailed in this it's also again if you have you start this in an afternoon just to kind of see especially when you're thinking of things like governance and quality or enterprise architecture how you're kind of supporting a lot of these is a helpful tool so in general when you think of EA you can think of it as an architectural discipline you can think of it as academic discipline I like to think of it as a really a holistic view to support business transformation so if your real goal here is business innovation and growth how do we do that how do we grow the business well we probably have to affect our processes do we try to do better sales processes do we need to hire new people do we need to organize our people in a better way are there capabilities we need to optimize are there new capabilities we need and then kind of on you know the technical side what data do we have to support this what applications are using this and I guess probably the most important one why are we doing this right what's the motivation we are trying to be the best you know seller of art supplies on the planet which is very different than you know we want to just you know stay stable and I want to retire in a few years and hand it off to my family you know whatever this type of thing is very important so it kind of provides that holistic view in a very from top to bottom top to bottom way to do that before we open it up into questions I did want to go through a few case studies because to me especially being a practitioner that's kind of where the rubber meets the road as they say kind of maybe push this in context so names obviously get to protect the innocent although these are all positive examples but this was actually we started we were going to do a master data management project but it really turned in to more of an enterprise architecture project where a lot of the focus we were doing was process modeling and linking data to process so this was international restaurant chain and they were doing a you know basically they were doing a company strategy this they weren't doing a data strategy they were you know all the executives and the CEO across we're all saying what do we what's our business assets what how we really grow what they realized that what's the core capability of their business and it was menus so they in particular had very innovative menus you know I'm going to sell a particular food item and they would change the menu and it'd be very when you think of it that's what a restaurant does right they're selling food on a particular menu and they realized once they did go a little deeper they didn't really have that single source of all my menus in fact they one of the comments they had was that I think the printer has better master data we do because they printed all the menus so they wanted and then when you a lot of the data for that was scattered across different systems from supply chain to how the kitchen prepared the menu to how marketing sold the menu to how restaurant operations supported the menu with you know how you actually built food and that kind of thing so they started a master data management program very wisely to say can we get that single view of our assets to support the business and the governance around that and all that but as soon as they started they realized you know we need to start with process and we actually built a very detailed process models from literally and this was way fun for me I got to go into the kitchens and talk to the chefs this isn't the actual chef but kind of look like that and talk to them and they got it you know that's what I think you know us data folks sometimes forget that people get data once it's put into context so we actually didn't know he was doing it but we started to whiteboard kind of what he did and it was a process model with some boxes for data so it really looked like you know they got kind of a workflow model and he kind of got that he had these different data things that were touched by it and we kind of did some high level data modeling and process modeling with chefs where he went to culinary school right so they non data people can get data and then we talked to each of these groups so we talked to the marketing team so how do they take and this was literally almost the flow how do they take that menu from the chef and make it be something great that people want to eat and buy to supply chain how do they buy the eggs and the broccoli and the milk and we hope that's not all one meal together and get it from the trucks and actually have it be you know a nice beautiful thing in the point of sale system where you click it and buy it and that was all data driven so we had very detailed process models with how data was updated across the way and that's what we had to do before you can even think about an MDM hub that really was the not the hard part because they didn't have a lot of data that was really just menu names descriptions prices that kind of thing it was more in the process itself so that was a great example where it was really the enterprise architecture around it the people the process the organization and one of the things we started with was a very high level capability model what are the groups even touching data before we start the process model so that was sort of in the form of one for all of us because we didn't think it was going to start there kind of ended up there another one we worked on was a merger of two big financial services and in this case the CEO actually said one of the reasons we merged because we knew we had a lot of combined data and if we can get that together right we've got more data on certain customer demographic than anyone else and we can really be strong in the international market so one of the things we did how do you get that common data foundation we started by building business capability models for each one of who this was beyond data at all again to start it was just these are two very different companies how do they organize their business capabilities are they overlap can we get the capabilities aligned first and then what data can use to be support this so do we have you know there's one column human resources and the other ones people innovation or something you know or how do you align the company and then the second step was how do you get that common data foundation but we didn't start with the data in that case because a lot of it was understanding the business first another one I'm quite pleased with because again for those of you who say business people don't get get data I will kick you if this weren't a webinar right I think the more and more examples we have people do if it's explained the right way so this was an international pharmaceutical company and they if you know anything about pharma huge part of their business is R&D if you want research and development if you want to stay ahead of the curve and have that new net next big drug that saves people's lives before anybody else that's all research and development clinical development how you get an idea to the market fascinating to learn all this stuff and then how you take it to market so our quote customer was the type of lady on the right they really phd research scientists and what they found helpful is we created these kind of they called the blueprints of the business so business process how you actually take this was so fun to learn how you actually take a drug from the concept sub phase to the product development stage to the testing phase to all of that and then they realized this was an issue how do you what data is either used updated deleted needed and they found this what was fascinating to me is that they had both process models and conceptual data models and you'd go into this brilliant research scientist office and she'd have a conceptual data model with either sticky notes on it or lines crossing it out and changing it around so they got it right it was a description of their business and they were pleased you have the so what they were able to find efficiencies in the process either hey we're all doing different things in a different way or you know if I could get that data from the other group and I could get that earlier in the process then that would speed up our cycle that was a big part of it that kind of information sharing I didn't know you guys had that off I could get that early that would be great or we're using the same definitions in a different way that kind of thing so this was you know using data and process to really map how the business runs we didn't again we call them blueprints we didn't say hey we're doing this fancy data architecture and you have to do a bpmn diagram that would have made their eyes glaze over right but we talked in their terms we built a flow of their data and their business and then they linked to the data and we're able to get process efficiencies there so in summary hopefully you've kind of seen the connection that really today's digital transformation is driven by data which is awesome and it's great for us as part of that you need to just put data in a bigger context so in some cases you can go to the CEO and say hey we need a new data analytics project just do this and they'll get it but really to do it right I think you have to take that broader look step back and look at the how it's going to affect the people the process the capabilities the motivations etc to do that and you know just as you model data you need to model the organization itself and I hopefully I showed you a few tools and artifacts to support that and then just with anything again you this can get overwhelming so focus on the areas with the highest business value generate some quick wins and then go back and finish the rest just quick about us if you're not familiar global data strategy we do management consulting across the globe if you need help let us know and before I go to questions here's my contact information always happy to answer questions after the fact and I wanted to just remind you of the modeling series next month is going to be a business intelligence and hope you can join us then so I will now hopefully we have some time to open it up for Q&A Shannon indeed Donna thank you so much for this great presentation as always just to answer some of the most commonly quest asked questions that we receive just reminder I will be sending a follow-up email for this presentation by interday Monday with links to the slides the recording of the session and anything else requested throughout as well as Donna's contact information so the first question coming in here Donna is should CDEs be defined on the conceptual layer or the logical layer what are the pros and cons so CDE I'm assuming you mean the data part so I would say both I'm a big fan of starting with the conceptual and that it kind of sets the concept for like that word it really gets the scope of someone we're thinking of kind of the business motivations the business drivers you know where are we going to focus our effort kind of also kind of a inventory do we cover everything or what are the big areas of the business and do we all agree what they're talking about you know one company it was something like well we have customers we have clients you know support calls and clients are they really different things is a different area of the business maybe brokerage calls and clients you know it could be a different is a different thing or is it a different you know name versus the same thing so I think that's always a great place to start I guess the risk of ending there is that you know it doesn't go deeply enough that's where I think the logical level is helpful and this is true across any of those artifacts I showed whether it's a capability model or a process model always a big fan of just starting high level just see where you're going I always give the analogy of a house right so I'm I want a new summer home you know just basics for the architect I want a big fancy mansion so I can bring you know all thousand of my family or I want a little shactical fishing in big difference right then you kind of draw okay this is kind of what it looks like and then here's the rooms and then the electrical dot you keep getting different levels of detail so I like it that way you can't always go top down I think the benefit of the logical is you really start to get some of the crux and sometimes it might be different people in the room for those I guess the downside of that is you can get stuck in the weeds and sometimes even when I've tried to keep it simple you know it just starts cascading in terms of the number of entities and the relationships but the risk of not doing that is you can miss some key stuff you know something as simple as okay is this I don't know different kind of policy do we have different policies are they treated differently the attributes on a public policy versus the private policy versus the academic policy differing all these kind of different things can have subtleties so I think you know both can be valuable just make sure you're you know the purpose and don't go too much into the weeds unless it's the time to you know at some point someone has to go into the weeds if you're building the application because there's a risk of not doing that as well just make sure you get the right audience and they're doing it at the right time fabulous are you now are you finding it harder to enforce the importance of a model within move to unstructured data are you seeing modelers being left out of the conversation or do you have and do you have any advice for those that are experiencing that I am actually ironically at the enterprise model I think this is a great way to start showing it so you know I'm working with a client now and we kind of very are at this stage of what are we trying to do with our data we're trying to get for example you know better customer support and then what type of what are we trying to do what are the process that need to do it it's online support it's online sales etc etc etc and then what types of data so we're all the very conceptual layer what are our motivations what are we trying to do what processes and capabilities we're trying to support and then at a very high level what kind of data that is and in this case it might be customer customer data which can be very broad and then you start delivering down and that's almost similar to the answer earlier and you can say okay where does this customer data live some of us are going to wear a very structured database some of it's online some of it is through chat logs some of it is through call logs it is totally unstructured and to me it almost helps sell the need to want to manage that and then often the business I've found often it's the business folks the easiest to sell because they're like well how are you managing that well we can't quite have to kind of create a high structure on Hadoop they're like we'll do it right because if we want to report on it don't you need some sort of thing to report on so again I think the risk of starting too much in the weeds and focusing on hey we have this great new technology to do real-time data analytics without the why and how and what else is being affected I think that's often where we can get lost in the shuffle because we structure data modelers because people are focusing on the new shiny thing and on the business value and not seeing how it fits in with other things if you can build this new shiny thing and I'm sure some of us on this call have been called in after the fact of now we've got the shiny thing and we can't link it with you know the rest of the organization and that's what a right architecture helps flush out so yes and no so someone said they like my voice that just made my day I've never heard that before I just read that comment all right and I and I we my IT manager and myself always talk about getting distracted by shiny things we love shiny things myself I saw a great comment I was like oh shiny thing are there any freely available industry standard business capability models there are and there's good group I guess with any industry model like the same answer with the data model I would say the risk the good idea if you use them in terms of hey I wonder what other people are doing let's take a look at the guide big fan you don't want to reinvent the wheel I you have limited usage when you want to just because every every company is different there's some for insurance there's so yeah so it's great to think okay so think about how an insurance company they'll kind of have all similar types of capabilities we can probably start with that but I would just take a you know because one of the unique areas of a company is how you have your capabilities and strengthen them and organize them I wouldn't just take it verbatim so that might be audience obvious but yes there are depending on your industry or what you're looking to do and there's also some kind of forums and things you might want to ask other folks what they can share sure and I think we've got time here for one more question you know we're seeing suffer as a service brought up more and more as of course as the availability availability is readily available to companies and how do you construct data models from software the surface apps or you only have access to web services well so in the context of this this organization in this presentation when you think of back to an enterprise architecture that's where you also kind of want to look at the other areas in terms of the applications using them that kind of thing often you know there are sort of XML structures and it's more less about it's more of the data in motion so you want to think of the interfaces between those services and how it might you know when you think of more of a oh gosh mental block any of the cup of coffee you want more of the it's a very common term like my first name but I can't say you want that common model to see what data is shared across the organization so then you would kind of think of here's you know that's very much at the high level so what data is shared between this service you're not really developing a structure for a database you're trying to think of what that common shared data set is and how to share it appropriately there's just kind of a different way to look at the same more data in motion process or the interfaces which is what's nice about some of this enterprise architecture kind of look at the applications of who's using them. Well I'm afraid that is all we have time for today Donna thank you so much for another fabulous presentation and especially being so engaged this week after our enterprise data governor's online conference yesterday we just certainly appreciate all the education you provide the audience here and thanks all of our attendees for being so engaged in everything we do we just love all the questions coming in and the active participation in our webinars. Just a reminder I will send a follow-up email by end of day monday with links to the slides links to the recording of the session at Donna's contact information and I hope everyone has a great day Donna thank you thank you