 Hello and welcome, my name is Shannon Kemp and I'm the Chief Digital Manager of DataVersity. We'd like to thank you for joining the latest installment of the DataVersity Webinar Series, Data Insights and Analytics, brought to you in partnership with First San Francisco partners. Today, Kelly, O'Neill and John Lathley will discuss keys to effective data visualization. Just a couple of points to get us started. Due to the large number of people that attend these sessions, you will be muted during the webinar. For questions, we will be collecting them by the Q&A 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 DI analytics. As always, we will send a follow-up email within two business days containing links to the slides, the recording of this session, and additional information requested throughout the webinar. Now let me introduce to you our speakers for today. Kelly O'Neill is the founder and CEO of First San Francisco partners. Having worked with the software and systems providers key to the formulation of enterprise information management, Kelly has played important roles in many of the groundbreaking initiatives that confirm the value of EIM to the enterprise. Recognizing an unmet need for clear guidance and advice on the intricacies of implementing EIM solutions, she first founded First San Francisco partners in early 2007. John is a business technology thought leader and recognized authority in all aspects of enterprise information management with 30 years experience in planning, project management, improving IT organizations, and successful implementation of information systems. John frequently writes and speaks on a variety of technology and EIM topics. His information management experience is balanced between strategic technology planning, product management, and practical applications of technology to the business problems. And with that, I will turn it over to John and Kelly to get today's webinar started. Hello and welcome. Good morning, good afternoon, good evening everybody, wherever you are. This is John and hello Kelly. Hello, hello. All right, we are ready to go. We're going to talk about data visualization today. When we started to pull this together, we thought the best thing to do would be just play all of the really good examples out on YouTube. But that really wouldn't have done much for some of the how-tos and all of that. So we are going to talk about the best practices, some of the nuts and bolts of getting visualization hooked into your environment. The myriad of things you can do that are under the category of visualization and some insight into the different ways you can use it. And then usually our takeaways and as usual, we will endeavor to get to your questions as they arise. So without further ado, and of course, I, John will be calling on Kelly here and to provide you our experience as well as our knowledge about this and a whole bunch of other topics. We want to talk, before we get started, let's just talk about the definition. Because this is one of those terms that when you say it, someone gets a picture in their head of what it is and that might not be exactly what you're thinking of. So there's a general term that it is a visual context for any data presentation. Okay, that's one way to look at it. Another way to look at it is to try to show patterns and trends and correlations that you might not see unless it was visualized. So data visualization is not just presentation, it's also something that can be interactive. It is primarily, however, a means of communication and that is to be clear. Of course, being visual is the old adage. I promise myself I wouldn't say it, but the picture does say a thousand words, right? And that's a really profound reason that visualization is such a strong component of the information management world. Kelly, anything to weigh in on our definition here? We do have a poll coming up here, but anything else to add here? Yeah, I think it's kind of a broad category because there's aspects of data visualization that exist in other categories of tools and capabilities. So for example, data visualization can be part of your BI and analytics solution. It can be part of your data discovery solution. It can be a capability to help you better understand quality levels of data, for example. So one of the things that is both interesting and complex about this is that it is a way of improving all of those different categories, BI and analytics, data discovery, and more traditional just data understanding, data quality. Yeah, absolutely. That is a profound point. The data quality example is perfect because, again, someone will say, visualization, you get in your head, well, I just have to show the results of my BI and reporting stuff. But it's a tool that you can use anywhere you're going to be interacting with data. And you have to step back and be really open-minded about that. Visualization can be used anywhere. And we do it a lot in our everyday lives and don't even know we are doing it. So let's move forward here. Then we have our poll. I think it's the next, there we go. Sorry, that took a minute there. Now, before everyone answers this poll, it's about our polling questions. We do this every webinar, and we get about a 50% response rate. And it's okay to be shy. I mean, it's all right. But something to think about when we adjust our content and our direction and our tone and context and scope of all of our work based on these polling questions over time. So if you're missing something and you want to convey something and you, for example, you're very experienced and you want to see advanced techniques and we're covering basic stuff. We need you to indicate that you're advanced in the poll. So we have a good readout of the population. So just for those of you that are a little bit shy, it's okay to be shy. But don't be reticent just because you don't want to, because it does give us a lot of feedback. So first question, are you or anyone actively using visualization? And we have a set of questions there. So please check off the box that is appropriate to your situation. And then what impact is data visualization having on your organization? And, you know, it could be really, really big in some organizations and it could be whatever in some. And we'd like to know that as well. So we'll take a minute here and let us know what you... Sorry, John. I jumped in when you started talking and I opened it and it's already closed. But let me publish the results out for you. Oh, okay. Well, okay, we'll get some data visualization here then with the results. Oh, there we go. Most folks are using it and are getting helped in several ways. And that's very, very good. So A, wins on the top. B, wins on the bottom. And we still have a lot of shy people, which is perfectly okay. So let's just move on then to our next thing. So why do visualization? And Kelly and I will kind of, you know, peel down this one one at a time here. Large amount of data understanding. Well, that makes perfect sense. It's really hard to look at a few million rows of something and you make some sense out of it. And of course that leads to explaining what you want to explain out of the data a bit more readily. And then of course that would clarify your reaction to what you're seeing. Data overload is a big problem. And when you are overloaded with data and someone tells you it's in the data and because it's in the data you should have taken an action, that is probably a good sign that you need data visualization. Even though that picture says a thousand words, that doesn't mean your thousand words are relevant or that the picture is relevant, you need to understand, you know, you need to have that intent of what are you going to do with it. And there's always a lot of good warning signs that you need to do visualization. Obviously the trends are there. And it's really easier to recall a picture than it is, again, a bunch of rows and columns. And it is kind of more interesting. Kelly, over to you. Yeah, I think so. We have been using, and I'm using the global we, I'm sure everybody on the phone, have been using data visualization in some form for a while in the sense that from a very simplistic, we would always want to translate our data quality metrics into something that was a bit more consumable by people. So from a very basic level, we've got the, you know, red, amber, green. We've got charts. We've got, you know, how many, you know, what is a null, what is a, you know, how many are considered to be complete, et cetera. So we've been using visualization for a while. I think what's making it different now is the vast amounts of data. And the opportunity to use more compelling visualization capabilities. And this is all to support the concept of storytelling. And I think storytelling is just becoming such a bigger communication vehicle that it's important to tell the story about the data, as well as the story about what the data tells us. So that you can trust the, what the data tells us if you can trust what's in the data. And again, when you're dealing with massive volumes, this is really critical. Yeah, absolutely. So I think we all kind of have a sense for what it is, what it means, and why do it. Let's talk about some of the best practices that we've come across and would like to convey along the way here. First and foremost, know your audience. Not everybody will react to the same picture the same way. We've used many formats and presentations of assessments and surveys. We've done in our own work and visualized results in our own consulting work. And I can attest to you that just this week, the actualization that has been perfectly adequate for five clients in the last several years was presented this week to a lovely group of people, a wonderful client, and the room was crickets. And it just was something that culturally they weren't tuned into. You really have to do know what is it you're trying to communicate. Is it multidimensional? Okay. Visualization is really good at above the X and Y axis, and you want it to go more and more and more. And Kelly, maybe you can help me here, the individual who kind of started exotic visualization who passed away earlier this year, has several videos out on YouTube that are splendid examples of multidimensional, multimedia-type visualization things that can communicate to a huge audience. So that's a really, really important concept. Kelly, do you remember offhand that person? I don't know why. I wrote it down and it's underneath something, and I don't want to disrupt the microphone or anything here searching around for it. And we'll get to it by the end of the presentation. Is it Hans Rosling? Yeah. Hans Rosling? Okay. Yes, yes. Just for those that like YouTube and such, go out there and see some of the work that he's done on visualization. And it's not digital either. He's done visualization with like Lego blocks and storage containers. You can do it with anything. It's pretty cool. I guess, Kelly, that gets to the next one. It's got to be simple, right? Yeah. Easy to comprehend and meaningful. A lot of folks are very, very used to an X and a Y and a bar graph or a plot or something, and then if you lay something on them that is trying to depict four dimensions, if they don't know it's four dimensions or three dimensions represented by using bigger circles and smaller circles on X and Y plots, then you're going to get the question why is that circle bigger than the other circle? All right. You really do have to make sure that it's meaningful there. And that means having a good legend and explanation. Go ahead, Kelly. Sorry. I was going to say self-explanatory. So one of the best purposes of data visualization is so that you can understand quickly a lot of content. So easy to comprehend self-explanatory. Yep, yep. The next one, this is one that Kelly will have a lot to say on because we have conversations about this all the time with our various present stations and things we do. You have to tell a story with your visual. Now what is a story? A story has structure and it has flow. Stories have beginning, middles and ends. There is a resolution. There is a, in a business story, there is usually a problem or a need to understand a presentation of a solution or insight into what needs to be understood and then some type of action or something after that. And visualization can get you to that without that's by PowerPoint. Kelly, you're really good at the storytelling and done a lot of research in that area. Yeah, you know, I don't think I'm any better than you, John, just to give credit where credit's due. But this is something that's really important in terms of getting your point across. And I think part of why John and I were really interested in having this webinar is because as data people, we do struggle with convincing people to do things, right? Why is this important? What's the business value? What's the business outcome? Why is this a good investment, et cetera? And so I think this sort of data visualization discussion can help all of us to make it more compelling. While we're, you know, name-dropping another author and presenter that I like a lot is Nancy Duarte, who wrote Resonate a while ago and her most recent book, Illuminate. And she talks about the flow of a story also. So again, when we're thinking about when to use it, or how to use it, it is a tool in our little kit bag about how we can make something that might be a bit boring or mundane, more compelling to an audience that has a lot on their mind, a lot to do, and many shiny objects to chase. So visualization to create that shiny object that they want to chase and invest in. Yep. While we're on this, we had a question just come by and a question is appropriate now. And that's someone whose organization seems to be kind of stuck with graphs in charge with Excel. And that's not dissing Excel, because that's kind of your entry point to visualization, and it certainly is a prototype even for more exotic visualization. But the question here is how can I help my organization understand the use of correct visuals? And Kelly, I'm going to offer something and then hand it over to you. I think showing them what has been done with visuals and all the examples that are out there is your absolute best way to do that. Your thoughts on that? Yeah, I would agree. And one of the things that's coming up in future slides is just samples as well as lists of the types of visualizations that you can tap into to address whatever the concern is or the question is. Beyond visualization that's in Excel, because realistically that is Excel visualization. So we can go through what some of those are and we will show you some of those visualization techniques to tell the story. To wrap up here, obviously the technology is really exotic now. And this is one of those few times you're allowed to deploy every bell and whistle so you can get your hand on and try it out. Hopefully something will stick as you try it. Really key, and we're going to talk a little bit more about this here in a minute, is the sourcing and methods and timing. A lot of time visualizations are done and they're not done in the right context or understanding of the source and they tell a story but they tell the wrong story. The problem with the visualization is you don't know it's the wrong story. And to use an overused term, you invent fake news without even knowing you've invented fake news. So that is the best practice. There's no where it's coming from, no where you're collecting it, and to know the timing and the timeframes with that. Anything else before we move on here, Kelly, on our best practices? Nope, we've got a call. We call that a no. No information. This is really important because this is important enough that we're going to dedicate a few minutes of this a talk to it. Standardization or curation or harmonization of the data is absolutely necessary for visuals. You can throw stuff into the lake. We've talked a lot about data lake architecture. You can throw a lot of stuff into the lake but if you just point visualization and figuratively draw a circle around a bunch of data and hit the button, you're not going to have any idea what's going to come out there. There is an awful lot of work to be done and you need to factor that in. A good friend of ours and a former client and colleague of ours who does a lot of data science now has indicated it's 50 to 80% depending on the project or what they're working on is just getting the data ready for the analysis and for the visualization. You really need to have that time to market as an awareness for yourself. Look at what it is you have to do. You have to get the right reference data. When you do visualization, you are plotting across x, y, z axes or four, maybe five even dimensions. Those dimensions are in essence reference data. They have to be correct. Those dimensions might live in a taxonomy. You have to address the correct level of the taxonomy. You don't want one dimension of your visualization to be three levels down in a hierarchy and another one only one level down. You're going to be looking at a sum versus details and that of course would be distorted. So all that has to be lined up after the gathering. A lot of times you have to pre-join. If you're in a data lake, that means some standardization zone which we've talked about before in our talks with. More than just the raw data, you have to shape it so that when you point your visualization to that or even move it into a relational format that the visualization people can better understand, get it ready so that it can produce the visualization the way you want it to be produced here. And again, this can take a good bit of time. I don't think people factor that in enough sometimes, Kelly. Yeah, I would agree 80%. I think you can find a lot of statistics out there on the web. And the 80% actually came from an article from Forbes over a year ago that talked about data wrangling and really the work that needs to be done in order to curate the data. So it is quite intensive and should be recognized in terms of resource allocation and investment perspective. Yeah, it's very interesting. Since the earliest data warehouse days in the early 90s, that 80% number holds up. We still have to spend a lot of time getting the data ready. Now, there's a good visualization to present to someone in management. Okay, moving forward here. And we're going to talk, I'm sorry, that went too fast there. All the patterns, just some examples here of what you can do with various types. We found these out on a nice presentation on the internet and nice enough that we cited the source and showed it to you. Obviously, you can compare, but then there's proportions. So A to B is fine, but A is bigger than B. Then is A related to B. Then what are A and B and C related to each other? Then, of course, what types of correlative concepts might there be? And then we can go to a different branch of visualization here, which is geospatial. Where is something, when something is somewhere, is it different than when it's somewhere else? Okay. If something is a part of a whole, does that make a difference in its utilization or its context or its appearance? So you're starting to see here, as you go through these, if you're new to this, why the dimensional and the reference data aspect is so very, very important here. Now, distribution, maybe not so much. Distribution would be something, I'm trying to learn something about my data. I'm just trying to see where it lands across a bunch of different views. How things work, of course, would be causality. Processes and methodology, how are things meshed together? What are the timing? What are the cycle times? Things like that. Obviously flows, patterns, ranges, data changing over time, year over year is the classic example of that. We can visualize that. Analyzing text is with, and now with that, you do aggregates of concepts with text. Someone is expressing a concept. It makes a great use of a bubble type chart where you hear a concept more in a certain context than another concept. And then, of course, it's just references where things are heard and where are things not heard and where are things used. So that is the start right there. Just what is it you would like to show on your visualization there. Then, moving on here, then the options within that are endless. And this is our squint chart for the day. I can do all of these prior things on the screen in more than some, and well, you can pick many of these various types of options to present what we saw on the prior screen. A lot of these sit in the Excel tool, which is why people talk about that. But you can get much different versions of these in other tools, plus you can animate them and things like that. Kelly, anything to add on this vast variety here on the visualization options? I mean, we can pick one and read one. There's more than you can almost comprehend. I would absolutely agree that like you just said, starting with what are you trying to convey and what is the message that you're trying to get across, and then choosing the appropriate visualization to get that message across. And so starting simply, starting with the basics, we want to describe an event or a challenge or a problem or what have you and use that to drive the choice of the data visualization. I think that's really the most important thing to make sure, like you also said previously, that the visualization technique that we choose conveys the proper message, and we don't get lost in the visualization itself, but we use the visualization to get our message across. Yep. You can also, one thing with all these options, most tools will let you overlay various presentations for example, I can have a data plot overlaid to show values and trending, but overlaid that over a bar graph to show volume in association with the plots. So I can have a value of a certain amount, but then the higher the bar is underneath that value means more intensity. It's kind of like volts versus amps for anyone out there who's an electrical engineer. So we can do that. Now some tools will let you do that easily, others will not. Then again, doing that might overcomplicate the visual and again you have to know your audience, they might be turned off from that. You might need to do that a different way. Key here is in any good visualization process you will probably try at least three or four visuals before you find that one that resonates with your audience across a broad spectrum. A really good way to approach this is to just take a really simple few data sets and just see what they look like across all of these visualizations when you get one of these tools. Literally just play around with it. You'll start to trigger your creative juices really, really fast. So moving on here. Yeah, so sorry, just before you move on, one of the things that I just wanted to just share and reemphasize is that the different visualizations will resonate with different people. And so sometimes it's what you want to get across, but it's also who is your audience. And you'll find that things that are more graphically oriented might be more compelling for those types that have a more creative perspective versus those that tend to be a bit more linear, more traditional might be more consumable for those that have a more traditional perspective. So for example, showing some sort of fancy scatter plot that's multiple colors or what have you, to a finance person might not be as compelling as just showing a very simple bar graph. So again, it's who's your audience and what message are you trying to get across to them in a way that they will be able to comprehend. And like John just said, sometimes you have to test a few before finding one that truly lands with that audience. Yeah, and kind of an extension to that. Again, I'm calling up a recent example I had out here in the field where we showed some benchmark data and the data itself was appreciated, but what wasn't appreciated is it made the organization look bad. And it wasn't going to help the case at all that we had the most elaborate, elegant visualization of the gap between activities within this organization and within other organizations that we were benchmarking against because it just made them look deficient. So it's not only just what visualization works, but also the politics and the culture and the tone that that sets. Remember, we react to numbers as human beings differently depending on the numbers. When somebody says there's a higher number and a lower number, we automatically have been trained in most societies on this planet that lower numbers aren't as good as the higher numbers, and people will start to get defensive and even understand the numbers that they're seeing, they're just worried about it. So you might have to rotate your presentation from a bottom to top to a left to right or a right to left or something like that to get your point across because if your audience feels bad when they look at it, they're not going to embrace the results. Even if the results are true, you still have to... The word spin is not the right word exactly because we don't want to change the results for acceptance, but we do want to alter the presentation to increase the relevance of what we've done with our data. So it's not as to just extend it. Kelly made a lot of really good points there about how you have to try and really understand who, but understanding who is a lot more than even their level of expertise or their profession it has to do with where they are and the entire context of the situation you're in when you're doing these presentations. Yep, good point. So there's some various categories here. So we just kind of collected these so we could just kind of show you what the wide variety of stuff is here in your graph and your plot area, which was really taking data and putting it out there by data point and then you start to visualize where things are either close together or lined up in a certain way and you can interact with the data with those types of things. And there's all kinds of them here. I'm not going to read them all. There's some, I think, the names, the old bubble charts and the radar plots and the scatter plots are a lot of fun, but candlesticks and Merameco are also really good way. And spiral plots are good ways of showing things too. So by all means, this is usually your starting point when you have a bunch of data and you're really not even sure what's in there and you have some ideas so you do that. The table presentation is good for that two-dimensional view of things, X and Y or left to right or by quarter, by month, Gantt chart type thing. Maps are terrific because you start to bring in the geospatial aspect and it could be a literal map such as a map of Europe and plotting things on that. It could be a symbolic or an abstraction of a map of an organization's distribution centers or something like that, but that's another good way to visualize diagrams themselves where you start to take the graphs and the plots and fill things in, smooth things out, and then there's the ever-popular category of other, which is things that don't fit in the other ones. Sometimes we just do things there, starting with the ever-popular pie chart and moving up through the pictogram and world-class type things there. Lots and lots of categories of visualization there. You could spend a lot of time becoming a really good visualization person. Anything on that, Kelly? We have some examples here. I think that that categorization to me, John, I think is really great because it can help people go through what are my choices and what are some of the things that I want to think about and use it as a reference point to say, okay, I do want to do something that's kind of graphical and plot-specific and I want to resonate with the audience that I'm working with. Or I'm trying to get across something that talks about issues from a global perspective. Maybe I can look at something that has to do with a map. I think this is a great reference that people could use to see what your options are. Absolutely. Here's some examples. These are public domain, easy to see examples. Simple ones like how people spend their time. So you've got your time and these aren't super, super easy to see. Our facility here doesn't let us necessarily zoom in. So various activities and then levels of time. Now that is a relatively simple technique there. You've got a time plot and a data point plot. This is an X and Y axis plot. It's really what it is. It still tells a profound story. Whatever that one is in the middle, they're spending a lot of time. I hope that's sleep. I can't tell. Maybe it's sleep. I don't know. So I just drilled into it. So the top one is sleeping. The second one is work. Work. Almost equal to sleep. Yeah. Yeah. Well, there's now this is one of those things where this is where you work with data. This is a good example. I mean, there was a lot of research that people either work too much or don't sleep enough, right? So this is, if you were a researcher in that area, you would start here. You would start and say, oh, we do a simple plot to access. And we go look at these points. They're pretty close to each other. Is there a reason for that? What does it mean? Again, visualization helps you work it out. If you were to look at millions and millions of rows of columns and things, you obviously wouldn't see this initially, but here it just leaps off the page at you. The next one, that's in the Grand Canyon. Red, of course, is something bad happened. And I can just tell you right off the bat right now, there are spots in the Grand Canyon that you want to be safe and not act irresponsibly by that plot right there. So you have a combination of geo and density and events. So that's a multi-dimensional type plot, but it's done on a map. And it's super, super powerful message. Did you realize on that one, Kelly, anything on that? Drilling into it doesn't actually help, but I would like to, again, just add to what you were saying, that the technique used here is multifaceted, right? So you've got the idea of plots, right? So this is a little bit like a scatter plot, right, where we've got concentrations in some areas, breaths in other areas. It's on a map, so there's a map component of it, but there's also a color component of it, because we think about red as something bad. And so it really is something that creates a very compelling, albeit not very, you know, positive message. But I think it's a good example of lots of facets overlaid in order to get a message across quickly. And, you know, from what I'm looking at, the little tiny thumbnail I have here on the WebEx, you know, I see red and I think bad, right? This thing speaks for itself. It tells its story without a whole lot of extra things. Again, powerful, powerful. With telling stories, and I'll let the next one to the right about snow leopards, that's an infograph, and that's becoming an extremely popular way of data visualization, where you're putting some words alongside some very, very powerful images, and the images are contextual, like the picture of the snow leopard. There are multiple types of graphical formats there. There's the map with the location on it, and then there's the shades of orange and brown plot there. But an infograph is a blending of a lot of techniques there, and again, super powerful, very creative message there. Absolutely agree. Yeah, infographics we're seeing is becoming very compelling walk around documents to get a lot of information across, using multiple techniques that are a way to consolidate, you know, let's call it a 25-slide PowerPoint presentation into a single sort of quick map. Yes. So another great representation. I'm waiting for the day when PowerPoint will have the infinitely-length slide, and you can just scroll that sucker. It's just paging. Here's an interesting one. The tweets, and the tweets are not just all tweets because that would be a wrong data set there, but tweets about GDPR. So this thing has, there's a conformational purpose. Where would you expect GDPR tweets to be? Europe. Well, there they are. Bunches of them. Where else might you think it? Well, places where you have a foreign presence, like the U.S. Eastern Seaboard, New York, Washington, D.C., financial centers, Chicago, and then the West Coast, the major centers like San Francisco, they're on the West Coast, and that makes sense. Now, where don't you see GDPR? Well, an area that is totally away from that, such as China, Africa, Australia, all right? So this is a way to confirm. This is a way to show intensity of interest. Importance of the topic. Again, really simple plot, but powerful with the G.O. presentation on it. And then lastly here, monthly average temperatures, Tokyo and London, and one of them is hotter than the other. So I know where I'm going on vacation. No, I have no idea. Again, this is something where you'd want to dive deeper into. But again, a really relatively simple plot. Again, there's nothing wrong with simple. If it tells the message, go with it. You've got the first one we looked at and the last one are very, very simple, two-dimensional, but they're very, very good, and they do what they need to do for the purpose they were intended, and they get their audience. So just a bunch of different types of things here. It doesn't have to be super sexy, super animated, but it does have to be relevant. And all five of these are extremely relevant right now. And some examples of tools we went through and here's the sources of Gartner and Forrester and our own sources and our own things that we run across. And on the left is what we see as a lot of the more popular visualization type tools. Some of these are purpose-built. Some of these are general, but these are the ones that are crossing the radar more often for visualization than other tools. And some of these, of course, are specifically built to do visualization. Obviously the first three or four there are really, really the most popular ones. And are almost ubiquitous. You're going to find one or the other of those just about anywhere you go. Now all the mainstream vendors, however, aren't going to be left behind. Microsoft Power BI has its visualization alongside Excel, but the other big players like MicroStrategy, Oracle, SAP IBM, they're all on the way or doing things like that. In other words, if you're going to do on there. And a question just came in on SAS Visual Analytics. Specific thoughts on specific vendors we don't do here, but I will say that was an error of omission. SAS should be on the traditional column there. They are actually the old guard of data visualization. Apologies. I will take the caller for that one. I left SAS off of there. But at the end, these are examples. Not going to get everybody, but just some examples here. Kelly, anything on there you wanted to add? Yeah. I think just to your point about whether SAS is on there or not, there are a lot of tools that have data visualization as a component of their overall offering or something that is data visualization tool that is added on to other tools to create that visualization component. And a lot of these have free trials or freeware that you could get off the web. So Tableau and Microsoft both have free versions that you can download. There's other solutions that we didn't indicate here because they're software is a service in the sense that you can input your data and they will give you an output. There's chart folders, a pretty simple one. You've got open heat map. It's another one, obviously, that does the heat map. So there's a lot more that you could try out and test other than just these traditional tools. So be creative. This is what it's all about is testing to see what's going to work. And there are free things on the web that you can play with more easily and quickly as a matter of fact in order to get a result. And then one other thing I wanted to add before we get off the slide. Sorry, are we running out of time? Am I getting cut off? No, no, no. We're actually good. Anyway, it's that there's also, you know, there's the natural language processing and querying component of all of these as well. And so the input mechanism is data, but ultimately the input mechanism for some of this will also be voice and other things. So that then the output is the visualization and we're pulling, you know, what's happening here is pulling together all different components of interaction with data in a way to present it in a compelling way. Yeah. Well, that kind of just leads to what I was going to say because this is kind of like watching, if you're a baseball fan, the baseball cardinals in the 1980s had a coach, Whitey Herzog, who would change the lineup every inning. And it's like watching a Whitey Herzog cardinal game. You need a scorecard. There's vendors coming and going on a monthly basis in this category because there's so much creativity and so many good ideas and the technology is powerful. The underlying technology stacks are so powerful and you're starting to see people like, for example, Kelly mentioned you can send somebody some data and they send you back a heat map or something like that. That's artificial intelligence. That's looking at the data, kind of picking out what maybe a good visualization would be and recommending one to you. That's pretty smart stuff. So we wanted to present just a list here. I had a click sense as another one. It wasn't here. Kelly mentioned a couple more. We could have filled this screen up with names and made it unreadable, but we just wanted to kind of touch on every corner of the spectrum. Someone did shoot us a question, which of these work with graph databases? Not many of them. And honestly, I don't know which one specifically works specifically with graph at this point because graph is just a horse of a different color and you need to know how a graph database works. Especially the traditional guys on the right are deeply embedded in the relational world. Most of the ones on the left are pretty embedded in the relational world or the Hadoop world. Graph is growing and perhaps when we revisit this thing again or next month we have a trends topic. Yes, this is a shameless plug. For our next presentation, we're going to look at the year ahead and do some forecasting. And one of the topics we are going to talk about are graph databases. So stay tuned on that one. And so now we're going to talk about this descriptive versus discovery. We wanted to spend a little bit of time on this process of using the visualization to just present something versus making it interactive and reacting to it. In other words, the visualization can present a conclusion. It can finish the story. The visualization can help you write the story. So if we take a look at descriptive analytics, what course descriptive analytics tells us what has happened and gives us a grasp of why something has happened. What is good for correlation and causality type things. And we can say that if certain sets of data set up certain conditions, this is going to happen. Or this is what's happened historically. We can also take our tool like even a simple plot and manipulate the data and go, wow, when this number was high, that number was high. So some visualization tools can be placed on top of data such that you can adjust the data within your tool and see what happens and see what looks different and see that there is a correlation with something then something else is going to happen. So it can be used for exploring your results as well. So it's really important to understand that this isn't just a presentation medium. This is a tool. Visualization is a way to work with your data and to noodle and to play and to be creative. Maybe you think there's causality but you know there's correlation. Then you can sit down with some of these tools and actually work with them and you don't have to necessarily be a super, super powerful data scientist to actually get to some profound conclusions. This type of process is leading to a lot of folks talking about the citizen data scientist where you've given a data set. It's a very elaborate data set but you're given powerful visualization tools and you can actually start to do a whole lot of data scientists just by connecting and dragging and dropping and billing networks and hyperbolic trees and things like that and just seeing what happens which is again why the list of tools in this area changes every month. A lot of cool stuff going on here but it's really important to understand whether you're in a discovery mode or a descriptive mode because the one descriptive mode better tell that story, better stand alone on its own, not need a whole lot of explanation around it like the Grand Canyon picture we saw. Others might want to just tell you something simply like the plot of the time people spend doing things but those will lead you to other thinking and other types of analysis. Kelly, anything to add to that when we're getting towards the end here? Not at all. You've got it covered. Oh, good. I did something good today. It's been that kind of day. All right. Our takeaways here. Well, I'm going to repeat myself because I think that's the key takeaway. These tools can be used to interact or manipulate as well as kind of an output type thing. And be clear what you want to do. If you haven't invested in a particular tool set yet, that would be a reason for a criteria for selection is how much interacting versus how much just telling a story you're going to do with that. Always tell the story with it. Stories have beginnings and ends. All right. And the Grand Canyon one tells a story of adventure and risk and risk management and when to be aware and how to be safe. And it does that with one page and some colors and a lot of data points. And that's powerful, powerful stuff. Play with these things. I don't know how much you've done with them, Kelly, but in the time, the rare times I've been able to sit down and play with a big pile of data, this is a lot of fun. You can really get your creative juices going. And there are free ones out there and set your personal goal for yourself to play with these things. You even know these will work with big data. You're going to have to noodle with it to get the context right, put some dimensionality in it. Make sure you don't have any outliers in there and things like that. And then lastly, don't forget that old statistical warning that you can run a bunch of data, look at it, go your rica and it can be absolutely totally irrelevant. Make sure that you're not getting your causality and your correlation is confused. Make sure you have a process to dive in and correct things. Kelly, anything to add to our takeaways? Yeah, no, I mean, I'm just kind of grinning to myself because we've talked about visualization. Well, if the data itself isn't standardized or is of poor quality and you're trying to show a result, you need to recognize that. If you're using data visualization to show that it's standardized, then that's also another way to do it, right? Garbage in, garbage out, you know, that never changes. So anyway, go ahead, Sean. Well, the first use, I will tell a story, my first use of data visualization was some time late in the 20th century, I'm not going to nail down the year, and it was a data quality effort and we were profiling and we do have a question out here, is data discovery the same as data profiling? Not in the context that we are using it in this discussion. Data discovery is diving into some data and finding out what it means and how the relations are. Profiling is, we use that in the classical data quality aspect. But anyway, it was a data quality effort. We profiled the data. We had enormous amounts of rows and columns that were supposed to match and didn't match and were supposed to be within a range and where and all of that. And we shoved it all, we did the 50 to 80 percent, put it all into Excel, did some preliminary things and then we found a fast tool that helped us with the visualization and produced some very, very powerful messages to an organization on the fact that the data was particularly bad when the perception was it was particularly good. And that kind of locked down the power of this particular sub-discipline within data management for me. Let's see here, Kelly, I think we touched a lot of our questions here already. Anything else to add here or throw in here, Kelly? If not, we're going to see if there's any more questions and then turn it back over. So over to you, Kelly. Sorry, I just wanted to, yeah, I just wanted to respond to the data discovery versus data profiling. I was trying to type into the chat and I don't know if it's my issue or the computer's issue. It's usually my issue. But anyway, data discovery also implies exactly picnic, our favorite acronym, picnic, problem in share, not in computer. Data discovery tends to imply location too. Where is my data located? Where is data profiling can be done within a system and once you know where that data is located. So that's just another kind of nuance of the difference. So anyway, that was all I was going to add. Thank you. That's awesome. That hits the questions. And at that point, just a reminder that because it's December already next month, my goodness, gracious, holy smokes. And we will take a look at our trends for next year. And Graf will be in there. And what's going on with the dupe will be in there. And a few other goodies like that. And we really hope you join us next month. And on that, back to you, Shannon, for a wrap up. Thank you, John and Kelly, for another fabulous presentation. And I'm going to start using that picnic acronym for that. Because that is just awesome. I think, you know, if you did a visualization on that, it would be a pie chart with a really teeny tiny slice where it was the computer. And a really, really big part where it was the person in the chair. That would be an easy one to see. I love it. And thanks to our attendees for being so engaged in everything we're doing, for attending today. As John said, we hope you to join us next month. And just a reminder, I will be sending a follow-up e-mail to all registrants by end of day Monday for this webinar with links to the slides in the recording of this session. Thanks all. I hope everyone has a great day. Thanks, John. Thanks, Kelly. Bye-bye.