Upcoming Data Visualization Workshops
Data, Designed | Jon & Stefanie Posavec | Amsterdam 1/16
Data Visualization in Excel | Amsterdam 1/17
Welcome back to the PolicyViz podcast. I am your host, Jon Schwabish. Happy New Year everybody. I hope you had some time to rest, and relax, and recharge, spend some time with friends and family. I was fortunate enough to take some extra time up and spend an extended trip with my kids. We went down to Florida to visit my mom, spent some time at the beach, spent some time at Legoland, some awesome time at Legoland because it’s Legoland and now back to work.
So, I’ve got a great set of episodes coming up on the show for you over the next few months. People from the fields of design, data visualization, some authors and speakers hopefully you’ll be able to use some of the lessons that they’re able to talk about to improve the way that you communicate your data, visualize your data and talk about your data in front of an audience.
So, just a few announcements before we get to this first episode of 2019. First of if you’re interested in supporting the show, which I would greatly appreciate, please consider leaving a review with iTunes, Stitcher or Google Play or your favorite podcast provider that helps me out, helps other people know about the show and gets it out there a little bit more.
If you’re interested in supporting the show financially you had head over to Patreon. I have a Patreon page setup where you can donate a dollar or $2 or $3 a month that helps me cover the costs of editing the sound, paying for transcription services, web services all the things that I need to help keep the show running.
Now, couple of other announcements about workshops. I am doing a couple of workshops in January this month. The first workshop I’m doing is going to be in Amsterdam with Stefanie Posavec. Sthefanie and I have taught a few of our data design workshops. I will be teaching another workshop in Amsterdam with Graphic Hunters on January 16th. I believe there are a few slots left, just a couple of slots left, so if you’re interested in working with us a full-day workshop, a combination of a lecture and actually sitting down and making things. Please consider joining us on the 16th. The next day the 17th I’ll be teaching a full-day excel workshop and helping people expand the capabilities of excel to create better and more effective visualizations. If you are in D.C. or anywhere near D.C. I’ll be working with Brittany Fong on the 31st at the Urban Institute here in D.C. we’ll be teaching a full-day data visualization and Tableau workshop. So, again we’ll be talking about best practices in data visualization sort of generally, but also in the tool itself we’ll be making a whole suite of visualizations.
So, the early announcements I got those out of the way and now we can move on to this week’s episode, and so to kick off 2019 I’m really excited to have Neil Richards join me on the show. Neil is a data visualization practitioner and he is an information designer. He is based in the UK. He is a Tableau Zen Master. He does some incredible data visualization work with Tableau. As I talk about in the interview with Neil he sort of came onto my radar about six to eight months ago, and some of his writings has really caught my eye, his visualizations have caught my eye of course, and so it’s really a pleasure to sit down with Neil and talk about the work that he has been doing and what he is looking forward to do over the next few months at least going into 2019. So, again a Happy New Year, I hope you had a great break, and so onto the first episode of the PolicyViz podcast for 2019.
Jon Schwabish: All right Neil. Well, thanks for coming on the show. Happy New Year to you.
Neil Richards: Thank you Jon. Happy New Year to you too. Thanks for inviting me on. I’m very privileged.
JS: Well, I am excited that you would even consider being on the show, it’s a privilege. So, I’m excited to talk about all the work that you’ve been doing, especially I think you sort of come on my radar the last six to eight or so, doing some pretty incredible stuff with Tableau. When did you become a Tableau Zen Master by the way?
NR: Well, I think I became a Tableau Zen Master towards the beginning of this year. They tend to be awarded for a calendar year and I think it was roundabout February the award came through. I was somewhat surprised to say the least. I had to re-redeem out a few times. I thought I was being asked to nominate someone else for a Zen Master and I thought well I’ve got a few good ideas. Now, it turned out they wanted me to be one, so yeah that was the circle.
JS: Yeah. That’s right. Do you walk around with a little medal hanging around your neck all the time?
NR: Well, I should do really. I’ve got some Zen Master t-shirts but that doesn’t really count. People don’t really know what that means in rural Derbyshire.
JS: Right. Well, maybe you could talk a little bit about yourself and your background for folks to kind of get a sense of the work that you’re doing, and then we can talk about the work that you’ve been doing?
NR: Sure. Okay, so we’ve spoken about Tableau and I’ve been using that for about three years, maybe three and a bit, and I started using that in various jobs that I was working in. I think most of what I’ve done and I’ve sort of become known for is being outside of the field of what I do at work. So, it’s been maybe visualizations I have put online or maybe community activities that I have taken a part in and I’ve been quite prolific really both in pushing some of the boundaries or putting out some, what I consider, different or interesting or fun visualizations just to try and learn, and to improve, and to put stuff out there. Also, I’ve done quite a lot of blogging, which hopefully has been sort of relatively well received and I tend to use it partly to maybe sometimes you know something I’ve tried or a question I’ve had and I like to pose a question out there to anyone who is reading just in case they have any different thoughts on what I’m doing. So, those are sort of the kind of the main things I’ve been involved in really just sort of bit of blogging and a lot of visualizing and I’m just sort of having fun with it as a hobby really to expand my skills and my reach out there.
JS: Great. So, tell me a little bit about the non-standard chart types that you use. We’ve talked about it on Twitter a bunch, and I am always curious about why people decide to move away from the line charts and bar charts. You do a lot of stuff with circles and radar charts and small multiples. I mean I usually find them great and fun to engage with, and I think that’s one of the things you’ve told me in the past that people sort of you know maybe a little more likely to engage with them. So, explain for us your thought process when you are creating a visualization that has maybe non-standard or non-typical graphs ones in them?
NR: Well, that’s right. I mean partly it’s because I still consider myself new and learning all the time, and so I feel I’m always going to learn something new if I think outside of the box I might know that the technically best chart type to this would be a bar chart or a line chart or whatever but if I can think of something more engaging then I feel I’m going to learn a bit more in doing it. But also it is very much in the fact of being noticed and I don’t mean maybe noticed but I mean in this Twitter world of things that are flying by it’s nice to have something which grabs the attention sometimes for the wrong reasons even. One of my blog posts was whether we take things too serious and often people might not be overly enamored with the analytical message that let’s say a joy plot well I’ve put out because you know they might say, well, look we can’t see what’s behind that peak, we can’t see this and that, we can’t read the exact numbers like we could on a chart or a bar chart. But I think it’s just a case of getting that balance and one of my – I have many influencers and heroes if you like, but I am – it’s always great to hear someone like Nadieh Bremer say look beyond the bar chart, look for a new chart type, don’t necessarily think of the first chart type that you think of and I think there is better influence than that as someone who can do beautiful work in the field just to push the boundaries of the so called Normal Correct Chart Type. So, that’s actually why I like to do it and I got this sort of, ‘mantra’ is too strong a word, but the fact intrigue leads to insight that’s kind of my thought somewhat I do. If I see something and it’s intriguing and I see something that it makes me look then I am likely to spend 30 seconds, a minute, five minutes whatever the requisite time is just to find a bit more.
JS: Great. So, you have the sort of public persona where you are playing around – not persona, maybe that’s a strong term but you have this public profile where you’re playing around with kind of interesting fun data, you are creating things that are sort of different and fun, and you’re trying to, like you said, make a splash on the Twitter stream as it were sort of flies by. But when you are working with more serious data and you are doing something maybe for work or something that you need a stakeholder really to look at. Do you think in the same way like I’m going to share this with my boss or a colleague or this is a more serious topic, I’m going to fall back on the more standard graph types or do you still say look I still need to make something that’s different and engaging in a different way, so I’m going to try these other form?
NR: It’s kind of a mix of the two answers. You do know 90% of the time that whatever you might be thinking the correct analytical answer or the most sensible thing to do is the bar chart or whatever. But I feel like if you have the mindset that maybe a Sankey would work in this situation. I do feel you’ll learn with experience that you’d know it wouldn’t work and it is not the right thing to do, but if you have these things in your armory or in your vocabulary and you’re prepared to think of them that will be that one shot in a hundred where you could think, no hang on we could do something different here and why don’t we instead of explaining this. I was listening to Cole Knaflic give a talk with Elijah Meeks at the Tapestry Conference where you were in the recent podcast and he said about how he trying to make a big splash at Netflix with a really funky Sankey chart and it died a death, but he said he learned from that and he held that back and a few years later when the project came back he thought about it and he thought it needs more explanation, maybe it needs a better annotation, maybe it needs a certain different circumstance in which it can be used, and he found that there were circumstances where it was really well received a few years later with that additional experience. I am sort of keen to make the distinction between, almost if you like, work-life and fun visualizations and I wouldn’t usually advocate that somebody goes and does a wade and wonderful album cover design to show the latest and profit figures of something like that.
JS: Yeah. I did a workshop with Stefanie Posavec a few weeks ago and we were having this conversation about – I was talking to the people who were at the workshop about pie charts or something, and she and I were sort of talking about them and she said you know I probably won’t use a pie chart because I’m more on the side of – I would try to do something creative and totally different and sort of outside the bounds, and for me personally I’d probably be more on totally the other side, right, like first instinct as someone who sort of went through the standard economics field, right, it’s still like make a bar chart. But there is this whole spectrum of graphs in-between and I mean for me at least is the struggle to move away from that one end and towards the other end, and I’m always trying to think of why would I use something that maybe isn’t perceptually as accurate as a bar chart, but it’s more engaging, it’s a little more fun and like you said maybe it makes a splash on the Twitter stream as it rolls by.
NR: I think it depends what message you want to tell. I mean to me if you need to tell an exact analytical message then you shouldn’t necessarily be going to overly creative with your chart type. But if your message is why that number is a lot bigger than that or of these categories look which one really stands out. If you don’t need to see the number, you don’t need to see the 86.3 is way bigger than 19.7 there are so many different ways that you can do that with impact. Everyone knows that let’s say you shouldn’t compare the size of two circles because the eye might be thinking of it in terms of the radius, whereas really you might be looking at the different ratio of the areas and who can tell which is three times bigger, I can’t. But that to me is an example of you can tell the difference between really big and really small. You could tell the difference between those two circles – you can see that there is a big difference between those two circles just as easily and in a different maybe more impactful way than you can between two numbers on a pay.
JS: Yeah. I think that’s right.
NR: In some ways it goes the other way. Almost the first rule that people learn when they try and sort of pick up so called Best Principles is pie charts are bad, and the way of thinking is almost if it’s coming around the other way. I would happily use pie charts for a two or three segment chart that it has a lot of advantages that others don’t. So, I would never rule out using a pie chart, and again it’s sort of challenging the perceived ideas you know sometimes you cut a square on your dashboard that you want to fill and a pie chart would do that perfectly or a donut chart. So, I would never – however much of a reputation I might had first it’ll come here with something creative I would never rule out the easy or the simple or the obvious either.
JS: Yeah. I wholeheartedly agree with that. I think the field has sort of turned back towards pie charts I guess, I don’t know, like we don’t hate pie charts as much as we used to. I think for me the past year or so has been an evolution in my thinking about when to use certain graphs, which graphs are good or bad and I think I’ve come down that the only graph that I would say never ever use are those like radial bar charts where the bars like wrap around the circle so that each bar has a different circumference of the circle because they’re just perceptually incorrect. You have two values that are both 50 and they end up at the same spot on the circle and that’s just wrong, but I think other than that – I don’t know if you feel that’s right but I feel like everything else is up for grabs.
NR: I think you’re right and one person that I’ve been particularly influenced by, if that’s the right word, and I need to thank you for this. In one of your recent podcasts Du Bois I mean his visualizations probably break every modern rule in the book and yet they are so striking and get the point across so well. You can’t measure the length of those spirals and compare them to some of the other straight lines in there, but you don’t need to. So, I think within reason I am sure we can come up with some other examples where that even I would say no, and most people would say no too. I think so long as you use with care, and you use with the right context, and you use for impact, and enjoyment, and discussion purposes yeah I would say most things are on the book. You only have to look at Xenographics to see the amount of fun, interesting charts types around there if you want to go different.
JS: Yeah. I think that’s right. Let me ask you this, so tell me a little bit about your Tableau experience so you’ve been using it for few years now, you are Zen Master, you’ve got the t-shirts and the gold medals, and walk around with your feathers out when you go to the Tableau Conference. So, just can you talk a little bit about what is it about Tableau that you like so much?
NR: Yeah. I’m not sure I’ve given a great impression here, I’ve been stressing around with my Zen t-shirt. That doesn’t happen too much, I promise. What it is I like so much is I am under no qualms about it. It is a BI tool first and foremost and the way that you should use Tableau is to plug your data in and to ask it a question, and to get a really quick analytic answer, sometimes a sophisticated quick analytic answer, which it can show in a pretty nice visual form. You can use it for that and you probably should use it only for that, but it has this amazing extra add-on ability to be used as a data visualization tool. So, it’s the combination of those two things that I really like, and at work most of us use it for the first reason that I mentioned, but you can use it as a really sort of nice sophisticated but accessible data visualization tool to do things that others might consider data art or any type of visualization we want. The reason I like it is because the chances are most things have been tried, or most ideas have been thought about, written about, discussed and there is a community of people who can sort of help you. You can be stuck on the most simplest of things, and you can Google it and you can be pretty sure that somebody from beginner to expert has answered it or if you are interested in some of the bit more sophisticated or a bit more out there then people, myself included, like to show people what they’ve done, like to put up how to do these things, so you really do have a good support network out there. So, ultimately it’s through something that I’ve used for numerous analytical purposes at work has introduced me to the first sort of creative hobby, if you like, that I’ve ever used using data to make some quite nice creative things.
JS: Okay. So, I’m curious about how you got into Tableau. What were the data or visualization tools you were using that ultimately led you to go into Tableau and use Tableau?
NR: Well, prior to Tableau I hadn’t done much in the way of data visualization. I was very much a data person, but all I’d really done was couple of simple charts in Excel, and relatively simple charts in SPSS, statistical software as well, and really all I’ve done was those sort of market research data tables where they just want to see every question, and they want to see every question tabulated by every other question. So, it would literally be pages and pages of Excel charts with pages and pages of physical printout tables, so I am old enough that was back in the pre-internet days that occasionally it would be 400 sheets of paper all printed off, and then put on the back of a motorbike and couriered from one agency to another in a different part of the country. So, it was very much moving from a dry way of presenting data into visualizing data sort of made me see the light, and I suppose Tableau, it could have been anything else, but it was the first time where I hadn’t just been in to me reading and presenting the data to someone else who was going to find the story in the data or present it visually. It gave me the first opportunity to do that myself and that’s what I really found interesting really. I’ve always like data and data visualization. I used to by David MacCandless’s books for the coffee table that kind of thing that’s a sort of a way into and trying that kind of thing for myself for the first time.
JS: It’s also interesting how you like a straight shot into interactive visualizations, into building dashboards as opposed to like an Excel graph or a SPSS graph?
NR: Well, that’s right. Yes. I suppose it gives you the feeling that you are presenting something a little more sophisticated, if you like, that you are creating something that’s half way between a static output and something that’s brilliant which is what I’m trying to say but you know something that’s almost more like add feel to it or something that allows you to drill down deeper because that’s always been the case, particularly in lot of the data that I’ve dealt with professionally that people want to drill down deeper into their data and see things sort of sliced and diced in different ways.
JS: As opposed to the way you were doing where it was 40 pages and a sort of do that exercise is like…?
NR: Well, sure. Yeah, I mean it used to be a case of rather than drilling down it will be sort of showing every layer literally you know layered in sheets of paper like that.
NR: It was a good introduction into doing some more interesting stuff. Yes, I’m doing some structured dashboarding.
JS: Right. I want to make a statement. I want to get your take on it.
JS: As someone who is working in Tableau and playing around with different forms and interactive – I going to make a statement, and just tell me what you think. All right, here we go, ready?
JS: Ignoring the technical considerations, so ignoring how difficult it is to create visualization. It is more difficult to communicate a message in static charts than it is in interactive charts?
NR: Well, I am not really sure that I agree with that. I suppose firstly because most charts that you will see or most charts that most people will consume whether it’s professionally or in the media, they are still static and indeed I think most chart packages will still focus on static things. So, I think there is a great skill, and not a technical skill, but there is a great skill in getting a message across in static charts. I suppose particularly if you talk about interactivity I wouldn’t necessarily always think of interactivity as clicking through and getting different views of charts that kind of thing. Although that can add a lot to what you do, but even let’s say your text, even your annotations or your well written titles they may not be interactivity in the sense of the word that we think it but you are interacting with the chart, You are moving from the data elements to the non-data elements, and it’s the non-data elements which are just as crucial in telling that particular story. I think if that’s done well then you can tell just as much with the static visualization as you can with a dynamic or interactive one. I think in many ways if you can do that, I mean if you can convey everything you want to convey on one page, one image, one visualization that doesn’t require interaction or that doesn’t require a click through or hover through into something else in many ways that’s more powerful. I know the question said – again you asked that being more difficult, but I don’t think it is more difficult and I think, as you say, ignoring the technical constraints it’s probably just as possible let’s say to get that message across in a really good static visualization.
JS: Right. The reason I posit that statement and I am not sure I have a strong view on it to be honest is that in some ways or in many ways I guess in interactive charts or at least in dashboarding I should say the creator has the option to give the user everything, and to say go for it you know you have filters and sliders and whatever and you can go take a look. Whereas, in a static chart you have to make a decision about what you want to say and that may just be the difference between a dashboarding approaches versus interactive you know a different type of interactive graph?
NR: I think you are right, yeah. I think there are different approaches which require different skills, but I don’t think one is more difficult and let’s fact it I love the other kind of chart that you say as well. I should be an Andy Kirk here and give my answer as it depends because I think it’s probably quite easy to give an argument for both kinds of answer, and from pure dashboarding point of view certainly a lot of the data I will work with professionally you know there is a lot of data in higher education and if you put it all out there and give your client the opportunity to slide by all the different dimensions that they might want to you can create a very powerful dashboard which can give the user a lot more flexibility and a lot more power. I am not sure that it answers your question about sending the message though. I think if you want to send your own message, also I think if you want to editorialize and you know I think let’s face it, most specialization is editorializing isn’t it? Because you’re already choosing what data you put in the chart. If you want to do that in many ways it’s easier and to do with a – we put a Sankey dashboard it’s a bit like exploratory versus explanatory isn’t it?
NR: And act like different horses or courses.
JS: Yeah. So, let me ask one more question. We’re at the beginning of the year do you have a plan or a goal or a series of things that you want to work on for at least the next – you know the beginning of 2019?
NR: Yeah. Certainly it will be good. Recently I was at the Tableau Conference and I sort of came out with some ambitions if you like over the next year or some almost New Year resolutions if you like so that’s *** [inaudible 0:27:36] I see. Really I’m quite interested in the idea of collaborations and a lot of what I do I do by myself, but I love to work with more people, to collaborate with more people, I’ve got stocks in there some great people at conferences who’ve said why don’t we this or I really like that, I’ve got this idea and the more people you meet the more people have similar ideas to you or might want to take something in a different direction. The idea of combining skills technically, and ideas, and creative thoughts really is quite interesting to me. So, I’d like to do something like that this year. I could talk for a whole podcast about it. I was quite inspired by meeting Giorgia Lupi recently as well and that sort of took me right into the idea of thinking what can I do to be more like her to do some of the kind of stuff that she does, and I’ve actually sort of had to go to a project it’s recreating some of her – the Ted Talk badges in Tableau and that’s been quite well received with quite some interest. Yeah, so just a few ideas like that and really I still have this imposter syndrome idea that here I am on your podcast and I listen to your podcast to hear experts and people in the field and yet I am slowly trying to become one of those myself, and I am still in a little bit of disbelief there, so to try and build on that and to do more. I love the idea of training and talking more, and just sort of getting more of these conversations going. I like the conversations that we have here or the conversations I try and start on my website that kind of thing.
JS: Yeah. I mean we didn’t get to talk about the, I mean, you had this great blog post after Tableau Conference, and then you had this great post about Du Bois graphic. So, I’ll link to all those in the show notes so people can check them out and of course your Tableau public page so people can check those out as well.
JS: Yeah. Well, I’m looking forward to see what you come up with this year you’re on my reading list. Well, so Happy New Year and thanks for coming on the show. This has been a lot of fun.
NR: Oh, thanks Jon. Happy New Year. Thank you for having me on.
So, thanks for tuning in so until next time this has been the PolicyViz podcast. Thanks so much for listening.