Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Episode #25: Andy Cotgreave
139 Plays9 years ago
I’m happy to invite my good friend Andy Cotgreave from Tableau Software to the 25th episode of The PolicyViz Podcast. Andy is the Technical Evangelist at Tableau software and writes and speaks regularly and widely about all things data: how to analyze...
The post Episode #25: Andy Cotgreave appeared first on PolicyViz.
Recommended
Transcript
Introduction to Juicebox and Guest
00:00:00
Speaker
This episode of the PolicyViz podcast is brought to you by Juice Analytics. Juice is a company behind Juicebox, a new kind of platform for presenting data. It's a platform designed to deliver easy-to-read, interactive data applications and dashboards. Juicebox turns your valuable analyses into a story for everyday decision-makers. For more information on Juicebox or to schedule a demo, visit juiceanalytics.com.
00:00:34
Speaker
Welcome back to the Policy Vis podcast. I'm your host, John Schwabisch. Glad to have you with us on this week's show. I'm very excited to have Andy Cottgrave here on the show. Andy is the technical evangelist at Tableau Software. Andy, welcome to the show. Thank you for having me, John. Great to be here. Great to have you. Good to see you again.
Role of a Technical Evangelist
00:00:53
Speaker
Why don't we start by having you describe a little bit about what you do over at Tableau? Yeah.
00:00:58
Speaker
Yeah, my title technical evangelist. What does that mean? It's a great title. It's an evangelist, man. I know. Yeah. My first job 25 years ago when I left university, somebody at that job at that company had the title evangelist. Yeah. And he seemed to have the best job in the world. Yeah. By chance, I seem to have ended up with an equally amazing job.
00:01:20
Speaker
I'm very lucky. My job is to basically help people get to grips with data and visualization and analytics through a Tableau prism. Primarily, I'm writing blogs and articles and speaking at conferences around the world, mostly in Europe, and that's it. Ideas come into my head and Tableau
00:01:43
Speaker
paid me to get those out into the world. That's pretty awesome. Who would want to be doing that? Get paid to just think about awesome stuff. Yeah, it's been amazing. I've been with the company four years and was a data analyst for about six years prior to that. All my previous roles and skills seem to have coalesced into this great
Insights from the Tableau Conference
00:02:03
Speaker
job.
00:02:03
Speaker
Right. So you just got back more or less from the Tableau conference in Vegas. Yep. I'm going to assume you lost all of your stock options at the blackjack table. It happens in Vegas, John. What data work happens in Vegas stays in Vegas.
00:02:20
Speaker
So first off, how many people are about or at this year's conference? We had over 10,000 customers attending the conference. Pretty amazing. It's astonishing. Yeah. So what were your top takeaways? It was about a week or so, five days, four days? Yeah, so it was two weeks ago with 10,000 other crazy customers. And for me, well, the first thing is just how big the whole scene has grown. You know, obviously from a Tableau perspective, it's big. You know, our keynote is filling
00:02:50
Speaker
the MGM Grand Arena. We shipped our stuff out and Madonna played the next day. It's like, what? What? That's crazy. And so that's obviously good for Tableau. But it's just reflective of the state of data in the wider community and industry. So clearly things are growing. People have this urge to see their data in better ways. And I think one of the key takeaway for me was diversity and what people are trying to do with data.
Expanding Scope of Data Visualization
00:03:18
Speaker
clearly the core of what Tableau customers are doing is business focused data visualization, trying to get, you know, we're going to talk about this later, effective information as quick as they can. Right, you know, whether it's storing it efficiently, and then visualizing again, exploring it as quickly as they can. And that's absolutely the call. But it's, you know, that there's a bigger scope and a bigger desire amongst our customers, I think, to
00:03:43
Speaker
to understand how to not just functionally communicate their data, but actually engage their audiences too, whether that's in data journalism or within the business itself. I think seeing more design-led sessions and having more design-led conversations is really the key takeaways I'm getting from it.
00:04:02
Speaker
Yeah, that's good. I mean, I would suspect there's a lot of data analysts, researchers, folks who don't view themselves as designers, but be able to sit down in our session and sort of learn some key design piece. Yeah, absolutely. I think people have realized we, you know, Tableau pioneered in sort of business intelligence industry and, you know, the let's move away from the terrible crimes of 3D XL horror. And then, and that was great. But then it was like, oh, but people, people don't always engage with that kind of stuff. Right.
00:04:33
Speaker
it's now understanding how to get the balance right is one of the key things. Right. So Tableau's, you know, obviously the thing that people love about Tableau is you can make interactive visualizations pretty easily drop and drag, dump your data and go and go to it.
Do Visualizations Need Interactivity?
00:04:47
Speaker
Do you think that all visualizations have some sort of level of interactivity because you sort of see
00:04:55
Speaker
oftentimes like a static bar chart that works fine as a static bar chart, but then you add some interactivity on top of it. And sort of piggybacking off of that, where do you see the future of interactivity? Are we going to continue doing those sorts of charts? Are we going to? Obviously, we have D3 and those sorts of tools. But for most of us, what's the future of interactivity? 20 minutes to talk about this. Yeah, only 20 minutes to talk about the future of visualization. You can divide things up into many different slices, right?
00:05:25
Speaker
For a lot of businesses in, say, operational environments, sales leaders are trying to track sales. Marketing leaders are looking at lead generation or success of their campaigns. In HR, we're looking at employee tracking, employee satisfaction, and things like that. For operational dashboards, if we're talking about dashboards,
00:05:47
Speaker
in business environments, then I think there's still a lot of space for interactivity because I as a, say I'm a dashboard designer, I can design something and anticipate some of the questions you might have when you consume my dashboard, right? But I can't anticipate them all. So at the most basic, interactivity and filtering allows me as a designer of a visualization to allow you to explore that little bit further. And I think in the business environment,
00:06:17
Speaker
there's still a huge amount of space and need for that. I mean, I think further than that, people in businesses environments also need platforms where they can dive further into those dashboards. You know, at the moment I'm talking a lot about the concepts of sort of dead end dashboards. You know, if, if I produce a dashboard, it's going to be consumed by my team or my organization and that inspires a question,
00:06:43
Speaker
and then those people can't then answer the question, then it's a dead end, right? And so filtering will get you some of the way, but the ability to actually just open that dashboard or get into it, change the view, add different dimensions and measures is vital. And so that's more like providing the actual ability to do deep analytics through interactivity, right? So I think in business, we need these operational dashboards, but then,
00:07:11
Speaker
When I look at visualization more in sort of the journalist or the public facing world, I'm seeing a bunch of trends that I think are reducing some of the interactivity that people need.
Trends in Data Journalism
00:07:23
Speaker
So I see interactivity changing, you know, this year if you think about some really influential stuff or some great stuff. The New York Times did the Dawnwall interactive exploration of the climb up El Capitan, the first guy to solo it, I believe.
00:07:39
Speaker
And there, as you scrolled down the page, the data changed and the visualization of where they were on the route of El Capitan and Yosemite changed. And that's interactive, but not of the traditional exploratory sense. Another example, Peter Oldhouse, who used to be at the New Scientist, now works for BuzzFeed, did a really, really interesting article about how he's beginning to use GIFs
00:08:06
Speaker
and animations more to show data visualization. And I think I've seen that a lot this year. Yeah. Animation is really engaging. It's not interactive. It's not interactivity, but it's it's kind of got movement. Yeah. And I'm actually a big fan of GIFs for data visualization. And then the final example of what's changing was Aaron Pilhofer actually on this very podcast a few months ago. Yeah. Said like in about the fifth minute that
00:08:31
Speaker
What he's seen from New York Times, now the Guardian, is that people want simple charts embedded in the flow of the text. So he's gone from doing rich interactive stuff, pushing the world forward through the New York Times, through the word there. So actually, people want simpler stuff. And that, to me, loops back a little bit to the Dawn Wall of the New York Times. I think interactivity is more being driven by the scroll wheel. If you think, I'm looking at a piece on my
00:09:00
Speaker
smartphone or my tablet, and I'm learning to interact by scrolling. And as we scroll, the whole thing's changed. Bits of data come in, come out as and when they're needed. And there's been various different examples of that this year.
Aligning Visuals with Storytelling
00:09:14
Speaker
So I'm seeing that a lot. So I'm seeing the scroll wheel being more interactive, being more driving interactivity. I think the other thing is I am seeing a lot of data journalism moving towards just show me one simple chart.
00:09:29
Speaker
Yeah, you know, it's like I've got my body of text and just embed a chart in it. I was at an event last night, a data journalism event last night, there was a book launch and I was speaking to the data team from the Times and they were saying this, you know, it's like we just want to do simple charts that are embedded in flow. So I think the interactivity, rich interactive dashboards are maybe not being used so much in data journalism.
00:09:58
Speaker
But when you're doing a simple chart, one simple bar chart, you've still got the tooltip opportunity. Here's a simple bar chart. Just click on the tooltip and I can tell you a little bit more information.
00:10:09
Speaker
Yeah. This distinction is really interesting between sort of the internal dashboard so that the team can sort of dive in and get more information versus the external front-facing, I'm putting a graph or two or three, maybe embedded within the text. I mean, what I see too often is that there's a graph or a visualization, but the story being told in that simple visualization doesn't match what's being told in the text. So the text is talking about something in some order and the visuals sort of popped in.
00:10:37
Speaker
but it, you know, resorts the data in some other order and so they don't match. So I sort of see that as, you know, obviously not for the Times and the Guardian, but for other groups sort of trying to do a better job of linking the visuals of the text and thinking about is it a holistic sort of single approach to telling the narrative. That's true. Hannah Fairfield gave a great example and a great Kino tapestry conference earlier this year talking about
00:11:05
Speaker
think she said the revelation defines the prize and she was showing again very simple visualizations each she was swiping left to right in that case and each left to right swipe just built the visualization a little bit it was just a simple line chart about deaths as a result of the change in helmet laws as I remember. And it was just a little bit of text and a little bit of
00:11:27
Speaker
simple visualization that actually told a very good story. So her message and again the message last night is data journalism is still driven by good stories when the interactivity loses you.
00:11:39
Speaker
It doesn't add, it doesn't enhance the story, right? What's the point? Great, so those are all great comments. The other thing about GIFs and animations is Nigel Holmes, who was also on the show a few weeks ago, made the same sort of comment that he sees animations, he sees GIFs coming to be an important part of the toolkit. He's obviously done a lot with that for the New York Times, I think he's done a bit too.
00:12:04
Speaker
Very good. So I'll put those links on the show notes page so people can take a look. So I want to move on.
Critique and Debate in Data Visualization
00:12:09
Speaker
Of course, this is a shortish show, so we don't have a ton of time, but I'm going to open a huge debate and we'll just talk more quickly. So Steven Few, as you know, wrote a blog post about Alberto Cairo, who's at University of Miami. Miami invited David McAnlis from Information is Beautiful to give a talk.
00:12:26
Speaker
And Steven Fuze sort of said, well, Steven Fuze has long known dislike, I'd say, for Candace's work. So he said, you shouldn't give this guy sort of a platform to talk. So you wrote a post, put a post on Steven's site. So why don't I sort of let you maybe talk about what you saw in Steven's post and your response and where you see that discussion.
00:12:54
Speaker
Yeah, okay. And if every listener should go and look at the blog post, it is an amazing blog post, an incredible debate with a, well, who knows how long the comment thread will be by the time this goes out. Right. So the debate, Steven Fu is passionate that data visualization should be about communicating data in the most effective manner. Steven Fu's opinion is that David McCandless is
00:13:24
Speaker
the antithesis of this and is preaching to businesses that McCandless's infographic, data art, inaccurate possible style is not the right thing to do. And therefore, Alberto should have, if he's going to give McCandless a platform, at least provided a counterpoint so that members of the public at that lecture could have understood that maybe that isn't the best way to do things. Right. Right. There's a great debate. Now, I made a comment
00:13:51
Speaker
In fact, in my comment, I opened it with who says data visualisation should inform effectively. Somebody further commented, said, I read that and I'm like, we might as well throw in the town. I was like, oh, really? Interesting. So what did I mean by that? I'm going to make three clarifications first. First of all, anybody who knows me and has read my stuff knows that I care about people
00:14:16
Speaker
being able to communicate data effectively, right? That's what I've written. That's what I've been working on for the last eight years. If you're trying to communicate something and you're trying to do it in a business environment in particular, then doing things with the most effective way possible is absolutely vital. Second thing is, I really like Steven Few. I love Alberto Caro and I think David McCandless does some really good stuff, which is hugely popular. We cannot deny the fact that he's tapped into
00:14:45
Speaker
something that people love. The third clarification before I actually get to my point is what I'm really fascinated about is, and I talked about earlier, is the balance between functionality as prescribed by Stephen Few and beauty as kind of shown off by David McCandless. There's a tension in visualization, even in a business environment, we need to do something that is functional but engaging. And I think all, you know, Stephen Few agrees with that.
00:15:15
Speaker
I don't think he disagrees with that. That's the clarifications. I do care about effective communication of data. My point is that this is two words, data and visualization. Data, visualization. Again, data and visualization. At what point in those two words does it mandate how you visualize data?
00:15:38
Speaker
If you want to use data visualization for effective communication and being effective as possible in the Stephen Few manner, then do. Most of data visualization is about that. Most data visualization is done by people in businesses trying to find what the data is saying, communicating that to the organizations in order to improve the business, improve the world, but solve cure diseases, whatever we're trying to do that. So what Stephen Few, his approach is absolutely right.
00:16:08
Speaker
But I just see it as a semantic thing. That's data visualization with the purpose of effective communications. Stephen and I have had correspondence on this offline. The sentence is about, but if you're going to visualize data for the purpose of effective communication, I'm like, there you go. You've just taken data visualization and given it a separate clause, a clarification. So I think if consider somebody like Stephanie Posovich, brilliant
00:16:37
Speaker
data practitioner now. If you go and look at her work, say her amazing air quality necklace or her explorations of some of the books where she makes these incredible treats, that's data visualized. I don't think you can say it is anything but data visualized. Is it the most effective way of visualizing that data in order for people to get rapid insight? No.
00:17:07
Speaker
But it's art. And I think where Stephen Few and I disagree is he would say, Stephanie is a data artist. Stephanie is not doing data visualization. Whereas I'm like, no, she's doing data visualization for the purposes of art. I actually think, you know, we're kind of all on the same page. I just have a fairly semantic distinction between these things. And then the last thing is, you know, Stephanie also says when she's thinking about her work, she's trying to inform
00:17:37
Speaker
She is absolutely trying to inform, but she's trying to do it in a different way. She's trying to provide the gist of what the data is telling her. What is wrong with that? There is nothing wrong with that. I wouldn't make my business dashboard on it. It's a brilliant way of visualizing data.
00:17:56
Speaker
I see both of them. No, no, I think that and I think it comes back to what you're talking about earlier versus the internal sort of the internal approach to visualizing data, which is we want a dashboard, it doesn't have to be the most beautiful thing. We want to be engaging because we want to look at it. But the point is to dive deeper into the details and learn something. Whereas an external visualization, you know, everything's a spectrum, right? So you may move closer on the on the on the
00:18:22
Speaker
aesthetics piece on the external facing stuff, but closer to the functionality piece on the internal piece. I sort of wonder whether this is, you know, the sort of the motivation for Steven's post was Alberto shouldn't have invited him to give him a platform.
Diverse Perspectives in Data Visualization
00:18:40
Speaker
And I sort of wonder whether this is, even if you don't agree with someone, do you still invite them to come and let them share their ideas? Now, I can imagine a case where let's just say I have the opportunity to invite two politicians or whatever.
00:18:56
Speaker
I would probably invite a politician who had an opportunity who doesn't agree with my political leanings, but I probably would invite a politician who would sort of come in and spew off hate speech and racist speech. And I kind of feel like Stephen sits to this side of, on the internal part, which is we need to be able to dive into data and understand it.
00:19:20
Speaker
On that side, his view of who you're going to allow to speak and provide their views is probably a little more hard-edged than mine would be, I would think. Yeah, I'd agree with you. Alberto invited him. Most of the audience were Alberto's students.
00:19:44
Speaker
Yeah, consuming this talk as part of a wide ranging talk on data visualization, practitioners theory and best practice. Intelligent people can make their own judgments. I think Stephen Few is more like, yeah, but there were members of the public that who, you know, it's like you don't want to start them off on the wrong track. You don't want their first exposure to this thing to be inaccurate representations of data. Because
00:20:10
Speaker
Steven Few is actually, it's not just the data art side of it, it's the inaccuracies that wind Steven Few up. Which I completely agree with. So I've seen some things of David's that I just, that's just wrong. Like that's just, that data is being used the wrong way. And I get his point about you don't want to sort of introduce people to the, you may not want, let me just say, you may not want to introduce people to the field on that side of the spectrum. On the other hand,
00:20:38
Speaker
I don't suspect that McAnlis was invited there to give a workshop or sort of teach people how to do data visualization. That might be a different take. And also, you would expect people, like you said, who are intelligent, who are working with data, you'd think they'd sort of look around to see what other people are doing and sort of try to figure out what are the best practices. Yeah. And I absolutely agree. And I think I'm slightly
00:21:06
Speaker
slightly more passionate about this because there was another recent post by Stephen Feud where he dismantled the thought starter from extremepresentations.com and I don't remember the author of it and it was a little thought map of you want to do data visualization, what do you want to do?
Learning from Visualization Practices
00:21:25
Speaker
And Stephen Feud ripped it apart. Every point was kind of valid but his main theme was this is how people start off
00:21:35
Speaker
you are teaching them the wrong thing to do. And fair enough, but that thought starter, I printed out eight or nine years ago when I was first trying to get into data visualization. And I saw this thing and I was like, wow. Oh my God, this is a field which actually has, it's not just choose a chart in Excel, it's actually think about what you're doing. Think about what you're trying to show and here's some ideas.
00:22:03
Speaker
I like to think that I then discarded that chart after it being on my cubicle wall for about two years and then learned that there are things in that chart, which are incorrect. But that was a gateway for me. And I think intelligent people can take that gateway, be inspired by things like this, maybe be inspired by people like McCandless.
00:22:26
Speaker
and then discover the true path. But that's what happened to me. I started on the wrong path with the thought starter and ended up on the right path with the starter. And I think people, as long as people are exposed to the walled side of things, then hopefully people are intelligent enough to make their own judgments eventually.
00:22:51
Speaker
Well, I think it's a really interesting post that Stephen wrote. And as you mentioned, the conversation going on in the comments is that alone is worth reading. And you have to give Stephen a mental load of credit for actually starting these conversations, which I think a lot of people are sort of wary to do. But he's willing to go out there and ask these questions, which I think is a real service to the field and to how people are going to learn.
Value of Debate in Data Visualization
00:23:15
Speaker
Oh, absolutely. Read the post, contribute, and yeah.
00:23:19
Speaker
He is super duper passionate about what he's trying to achieve and well done, Steven, honestly. Massive respect for him. Great, Andy. Well, I'm sure we could talk about this for hours. And we have and we will, but we're going to wrap it up here so people can get back to their normal lives. Andy, thanks for being on the show. I appreciate you taking out the time. Thanks, John.
00:23:39
Speaker
And thank you listeners for tuning in to this week's episode. If you have comments or suggestions, please hit me up on Twitter or on the website. And please rate the show on your favorite podcast provider. It does help other people learn about the show and get involved in the comments. So thanks again. And until next time, this has been the policy of this podcast. Thanks.
00:24:10
Speaker
Again, thanks to my sponsor, Juice Analytics. For 10 years, Juice has been helping clients like Aetna, the Virginia Chamber of Commerce, Notre Dame University, and US News and World Report create beautiful, easy to understand visualizations. Be sure to learn more about Juicebox, a new kind of platform for presenting data at juiceanalytics.com. And be sure to check out their book, Data Fluency, now found on Amazon.