 Hi everybody, welcome back. I'm Jeff Kelly with wikibond.org. We're here on theCUBE. We're live at Tableau Customer Conference 2013, just outside the nation's capital. Great conference. We've seen, we've had a great show so far today. We had Nate Silveron just a moment ago. We've had Tom Walker, the CFO, for Tableau on a little bit earlier and we'll have Kristin Chabot, the CEO, actually joining us in the afternoon. Thanks again for watching. Today on this segment, we've got a couple of really expert data visuals as practitioners. So at this conference, it's all about customers. About 4,000 people here, probably two thirds or more are customers. But we've kind of cut the cream of the crop here. Joining me today, Ryan Sleeper, who is the, works with a company called Avalytics, covering data analysis for that company. And also we've got Jim Wall, who is the principal at Broadband Metrics. Guys, thanks so much for joining us today. Welcome to theCUBE, first time here. Promise it won't be painful at all. So you guys won a contest, two contests, each of you, leading up to this event and you'll be competing in the Iron Viz competition tomorrow. So why don't you, Jim, start with you. Tell us a little bit about the competition and a little bit about how you kind of were able to secure the top spot in the qualifying round, if you will. Sure, so Tableau runs three contests throughout the year and the winner of each of those contests then comes to compete for the Iron Viz title at the conference here. So the contest I was in was a Civic Data Contest, which you took whatever data source you wanted and try to present that data in an interesting format. So I used some volunteer data that was available from the Pew Center for Research. And I think with any challenge like this, you got to find the right data set, but then you got to play with it and that's the great thing about Tableau is it helps you find the questions. Does volunteerism in America vary with red state, blue state? Does it vary with the amount of volunteer organizations that a state has? So you would never know to ask those questions unless you got your hands dirty with the data. And that's really, I think, the key to building a great Viz and understanding a good problem. So Ryan, tell us a little bit about how you got to this point. Take us a step back, what's kind of your day job and then how did you get involved in the Iron Viz competition? Sure, my day job, I'm a manager of data visualization and analysis at Evalytics outside of Kansas City. And I run a blog on the side called Online Sports Marketing Guy, which has really kind of become my sandbox for experimenting with Tableau and doing some fun things that clients, because we're an agency and a client doesn't necessarily need some of the fun design stuff or bubble charts that I might throw in in my own personal blog space. And the contests that Jim alluded to, I was actually in the first of the three contests and it was called the Elite 8 Makeover Contest. And it was earlier in the year and the whole point was to take a visualization that you had created in Tableau version seven and then make it over with some of the new capabilities in Tableau 8. So I used the design from one of my blog posts, Online Sports Marketing Guy, and it's a visualization that measures player value and major league baseball and just prettied it up. There's Tableau 8 unlocks a whole new level of design possibilities through their free form dashboard design. Ended up getting into the final eight that was set up like a bracket, kind of like March Madness, and then from there you were voted in and I won that contest. Well congratulations and Jim, why don't you tell us a little bit more about specifically the visualization you worked on to kind of get you into this competition? Yeah, it was on volunteerism in America. So how does it vary by state? What's it correlated with? What are some interesting trends between states and within a state, both ethnicity or by age group or by economic factors? So, you know, it was a single page of his but I think it maybe generates a conversation. And I think, again, the nice thing about Tableau is it's not a static visualization. I mean, someone can take that data, take the visualization, open it up in their desktop and dig into it to answer their question. You know, maybe they wanted a different group of ethnicity or maybe they wanted to compare two states more closely over time. So I think that's, I mean, that's part of the process you go through when you create these visits and I think it doesn't end with the visualization. I mean, hopefully it continues on with people that use the visualization and they can answer their own questions. So yeah, so it sounds like for you, it's, as you said, one of the roles anyway of visualization is to start that conversation. Yes. So yeah, I mean, so maybe expand on that a little bit more. What, I mean, what is, fundamentally, what does data visualization bring to an organization, to a business who's trying to become more data-driven? All right, so I spent 10 years working on telecommunications data where the process was typically, senior management would come up with a question and then the operational group would spend two weeks trying to answer it with data. And by the time you came back with an answer in two weeks, either the question had shifted somewhat or there was a follow on that would take another two weeks to answer. So I think the goal of visualization and the real value of Tableau and the work I do is that it's no longer a two week delay between question and answer. And the people asking the questions can actually answer the questions. What a novel idea. Yeah, exactly. And such a critical idea for really understanding the problem more fully. So that's been a, it's really been a complete night and day change in the way I work. And Tableau, you know, you don't need to know SQL. You don't need to, if the data's clean, you really don't need to be a data guy. You need to be someone who's interested in the problem, right? Which is a lot of people. So that's great. And hopefully the data viz and these competitions kind of show you what's possible with the good viz. Take it beyond replicating in Excel. What we used to do in Excel, we now do in Tableau. I think these contests and some of the work in the public gallery show you what you can do with Tableau if you push it a bit. Interesting, so Ryan, what's your view in terms of the role of data visualization? Do you agree it's kind of a, the goal is actually to find some, find questions and start conversations more than necessarily answering question A or question B? Yeah, to piggyback off Jim there, I definitely always design with the end user in mind so that we don't necessarily, as the analysts have to know every single question that the end user might ask, they can find their own stories in the data if you can make your dashboard in a flexible enough way where they can find their answers. And then from an agency perspective, another benefit of Tableau and data visualization is all this interactivity that you can create and all the automation that it allows, particularly when you get to a level with Tableau server. You know in the past from an agency perspective we may have to spit out a PDF report for example on a weekly basis. Now we can create a dashboard up front, upload it to Tableau server, connect it to a live data source and the end user at our client can log in anytime they want and get the answer as current as it possibly can be without having to rely on us to do the manual work of creating a flat file each week. So that's been a great benefit. So we just had Nate Silver on and he talked a lot about as well as in his keynote he talked about the whole causality versus correlation argument. As a matter in the big data world some will argue that it's more about correlation. It doesn't necessarily matter why something is or if something's going to occur. It just matters that it's going to occur and now that you know it you can exploit that. Versus the causality argument that it's important to understand the underlying reasons why something's going to occur or what the data's telling you. Where do you guys stand on that? Jim what do you start? Well causation is always better than just a correlation but it's also much, much more difficult to find. So I think nine times out of 10 the causation or the correlation is enough to better understand the problem and maybe that leads to a subsequent experiment that is trying to find the causation. So first of all is just finding the questions, right? And then picking one of those questions say well gee we see a correlation between the number of groups that a state has for volunteers and the volunteer rate. That's a correlation. Well is there a causation there? And maybe you can set up an experiment or maybe you can more carefully think about that question once you have it. So yeah, correlation, I mean correlation I think often is the first step to finding a good theory or a good hypothesis for causation. So of course as I mentioned and you described you guys are the winners of the competition of your individual competitions and you'll be competing tomorrow in the finals with Kelly Martin who's not here with us from a biz candy I believe. So tell us a little bit about the final. Sounds very nerve wracking, it's a 20 minute, you're live on stage, you've got 20 minutes with a new data set to create the best data visualization you can. Do I have the ground rules correct and how you guys feel about going to that tomorrow? I'm just excited to be here and both Jim and Kelly they're just so talented. It's great. There's not too much to prepare or get nervous for. Just excited, I think it'll be a lot of fun. I think they have a couple of surprises in store for everybody. So I think it's going to be a great time. It's nerve wracking and I think it's going to be fun and it's going to be interesting because I think there's always multiple ways to look at the same set of data and when you have three people that are sort of familiar with this kind of thing and doing it live I think it's interesting to see how different cooks mix up the ingredients. So that kind of leads me to my final question. We've got people watching who probably have a lot of customers and tablet users actually watching now as well as others that maybe use different tools but really are trying to get better at analyzing visualizing data. If you could boil it down to, I mean certainly there's multiple characteristics but what do you think is one of the, what is the most important characteristic in a person if they're going to be successful at really visualizing and communicating data to people in their organization? Why don't we start with you? I think it's to ignore the history of reports that your company currently has. I mean they're great and they have a long history and people are familiar with them but most often they're not making use of best practices, they're not really communicating as effectively and they don't take advantage of some of these newer tools. So I think that's the biggest challenge and I think it's really the key to success is to try to break out of the way we've traditionally presented data into kind of a new data visualization or data-centric view of the data. Yeah, I'd say there's definitely, people need to break away, I think anybody that's here at the conference understands that there's a wave with Tableau coming over and there's an evolution in data visualization and I still always, I don't try to pit Tableau against other software tools, I think they serve a place but with Tableau you're able to just take the visualization one step further, the interactivity one step further, the automation one step further and I don't have just one trait that will help people get to the point where they're good with Tableau but I think that Tableau has done such a good job of setting up a community that's very willing to help and one of the best ways I've learned is just to connect in Tableau public, Jim alluded to it earlier but you can download any workbook and you can really reverse engineer those so if you see something you like or something you may want to build in your own reports you can definitely download other people's work and reverse engineer it and that'll really help you grow through visualization skills. All right, very good. Well, Jim Wall, Ryan Sleeper, thank you so much for joining us, good luck tomorrow in the competition. If you're here at the event, definitely check it out. It's going to be, as you said, a little nerve wracking but I'm sure very exciting. So thanks again guys for joining us. Stay with us, we're going to take a short break coming up but we'll be back, we're going to be broadcasting all day, day two here at the Tableau Customer Conference live inside theCUBE. I'm Jeff Kelly with wikibon.org and we'll see you in a few minutes.