 Live from Seattle, Washington, it's The Cube at Tableau Conference 2014. Brought to you by headline sponsor, Tableau. Here are your hosts, John Furrier and Jeff Kelly. Okay, welcome back to The Cube. We are live in Seattle, Washington for Tableaus, user technology conference, data 14 is The Cube, our flagship program, go out to the events, extract the civil noise. I'm John Furrier, the founder is looking at angle. I'm John Furrier, my co-host Jeff Kelly, big data analyst at wikibon.org. Our next guest, Jacques McKinley, VP of visual analysis at Tableau. Welcome back to The Cube. Thank you, it's great to be here. Jeff and I love talking to your customers because they all rave about the product. And Dita, the one thing that they all love is just it's easy to use, it's intuitive. And that really speaks volumes to some of the work you guys do around user interface, user experience, user testing. And as such a liberating tool like platform like Tableau, it's really changing how people are impacting their work and their companies. And that's what we're trying to do. It's very important to make it very easy to work with data because data is complicated and hard. And if you combine that with a difficult, complicated to use software, people can't get it done. If you make it really easy to use, we get very happy customers because they get useful work done. That's a hard nut to crack. Everyone's been trying to do that. I've talked to everyone in the Valley, startups, the big companies, they're trying to hire UX guys and gals, right? So okay, but it sounds easy on paper. Share some insight into what goes on behind the curtain at Tableau around some of the work you guys do, methodologies, mindset, guiding principles. Could you share some insight? Well, the place I'll start is that we're one of those lucky software companies that is working on software that we authentically use inside of the company everywhere. So our sales team is using Tableau to look at their pipeline. Obviously the marketing people would look at their marketing survey data, finance people would look at their finance data. The development team, they're looking at performance data and whatnot. So that gives us an authentic understanding of what our users are doing out of the box without having to go off and do studies and whatnot. We do the studies, of course. We're a data-oriented company. We love data, and so we do all the standard techniques from user research and blend that into the mix. But the core of it, I think, is that we are really on this mission to help people see and understand data. We actually live the mission, and so that's what drives our usability forward. And the staffing is always the conversation we always have to save about cloud first, mobile first, now data first, is not the first call. We coined that term because data's everywhere now, so it's not just a conference anymore. It's like if we were talking to someone who knows this is a big data conference, the guy from Amazon, I'm like, no, no, this is a software conference, application conference. The data is resonant everywhere. Absolutely, and growing everywhere as well. I mean, it's growing as fast as Moore's law. People are realizing that they can transform their industries by using data if they have effective tools for it. I agree with it completely. I like data first. So the culture or mindset for companies that are transforming you guys, obviously you're using it internally, your customers are all happy, you've got great product market traction across the board. As you go forward, companies that you're helping are transforming themselves. Absolutely. And they might not have the staff and experience. So what's the cultural mindset for the customer about how they should hire, how they should start the process to get those right people on the bus, so to speak? Well, often they probably actually have the right people already in their organizations and it's empowering them and creating self-service analytics, the ability for those people to actually use the data to work on it effectively and then grow from there. And that's the process, of course. Those people will know who else to hire and it builds from there. What's the psychology behind you guys, your methodology? I mean, you mentioned you had some different kind of, the psychology of sociology of user experience. Could you share some things that you have vision around in terms of good software should be intuitive? Yes, okay, but now also you can't figure out every human's need. So how do you find that middle ground? What's the techniques? Okay, this will be fun. So there are two parts to this. I mean, I'm an expert on visualization. So one part of it is that humans have very powerful visual systems. And if you can put the data into a visual form, it makes it much easier for people to understand it. But Tableau at its core is not a visualization company. What we are, what we call ourselves is the visual analytics company. And I put the emphasis on the word analytics. It's about helping people see and understand their data, doing an analyst process. And that's where the human computer interaction side comes from it. So you want a person to get into the flow of analysis, to move through visual views of data to get to an answer really, really easily and smoothly. And the core thing we do when we hire user experience designers is we look for people who are very good at understanding the cognition that people go through in the process of it. Jock, I know it's interesting because, you know, I remember in the 80s when I was getting my CS degree, I had to take a mandatory class called humans and social, computers and social change was like a mandatory class. Oh, email's going to kill social interaction. So there's always that kind of fear out there for the folks who are in the trenches. But you guys from hearing from your customers, their success is creating more robust interactions in and among the company. Explain why that, why is that happening and the benefits of that. There's a couple of things. The first thing is that if you go through a process of discovery like you can do in Tableau and you find out something interesting, the first thing you have to do almost always is go tell your colleagues about it. So it's actually that moment of discovery is a spur for highly social interaction, which is like yesterday I was the person on the keynote who talked about storytelling with data. And we see that happening all the time when you're, if you can discover something in the data, you go into a storytelling process and you need to tell other people about it. It's funny. We always roll our eyes when we hear storytelling because we're such big fans of the concept. And but in this world, people are trying to figure out what the story is to tell in that part of the process. So you guys tell stories really well. So I don't think that's our problem. So the top story that people are telling is, we have a story. Here's the story. So it's that process. So you got to set the data up. So that's the process. And is the art of storytelling changing at all from your perspective? I mean, it's a great creative thing. It does spark innovation. Well, I want to be a little careful, particularly I'm live here. Storytelling is a grandiose term. And my chief scientist, Pat Hanredhan, was one of the early people at Pixar. And he reminds me that Pixar storytelling is not only about a narrative arc, but also about that creating that emotion in people's lives. So often when you're actually talking about data, you'll have the narrative arc. But what you're really doing is you're explaining a finding. So I don't want to be too grandiose about storytelling. Sometimes, of course, that finding can have that emotional impact. So it does happen. But what's interesting and new here from the long history of storytelling, and storytelling that goes all the way back to ancient times, is with modern interactive computers, when you're trying to explain the finding in data, you can actually do that in very rich and interesting ways. And so it's a new medium that people are just beginning to figure out how to use. Us, we're co-evolving that with our customers. We listen to really great storytellers like data journalists and other people like that. Neil deGrasse, who talked earlier today, great storyteller, does a lot of stuff with data. And so, yeah, that's the interesting new part about it. Yeah, I like how you decoupled the narrative arc from the emotional delivery. And that's where the users come in. Absolutely. So I think it actually helps for a lot of people to understand that a lot of storytelling with data is in that explanation place. And so you don't have to be a skilled storyteller, as say, the people who design Toy Story or something like that. And it really is. It's there, the core of it. And this is actually also true. I know you're interested in how we do really good user experience design. It's a process of iteration and design critique. And so if you want to tell a good story about your data, you can iterate through the story and get it really well. You need a fast, easy tool for doing that. It's like doing a dry run of a presentation. You never get it right the first time. Exactly, so, although we're live here. We have to get it right, either way. We get it right or wrong, yeah, so, yeah. So, Chuck, how do you do, from a research perspective, how do you do your research? So you understand that humans interact with data a certain way and you're always learning new ways that, you know, to tweak the product, to approach visualization. How do you actually get that feedback and that research? Are you working with specific customers? I mean, what's your kind of research approach? So, a tableau of the research, that innovation comes from all directions. From the earliest days of the company, we've gotten tremendous amounts of feedback directly from our customers at the Cabell Conference once we started it. The sales team is, of course, talking to customers every day, and so they give us feedback as well. We also are very active in the research academic community. That's my background. Now there's a research team here. There's actually a bunch of PhD level people on the software engineering team. So, research at Tableau is actually infused out in the entire organization. Everyone is listening and communicating. It's because we're a company that's on this mission. We really are authentically on this mission to help people see and understand data. It infuses into the entire organization, and so the innovation is driven from that. Like, for example, in development, we have what's called in development a hackathon culture where one day a month, everyone does blue sky work, and some of those get into release, every single release, there's cool stuff that gets in. Are those like just spontaneous hackathon, like code jamming sessions, or are they kind of like scheduled out? So, everyone knows they have a day, once a month when they can work on whatever they want. They tip, it's not code jamming, it's more people working on an idea, sometimes collaboratively, sometimes two or three people. Like, often the members of my user experience team are collaborating with software engineers because you need both that, for us, you need both that sort of backend data part and that visual front end part at the same time. And so, it's a natural collaboration to have. Yeah, I love to study startups and we like to tell the stories, Jeff and I and Dave Vellante and theCUBE here. And if you can make something simple and elegant and easy to use that reduces the steps to do something, usually value, that's a good business model. Yes, that's our part of our... Exactly. So, I got to ask you to share with the folks that are watching or will watch that are developers, there's a lot of people who really want to make great apps and great user experiences in their design. What would you share with them for folks out there hacking away, building a new product, building a company? I think one of the things is you have to infuse it into the entire development team. This is a classic thing that I say to people on the development team. You can spend all your time polishing up a web page or of a dialogue, but you need to ask the question whether you even need that web page or dialogue at all. Cause if you can remove it, as you said, you remove a step and then you get to something more elegant. And so, that has to be infused out in the entire team because it's natural enough to just do the craft work of fixing the thing that's in front of you. But if you ask that deeper question, you get to the more elegant results. We were saying yesterday, because of the timing of the Apple announcement I watched, you guys are a lot Apple-esque, like you have a great team. You're flatter. Well, I mean, this is feedback from the customers. I mean, we see it and then they all say the same thing, but that is the culture of minimalism, right? To make something simple. Yeah. And it's not about the bells and whistles. It's to be really precise because it's important. It's both simple and useful at the same time. It's relatively easy to be simple. And if you're not useful, no one will care at all. It has to be useful and simple. And if you do that, then you get the magic to it. So, that's the core of what we do from using science. I love this topic, but Jeff, you can get to work. I just want to ask this question. Could you share some interesting insights about the way humans interpret visual, visualizations and data, things that you've discovered that maybe are counter-intuitive, whether we hear about companies testing different colors for a button on a website. Just what are some of the interesting ways that we interpret data? I'll take you through one of the standard ones that I gassed all the time. A lot of people, in fact, I was asked yesterday, are you guys going to do a 3D visualization? And so what I say to them is, well, 3D visualization is actually done today all the time. They're great products on it. They're used mostly by scientists. And it's because when you're working, say, on a 3D model or something like that, it's a very natural way to look at it. But from a user experience point of view, it's difficult. You have to worry about what point of view you're looking at. There's issues of inclusion and on and on like that. And so one of the things we know for how people work with data, it's actually better to stay in 2D and then use interactivity to move through views very quickly from one view to another. So there's probably others, but that's the one that popped into my head. Well, that's interesting because we see, obviously we see in 3D, but when you're, so it'd be natural, maybe that's what we want. But we read in 2D. Right, so. We're actually very, we're very skilled readers of surfaces of pages, so. I think that goes along with the simplicity aspect. Don't just kind of, 3D is, or it's kind of cool looking, but if it doesn't help you understand the data better then why do it? And the second part of it is, most of the problems people are working on, including scientists, by the way, are multi-dimensional in their structure. In fact, most of the problems people are working on are in multiple data sources at the same time. It totally overwhelms 3D, the number of dimensions you have and the number of different data sources. And so it's the interactivity much more than just the number of spatial dimensions that's actually the key for being able to understand data. I got to ask a computer science question, because again, just popping into my head just randomly, but the curriculums are changing. You're seeing data science come into disciplines. And in some cases, computer science in other areas. Coding is different now with the cloud. Obviously we hit Amazon earlier, this born-in-the-cloud younger generation are doing things much differently. But software business is somewhat the same, but can you share some insight from your perspective of the computer science curriculums out there, what they're teaching the young kids, what's good, what needs to be tweaked, any commentary around that? So I'm no expert on what the current curriculum is. I graduated a long time ago. What I can say is that the software engineers coming into Tableau are as cutting edges hired by Google and Facebook and all those companies like that. They're very well-trained, they're very skilled. You said something that was interesting to me, which is data science in particular is very close to what Tableau is doing. But the people who are called data scientists are typically extremely skilled people in programming, statistics, machine learning, database architectures, information architectures, things like that. If you're a person like that and you can learn all those skills, any company on the planet will hire you. It's a great job. It's not, however, the people we're building software for. We want to hire people who know how to build software for regular people who are trying to answer questions with data. It might be infused with machine learning. It certainly will have visual techniques, very sophisticated architectures for high performance computing and whatnot. But our end goal is to have really a wide set of people be able to work with data. You know, it's interesting, you mentioned those skills, machine learning. These are high-end skills. And we were talking with George Matthew, who's one of your partners at Alteryx and Nick provided the engine under the hood, if you will. And what's interesting about your success as a company is you've abstracted away those complexities in software. So essentially, you are a front-end to all of that magic. It's under the hood, but there are people that need to go under the hood and do the work, but you guys are doing that. So this is built into our fundamental DNA. The breakthrough from Stanford 11 years ago was partly that we could direct connect the databases. So we connect to some of the biggest data computing engines on the planet. And yet the people who are using our software don't have to know anything about databases or writing queries or any of that sort of thing, but they get to take advantage of that technology as well. So we're continuing down that path. We continue to add things like, like for example, a couple years ago, we added forecasting. And if you're a statistician, what you would know is how to use R or some package like that and all the millions of different forecasting models. What we did instead was we wrote some software so that we would automatically pick the right model for you because if you are not an expert in statistics, that's the sort of thing you need. That's our style of doing it. Lots of power under the hood, but packages for the regular person. Well, it's a great testament. We certainly, the testimonials are unsolicited when we have your customers on. I love this conversation. Thanks for coming on theCUBE. I want to extend a little further. We have a little more time. Well, we do, good, cool. Since the iPhone came out in 07, really that was a Mac-like moment that changed the smartphone business. You guys had a great demo yesterday with Elastic on an iPad. How is the connected device, then these new native users impacting some of the research you're doing and how do you look at the societal human impact to the always connected, my kids are all on the feeds and the Instagram. So the user expectations and user preferences are shifting. Some can argue good bad, but ultimately they are shifting. Sure. They're connected. How do you guys look at that? What is some cutting edge insights you guys are gleaning out of this and what's evolving? Well, there's lots of ways to answer this. The first of all is that we actually want to help people see and understand their information everywhere. And absolutely, we're all connected now and we have devices. And when you're out in the world, your data should be there with you and help with you. Now there's a lot of rich architectural issues there under the hood that you need to solve. We're working actually rather hard on client server architectures because the mobile devices just aren't as fast as the servers that they're connected to. And so it's fun for computer scientists actually. There's all sorts of interesting things that we can do and we are doing with that. So it's a real playground for the software engineers. The other thing is, of course, as user experience professionals, we know about how tactile humans are and how they can take advantage of touch devices. And it's mouthwatering only fun to innovate in that area to help you actually reach out and to actually authentically touch your data and play with it. So it's an area of innovation for us. I don't know if you talked about the Elastic demo yesterday on our keynote, but that's just the beginning of the innovation for us. There's a lot to do. And of course, when you're out there, you can also be disconnected from the server and you have to deal with all that engineering as well. So there's a lot of background engineering for it, but it is just natural for people to want to work with data. The other thing I'll say about it is of course, the fact that people are connected like this is increasing the amount of data that people want to do analytics on. So for example, we actually, this is another place where we do it authentically. We have a free version of Tableau called Tableau Public. Since it's free, we actually collect, you know, vast amounts of data about what people are doing there. And it's useful for our research. You were talking about their research before. We can actually there because we have the right to have access to it. We can actually validate whether our research ideas are good or bad and do it that way. And that's large enough that we're actually also authentically experiencing the size of data that some of our large customers have. So the actual core data in Tableau Public is in Hadoop. And then we can pull like Tableau Extrics out of it to do the hot analytics that you need to do. So, yeah. Kind of a related question, slightly off topic, but I'm just curious to get your response of talking about kind of looking at user data, experimenting with that and using that to support kind of your product development. What do you weigh in with, if you think back to that, the Facebook controversy over them kind of doing experiments with their user data? And you know, I think it was around, you know, if they show people more negative commentary, they have, it affects their emotions, that kind of thing. What do you weigh in on kind of the ethics of using data, user data, consumer data? And where do you kind of draw that line? How do you look at that issue? You're going to get a deep answer from me. Fantastic. So I'm not going to comment on exactly what happened with Facebook, except for obviously, they got some bad press out of it. Sure. But I want to talk actually at the deep level about what is the true value of data for people, including the kind of data you would get off of a site like Cabo Public. Well, the true value of data, well, the true, the true wellspring of innovation is actually, is an affable thing that people go through. We are very creative when we come up with ideas. The value of data isn't necessarily the source of innovation, it is a place whether you validate whether your hunch or idea is actually a good one or a bad one. And so if you think that you can just take a pile of data from Cabo Public, say, and figure out how to improve the user experience, you're not going to succeed. There's just, there's a lot of data there and you can look all over it forever and ever. Now it might encourage a hunch in you if you do that, but the value of all that data is if you think, oh, I could make it, you know, people are doing this and I could do that, you can actually go look in the data and see whether they actually are doing this and whether that would be worthwhile to do. Now I don't know the details of what Facebook was doing, but if they were trying to gather data to go after a hunch like that, they should have at least told people they were going to do it and not suffered the bad publicity about it, but that would have been the right way to use data. If they were like just wandering around and doing random experiments, then they weren't going to get innovation out of it so they shouldn't have done it at all. And it doesn't scale very well either from a scale standpoint. So anyway, I told you a deep answer. Yeah, well, I mean it's interesting because I think that's an issue that's going to become more and more important as this so-called big data theme technology approach expands. And we kind of live in this bubble here where we're talking about big data and analytics all the time, but I think the general public still doesn't quite get what's going on with all their data. I mean, there is a revolution going on with big data and data scientists in that with huge volumes of data and very skillful people to be able to process it, there can be, you can answer questions that probably couldn't have been answered before. The thing is that there are many, many questions we have that just good data analytics would answer. You don't have to be the super skilled data scientist to do it and data can help you, rather than just following your intuition that might be wrong, data can actually help you hone your idea and make it really effective. And so that's just the difference in style of, there are companies that specialize on those really, really big, difficult, hard, new kinds of problems, there is a revolution there, certain kind, and the public should be concerned that they might find out things about the public or individuals that they couldn't before. But for the regular person who isn't skilled in statistics and programming and all those sorts of things, we still have all this data here and we can use it to make our lives better. And that's more of the mission that Pablo's on. Jack, share with the folks out there. Let's bring this question on you. It's really designed to get a response from you in terms of just some random response. The most amazing thing that you guys have done in the team with the user interface. What are you most proud of? What can you point to saying? What we did here was amazing. I don't know. The interesting thing is, yeah, the interesting thing is, as I was going to say, the most amazing thing we've done is that we're building a Rockstar company, but then you narrowed it down to the user experience. That's good. We'll add that in there. That's two things. It is an amazing company. Yeah. And there are, you know, one of the things we are is, we're sort of humble. So I'm going to probably just skip the answer there. Because that's the area that I focus on. That's a good answer. Being humble means you don't want to, you don't want to pick one child, as I always say to my wife. You can't pick, you can't favor one child. They're all equal. And by the way, I will add on camera that we have a long ways to go our user experience could be improved significantly and that's what we're working on. Yeah. And you guys do a great job. Thanks. And for the folks out there, building software, again, any final insights, you want to share with people who are envious of the success of team that you have, because you have a great team of people. That's the core of it. If you're starting a company, it's to build a really great team. And the psychology, mix of talent, I mean, we're seeing. Multidisciplinary. Yeah. Is there anything that you can share? Like certain disciplines that you've seen work well, musicians, archeologists, we've seen a lot of big data folks come from all kinds of different. All over the place. And a range of people. Yeah. Okay, Jack, we got a break here. Thanks for your insight. Well, thanks for your time. Great conversation. Very engaging. I hope we told a good story. You got the data out of your head and share it with the audience. This is theCUBE, extracting the data, sharing it with you live here in Seattle, Tableau's Conference Data 14. Right back with our next guest. I'm John Furrier with Jeff Kelly. We'll be right back. Good.