Brian Suda on Designing with Data

By Brian Suda

Scrunchup interviews Brian Suda on his recently published 5 Simple Steps book called Designing with Data

Listen to the interview

Download the audio (.mp3, 7.4MB)

Audio transcript

Designing with Data book cover

Anna: So today I’m talking to Brian Suda who’s just published a book under Five Simple Steps called A Practical Guide to Designing with Data. So could you tell us a bit about what it is and who it’s aimed at?

Brian: Sure. The book is called Designing with Data. Originally we were kicking the idea around of calling it Designing with Statistics, but that was a bit probably too heavy for most people, so we went with the alliteration and talked about Designing with Data.

This kind of stemmed out of just seeing lots and lots of bad charts and bad design here and there from just people churning out pedestrian Excel graphs. I mean I’ve seen local newspapers which write…spend loads and loads of time copy editing and obviously printing and laying things out, and then just have this horrible pie-chart smack there right next to the article. And there’s lots and lots of books about visualisations and how to make these really cool; flow diagrams and all this sort of really artsy appealing stuff, but there didn’t seem to be a whole lot of really just basic how to get a bar chart; when do you use a line graph; when do you… what’s the difference between a pie chart and a doughnut chart, so we sat down with the Five Simple Steps team and we kind of hammered out, the rough twenty five chapters on what we kind of wanted to see in the book and we finessed it a little bit and went from there.

Anna: I mean the thing that I noticed is that you barely ever use the word infographic, but it’s a word that’s thrown around a lot, so why have you chosen to stay away from this?

Brian: I think it’s a bit like, you know, any discipline; to get to being able to design and develop visualisations and infographics, you can’t just jump in at the deep end. You’ve really got to learn the basics and what I tried to do in the book is just cover some of the basics such as when do you use colours and why do you use colours and what can they be used for, calling out data or highlighting different parts of a charge in a graph.

I think once you’ve become expert at just generating really well thought-out and well designed bar charts and line graphs, which if those can tell your story really well, then you can move onto infographics and visualisations.

Anna: I like the bit where you put…”if I find you’re using doughnut charts, I’ll hunt you down”.

Brian: Yes, those are… I think that’s a horrible sort of feature-creep function. I’m sure somewhere, in some version of Excel, some guy was using pie-charts, and said “What if we put a hole in the middle of it and call it a doughnut graph?” and it’s…once you put a button in there, you can’t take it out and it stuck around in loads of successive versions.

Anna: I’m still waiting for bagel-charts! So why is it important to present data in this way?

Brian: I think lots of reasons. People forget that charts or graphs are complementing the article, the story, but at the same time, they’re telling a story themselves, and you’re always going to be introducing your own bias and your own view on the data by both what you include and what you don’t include.

I mean maybe you’re doing fiscal charts: how much money the company’s made year after year or even for your website: how many visits do you get month by month, hour by hour, and you’re always trying to tell a story, so if you can be able to create good charts and graphs, it’s all about how to….better story telling.

Anna: You talk a bit about things like chart junk and data to pixel ratio, so could you explain what these are and how they help people make good charts?

Brian: Sure. A while ago, Edward Tufty coined the term “data junk”, and he talked a lot about ink to data ratio, and in his book, ink to data ratio is the amount of ink that’s used versus the amount of ink that is actually representing data, so on a standard kind of bar chart, you’ve got labels, you’ve got some axis grid lines; those aren’t actually data. Those are just there because it’s part of the framework. Then you spend the ink to actually draw the bar charts, so you get some sort of ratio on the amount of ink that’s used to make the data versus the total ink.

But on the web and on-line and on computers, we don’t think about ink, because ink doesn’t cost money, so I went with the idea of pixel to data ratio, which is just the number of pixels being used to represent data versus the number of pixels used overall in the chart. So when you get the chart junk, that’s just when you end up using loads and loads and loads of pixels which aren’t necessarily advancing the data at all.

Anna: I guess this could be applied to web design as well. I mean now we’ve got things like iPhones and iPads, it’s so important to only show the most important information. So do you think there are common practices in designing with data and designing websites?

Brian: I think there’s a lot….again it’s when you’re doing charts and graphs, you are telling a story. And I guess when you’re developing a website, you are trying to display some sort of…this is my brand, this is the story about my company or about myself that I want to tell. Sometimes that might be lots of colour and lots of flash and lots of random junk. But maybe that’s the story you want to tell.

With data, I guess you’re always wanting to make it easy for the person to actually understand the values and the numbers and what it means. On the web I guess you’re trying to do that as well. If you’re a company, you’re trying to make it easy for people to contact you or buy your product.

Anna: So you cover accessibility as well in your book. So what problems can people have in reading a chart and what steps can designers take to make the visualisations more accessible?

Brian: There’s a whole chapter that I talk about with colour. Mostly we forget about that on the web, we can represent whatever it is, 16.7 million different colours. But that doesn’t mean that everyone can actually see all of them.

So in the book I kind of break it down, all the different types of colour blinds and disabilities where you might actually, if you say check out this red line, that’s our annual growth, and someone who’s red-green colour blind just might see two shades of murky brown or something like that, so kind of addressing some of those issues as well as I don’t have a colour printer. I just have a boring black and white laser printer, so if you also make this really nice, crazy visualisation where colours and mean heat maps and stuff, when you print it out, it’s completely lost.

Anna: Are there any websites that you think do this really badly, specifically?

Brian: Not off the top of my head.

Anna: Are any that do it really well?

Brian: Again, not off the top of my head. There’s lots of tools out there. It’s probably one of those things that if it’s done well, you don’t actually notice, because if it’s done well, you don’t refer to “see the blue chart” or “see this in green”. The person just says “the third line” or “the thick line” or “the line that starts with X” or something like that, where that might have been a logical…they’ve thought that through because of an accessibility issue, or it might just have been dumb luck that they didn’t use colour and therefore it is accessible, so I guess it’s hard to tell. When done well, you don’t even notice it.

Anna: Yes, I guess it’s something you’ve really got to bear in mind, especially with things like say pie-charts and line graphs as well.

Brian: Yes. Especially like with pie-charts when you print those out, if you’ve got a light blue and maybe a light green next to each other, when they all just print out in grey, all of a sudden you’ve lost where the division is.

Anna: Yes, and it’s lost all meaning

Brian: Yes, exactly. So that’s one of the bigger issues I guess with accessibility. And that affects everybody.

Anna: Oh yes. And you talk a bit about the new technology that developers use. Things like Canvas so they can start creating their own charts using live data. So what have you found that’s out there and how far have we come from where we used to be?

Brian: Most recently I’ve been playing quite a lot with SVG and that comes partly because in the book, we needed to be able to make a nice pdf which would be seen on screen, but at the same time, it’s going to be printed in a proper print book, and therefore the dpi is going to be much, much higher. I think on screen it’s 72 to 96 dots per inch, pixels per inch, but in print it’ll be 300 or more, so the quality is quite high, and therefore I was playing around a lot with SVG in vectors.

I think on the web as well, SVG is becoming more and more commonplace as more browsers are supporting it, so we can do nicer, smoother curves and graphs, so when people do zoom in or zoom out, or even print it, they get a much higher quality.

Anna: And what did people used to do to show charts? Did they just put a screen-shot in or….

Brian: That…I mean Google chart api has been a wonder. I’ve used it as well, but it just generates gif images or pngs, so….well there’s also a Flash version as well, which works great because you just throw it some values and it gives you a chart back, but that does mean that you are limited to what they give you and it’s going to print as a 72 dpi bar chart unless you make it really, really big, then scale it down for print.

Anna: Cool. So what was the process like for writing the book? Did you set your own deadlines? Was it quite difficult to get everything in, or did you find it quite easy?

Brian: I tried lots of different things. The Five Simple Steps team was really great in that we took it little by little and we tried to have just monthly phone calls, just to kind of check in and see where everything was. We also didn’t know exactly when everything was going to unfold, because they were lining up all these other authors and they didn’t know when they wanted my book out exactly, so in the beginning, we didn’t have a final deadline. But it’s also very nice because it is Five Simple Steps. It’s five sections, and within each section there’s five chapters, so it makes it very easy for me to just sit down and say, OK, today I’m going to write two thousand words for Chapter 1, and you know, we were shooting for…the book ended up to be forty three thousand words.

We were shooting for somewhere in the ball-part of forty to fifty thousand, so that means about two thousand words per chapter, which is kind of just a long blog post. So that’s kind of how I was always thinking about it. Like, OK, today I’m going to focus on Chapter 1 and I’m going to try and write two thousand words. Then tomorrow I’m going to maybe take Chapter 10, or if I find some great links, I’ll read up on them and see, you know, maybe something I read sparked my imagination and I’ll pick a random chapter and just kind of work on that. But it was really nice just to be able to sub-divide it into twenty five blog posts, as opposed to one book.

Anna: Sounds a lot more manageable when you put it like that.

Brian: Yes. And I don’t think there would’ve been any other way I could’ve managed to do it.

Anna: So what route did you take to get where you are? Did you go to University and what would you recommend to people who want to do what you do?

Brian: I’m not even 100% sure what I do, so….I went to school originally in the US. I was studying Computer Science and Software Systems, so I understand…I’m a classically trained programmer. I understand project management and estimations and Agile and all the sorts of these different methodologies.

I worked for a year, year and a half for a very small company. It was a lot of fun; I really enjoyed it, and then I moved to Edinburgh up in Scotland where I studied Informatics at the University of Edinburgh and did my Masters.

Informatics is still one of those weird terms that no one is 100% sure what it means and it means slightly different things to slightly different companies and universities, but it was a good opportunity for me, because I studied a little bit of artificial intelligence, dug a little deeper in databases, did some more kind of project management courses, did an entrepreneur course, so…it was good and it was a full one year extra Masters, but I think it was like eight different classes, and there was all sorts of different things, so it worked out really, really well and I got my feet wet in a lot of different things.

From there, then I moved here to Iceland and I’ve kind of been working for various companies since, doing some start-up companies here and there but basically I’m just quite curious, and I love learning more and more things, so I’m always reading, I’m always….I’ve got hundreds of things in my RSS reader, checking those every once in a while, just trying to stay current, as well as contributing back because anyone can read loads and loads of stuff and that’s great. But unless you kind of formulate your own ideas and just get out there and volunteer and…I’ll write and article for this, I’ll do an interview here and there, and then people will comment about your stuff. Some people will be, “oh this is great, thank you very much, I didn’t know you could do this with CSS3”, or whatever. It’s both a good feeling for you because people are actually using some of your stuff as well as you’re contributing knowledge back to the community, and you’re learning stuff yourself, because someone is definitely going to comment, and be like “could be a little faster if you did this”, or “actually, this doesn’t work in this browser” or all sorts of different things, and you continually learn both from other people criticising your own stuff.

Anna: So is there quite a big geek community in Iceland?

Brian: Well Iceland’s not that big to start with, but there is a good core group of people here who always like to meet up and chat and it’s good because it’s so small, you can just send a direct message to somebody and get your problem solved pretty quick.

Anna: So having the internet there it’s really nice to keep in touch with people?

Brian: Yes.

Anna: So I guess my final question is, what next? Have you got any more books lined up? Are you doing any speaking?

Brian: Not at the moment. I think I’m going to take a little time off from writing massive books, but it’s probably just moved into promotion mode, so always going to be writing more and more articles, try and just talk about ideas in the book and then continue to refine it. I posted a few links here and there. The book has a website, designingwithdata.com

Anna: I was just about to ask you the link for that.

Brian: So as I find new things or contradictory things, I’ll continue to past them there. I just wrote a quick short little article for Fifty Two Weeks of UX.com talking a bit more about chart junk, which isn’t in the book, so that’s the other problem with books: there are deadlines so as new stuff comes in, you’ve just got to be like…it’s not going to make it into the book.

Anna: And also I guess it’s only five sections as well.

Brian: I could ramble all day long on doughnut charts, but got to cut it off somewhere.

Anna: Well it’s been really great talking to you.

Brian: Yes, it’s been fun

Anna: Thank you so much for joining me.

Brian: Thank you very much.

Brian Suda

Brian Suda on Designing with Data

Brian Suda is an informatician currently residing in Reykjavík, Iceland. You can find out more about him at suda.co.uk