This autumn, I had the opportunity and pleasure of interviewing three of the top and world- renowned influencers in the field of data visualisation – Stephen Few, John Grimwade and Nigel Holmes. They arrived to present at the University of Miami’s Places & Spaces: Mapping Science conferences which showcased 100 visualizations intended to “inspire cross-disciplinary discussion on how to best track and communicate human activity and scientific progress on a global scale.” The exhibit and talks were brought to UM by its Center for Computational Science, the College of Arts and Sciences, and the School of Communication.
Whilst these three thought leaders share similar core perspectives in some areas, even to the point of referencing one another, the differences in their approaches and the angle of their specialties were a treat to witness. Here are some of the most inspiring angles they talked to in the three and half hours we had, slightly edited in length, of course.
Interview with Stephen Few
Big Data is just a marketing campaign. “I have an opinion of Big Data (BD) that runs counter to most opinions out there. Big Data as a functioning term is just a marketing campaign, and most of the talk about Big Data is generated by thought leaders and technology vendors that benefit from selling things that are related to what they are calling Big Data and they pushed the use of this term because it has ignited a lot of curiosity about what can be done with data, especially through the use of technology. The problem with the term, with the likes of other terms that become popular with technology, is that it is ill defined; no two people define Big Data to mean the same thing. The problem that I have with it is that it tends to encourage people to focus on the things that are not particularly helpful – and that is, chasing the latest technology to solve all their problems. What I call ‘Technological Solutionism’. That somehow if they just purchase these Big Data technologies and start using them, suddenly they will be able to do things with data that they’ve never been able to do before. And I think that is absolutely misleading. There is nothing in these new technologies that will actually let them do things if they haven’t developed some basic skills for data sense making.
“The only thing I find useful about the term is that a lot more people are talking about the potential benefits of using data for decision-making purposes and I think that is a good thing. We are facing a lot of challenges in the world that require we really understand what is going on. And we can only understand what is going on based on real evidence. Using data is a great thing, but we have to make sure that we use it in ways that are meaningful and that will make a difference in the world. To do that, it requires some basic skills to look at data, make sense of data, and communicate what we find to others in understandable ways. No technology is going to make you a data sense-maker, a data analyst or a data scientist. That’s what lacks from Big Data campaigns: there hasn’t been enough recognition that we need to learn some basic skills. That is really apparent to me as my work is focused on teaching those basic skills and so I see that as being left out in the Big Data push right now. The good news is that those basic skills are not that difficult to learn, but they are not obvious.”
Our brain is a critical consideration for worthy data visualisation tools. “Technology is essential, but only good technology is going to be helpful. Many of the technologies being made today don’t work very well. Technology for sense making serves as a tool for extending and augmenting the capacity of our brain – where our brain is limited, providing ways of working around those limitations. Data visualisation technology is able to do that when done well, but a lot of the tools being developed for data visualisation, are not made by those who truly understand the human visual system or how our brains work. And how do we build tools that help us think better and find what is meaningful in data in more effective ways? I don’t think what people are calling Big Data has much of an effect on data visualisation. I think how we visualise data in useful ways isn’t affected much by the volume of data we are working with, the speed at which the data is being delivered to us, or the different varieties of data, meaning the three V’s (volume, velocity and variety) of how Big Data is being defined. Because what works is based on how the human brain functions and that is not changing. Our brains evolve at an incredibly slow pace and so the good tools for data visualisation are based on how our brains are already wired. What we can develop in the future in data visualisation will have to be done based on a few fundamental methods because there are only a set of things that work for the human brain.”
There is no one data visualisation tool, only a handful for diverse purposes. “The tool you use should depend on what you are going to be doing with data in your work, because there are different tools that work well for different purposes, but not one tool that expands the spectrum for the purposes of data visualisation. For people who are going to explore data to discover what is meaningful in the data and then analyse what they find to make sense of it, there are only a handful of tools commercially available today that support that process in a way that works. These include only a handful like Tableau, TIBCO Spotfire, and SAS JMP (for the statistically trained). There are a lot of vendors that are trying to chase after the success of these top tools by emulating them in a surface way. But if they don’t understand why those tools work well; or how to develop their own tools based on the functioning of visual perception and cognition of the brain, then they will build things that don’t actual work, like some vendors already have, for example SAP. Some of SAP’s efforts in developing data visualisation tools for data exploration and analysis have been huge failures.”
It won’t happen by magic. “There is something I have noticed working with small to very large organisations. Even organisations that we think of as the most analytically sophisticated, such as some of my clients like eBay, Cisco, and Apple, are no more sophisticated than the typical organisation in terms of making better use of their data on a day by day basis. What is different is that they have pockets of specific expertise. The truth is that a lot of what an organisation needs to do typically on a daily basis to make better use of data can be done with relatively basic skills supported by decent technology. The problem lies in that the skill set that is required to do that just isn’t being taught in organisations. And I think the reason is organisations assume that the skills become automatic once they install technologies that are designed to support those skills. As an example, they decide let’s install a new business intelligence suite of tools with a product that is designed for doing visual data analysis, and assumes that everyone will be empowered to master it, but people have to be taught these skills and how to use the tool. They can get a great tool but if they don’t’ know basic skills, it will cause a breakdown. Very few organisations are willing to invest time and resources in helping people develop basic skills to use data better, they just throw things at them and expect to somehow in the midst of their regular work, develop those skills, although there are rare individuals with an aptitude that can do it on their own. What I have to say to those organisations is they need to recognise that it will not happen by magic and that the organisation has to make an effort to help people develop those skills.
For example, when communicating data, how do people judge what is the proper way of displaying the data they have received in reports or dashboards? Would a graph or table work best, or maybe both? If they use a graph, what type of graph – a bar graph, a scatter plot, a line graph? To make those decisions requires some basic understanding of how these different forms of display work, how they work for our brains, and when you would use one type versus another. Those things are very easy to learn, but they are just not intuitive; they don’t happen automatically. Those basic skills are some of the teachings from my courses, so people learn what form(s) of display will work best to tell a specific story contained in the data, based on what people need to see in the data, and what decisions they need to make with that data. At the moment, I’m focusing my attention on the basic statistical skills that are required when identifying what is meaningful and what is not within a data set (when there is something you ought to be doing with the data versus what is routine or random variation in the data).”
How we use data effectively needs to be questioned. “Long before you are concerned with how you visualise the data, you have to have some sense of what data you should be looking at, and what is going to matter to the organisation. There is a lot of time wasted tracking information that doesn’t matter to the organisation, and many times it’s because it’s historical; something that has always been done. People cannot imagine not tracking that information anymore. But if you were to ask them when was the last time you ever did anything in response to this series of reports you’re looking at, it’s amazing how often they say they never do anything with it in response. And that has to be challenged in organisations. You need to understand what data is going to inform you in ways that will better meet your organisation’s objectives, and unfortunately many organisations don’t even have a good sense of what their objectives really are. In general, organisations need to think more analytically so they can ask the right questions of their data and focus on what really matters to them. That means a culture change and has little to do with data visualisation but is foundational to use data in a more effective way. At a fundamental level, I think of my work as not being data visualisation; my basic work is to help people in organisations think and communicate more effectively. I just happen to focus on data because that’s where a lot of the evidence of what is going on is contained. And I happen to use data visualisation simply because a lot of what we need to understand that exists in data, can only be understood if we see it visually – in patterns and in relationships; and I use it because it works.”
Few’s work is inspired by Edward Tufte, and others. “I had been working in business intelligence for many years before I was really exposed to data visualisation that opened my eyes to its potential. And that was through the work of Edward Tufte, [an American statistician, professor and pioneer in the field of data visualization] when I attended one of his one-day seminars about 15 years ago and thereafter left that experience very inspired about all its opportunities. In the next few years, I dug into data visualisation further with the work of Edward Tufte and others and I realised that I had an opportunity to help people within the realm of business intelligence to take better advantage of data visualisation. Early on not only Edward Tufte, but also [Purdue University statistician professor] William Cleveland and Princeton University statistician John Tukey were very inspirational to me. Today, I get most of my inspiration from either my students who I also learn from, those few who do a great job at information visualisation research, and because I think being cross- disciplinary is important, I also look to inspiration outside my field. For example, from Daniel Kahneman, who has a background in cognitive psychology and is author of “Thinking Fast and Slow”, Daniel Levitin author of “The Organised Mind” and [University of Miami professor and author] Alberto Cairo, who points me to different things I have not seen before. People in any field who tend to contribute the most, and have the greatest insights are not focused
exclusively on their field. I seek to have exposure in a broad range of disciplines that allow me see connections and use critical thinking that feed my work.”
Interview with John Grimwade
A good graphic has a fine balance between information and art. “Having an inquisitive nature is a good start to being successful in information graphics, followed by having the enthusiasm to explain that to people. You have to have a sense of hierarchy, sequence and editing. It’s journalism of course. You have to be telling a story, and unfortunately that is what is missing in a lot of the graphics that I see today. They don’t start out in a clear way, as every journalist knows, of having a headline, an introduction, a clear, sequenced story. Those principles of writing a story in journalism apply exactly the same to visual journalism. The story has to be clearly arranged so that everyone can understand it through the graphic. It’s all about explanation, “turning on the light”, and clarifying. So I’m nervous when I see many graphics today that don’t turn on the light and I end up more confused than when I first looked at it. That worries me, because they were not following these principles and that person was not thinking of their user. You have to think about it as taking someone by the hand and leading him or her through a complex story. If you don’t do that, you will leave the user looking at something that may look good, but to them is baffling. Often when I’m explaining this to students, I feel like I’m telling them something so obvious it’s embarrassing. But it doesn’t seem to be obvious to them. I think it has to do with art, when art intersects with information, trouble can begin. People are thinking about the images, but they haven’t stopped to think about the structure and the story telling aspects of their graphic. They are too interested in making a mind-bending image that is dominating the proceedings, and the proceedings need to be dominated by telling the story. It sounds so obvious, but it’s frequently lost in the mix. The actual purpose of the graphic is completely lost in this visual attack when it should be about balancing information with art. What I think could help is to write down the stages of what you are trying to explain first. Daniel Velasco, the former graphics director of National Geographic, who now has his own company, has a good system that I can offer as an example. He first writes down a shopping list of the items that he thinks need to be explained and in which order and then he goes into the visual stage from that. Write down the five things that you must get across to the user, to the reader. Essentially have a structure, have a plan. You should not have the user put the information into an order – that’s not their job. You don’t just throw everything at them and expect them to organise it by themselves. It’s so simple, I think, but we overcomplicate it because we want to make art. It’s hard to generalise across all infographics, but those principles, which have been the same for years have not changed at all, and are exactly the same principles of journalism 101.”
The new trend in data visualisation… is storytelling? “How come it’s taken the industry about a decade to figure out that data visualisation is about storytelling? If you’re not telling a story, then what exactly are you doing? Perhaps it’s because data visualisation is in its early stages, as Stephen Few pointed out to me. There has been a lot of criticism of whether data visualisation is of good use to everyone versus just the scientists. There was the cool factor, suddenly people had tons of data, in fact a data tsunami (as Richard Saul Wurman originally coined it) and new 3D software and all went bonkers. And perhaps now we are in the consolidation phase, which is why we are hearing about storytelling so much, because now after the wow phase, people are starting to understand it. I think now the industry is responding to that and making visualisation for the average person, not just what will wow scientists at conventions. I am hopeful that with education it will become more fine-tuned to
the user. The lack of educational support was a huge problem and the discipline had not been taught. Today, with people like Alberto Cairo and others who are teaching data visualisation we can expect to see huge improvement. We have every reason to be optimistic about a big improvement, and the potential is amazing. Data can be our best friend in the 21st century and by unlocking data we have the potential to change the world and improve people’s lives. Perhaps infographics can save the world.”
Grimwade creates inforgraphic experiences that have to interest him first. “I would start with asking myself what would I want to know? Would I want to know more about this or about that? I have tried to make graphics that told things that people wanted to know in a way that was easily accessible. And although it may sound bad that all the graphics that I have done, I’ve done for myself, if I did not want to read it, who else would? Positive feedback told me when I did it right and I’ve let that guide me. Let me tell you about this one project when I was at Conde Nast Traveller magazine that started at London’s V&A Museum while I was on vacation. The V&A [The Victoria and Albert Museum, the world’s largest museum of decorative arts and design] is completely impenetrable. They gave me a leaflet at the desk, as they typically do, and as it is an old building [founded in 1857], it has all different levels, different extensions; it was a nightmare. When I got back to Conde Nast, I told the editor traveller my experience and that many of our readers travel to London – it’s the most visited place for our readers. Wouldn’t it be great to do something for our readers to make it easier to navigate through the V&A? We went ahead and developed a pocket-sized booklet like the ones we’ve done before where we got a fantastic writer to pick a route through the museum, marking the really great museum pieces as a line all the way and out the exit. Don’t bother with the rest of it because it’s too tiring anyway. They handed the booklets out at the US embassy and they were so incredibly popular. To me, that’s where information graphics can really score. I made something that would make people’s experiences better. Instead of going to the V&A stumbling around, confused not seeing half of the best things there, you took this little booklet and followed the line and saw the objects numbered and came out the other end having had this incredible experience. I field-tested it on another trip, and that way I was able to make certain there were no brick walls in the way. Not only did I field-test the booklet at the V&A, we also had another person in London field- test it separately and we compared notes on it to make certain every left and right turn was do-able. And when I came out the other end of the museum, it was just a great experience. It took you past all the key items. Done! I had many emails from people raving about this booklet. That was done in 2004, now it could potentially be done digitally inside the V&A. I’m intrigued by a digital version of this.
The Red Line System in Grimwade’s mind. I’ve also been known for this thing, people have called the “red line system” but it’s nothing originated by me. The idea of having a red line along the ground that you follow, in my case, this imaginary red line exist through these complex spaces, which takes your through the best experience. Some people think I got the idea from my visit to Boston where they have the “Freedom Line”. It’s a red line in the sidewalk and you walk on it, and it takes you past all the best places and monuments in Boston. It takes about an hour or so. The red line is physically on the sidewalk. In my mind, there is always a red line magically on the floor, so then I suddenly realised, there’s a place where they actually have a red line on the floor! You could actually have a red line on the V&A on the same route. In other words, my guide becomes a red line on the floor and you follow it for the highlights collection. So you can make it a physical thing or a virtual red line. Those types of infographics challenges are really interesting to me, and they make people’s lives easier.
Feedback on your graphic is important. “What’s the point if your work does not make sense to the person in the receiving end? Show it to other people and see their reaction; that tells you a lot. In an editorial environment you get a lot of people look at it and you get a lot of feedback. I’ve even had comments from a cleaning lady. She looked at a graphic I had left laying out on the desk one evening, and she put a post it note on it and the most incredible thing, she said, “Check this corner, I don’t think it’s right.” It turned out she had a degree in chemical engineering and was in the process of getting her degree certified in the U.S. And she was absolutely right, it was wrong! When you’re doing infographics all opinions should be welcome. You cannot put a wall around your work and be super protective. It’s all about user experience and being correct. Being wrong is painful but it’s good to know. It’s better to hit me with it upfront and “tell me now”, and not after my work has been printed in a million copies. Several of the graphics I’ve done, I’ve redone them again because I realised from feedback, sometimes from a student, that it can be better designed in another way, it can be improved. And I learn from these re-iterations to not make those mistakes again. It’s important to learn.”
Grimwade on inspiring the new generation of infographics students. “Be enthusiastic, and love information. Have that passion for explaining things. All the tools and methods are secondary to that. Remember what you are trying to do – to turn on the light. Wherever possible inject some warmth and fun. Remember there is a user who needs to be drawn into this process. Nigel Holmes speaks about that and I do like to see some warmth. I use “ice breakers” to break up information to make it more inviting. Without that, data can be intimidating and cold. Always keep in mind your audience. And when I mean keep it simple, I don’t mean to be oversimplified. I don’t want it to be just simple, but clear. Clarity is what you have to be looking for. You have to look to clarify things to people and they will be grateful. And never set people up for something you can’t deliver. It’s an exciting field to be getting into and could become a mega career.”
Interview with Nigel Holmes
Emotion in explanation graphics. “Emotion is memorable. I use a sense of humour, or friendliness as an approach in my explanations graphics work, especially when I try to get some response from the user at the moment of seeing my graphic, and then I have a millisecond to deliver the information. My father was a serious man, but back in England, I was influenced by the likes of Monty Python, The Goon Show, [a radio comedy programme], and the silly lyrics of Edward Lear and Punch, [a British weekly magazine of humour and satire] where I actually sent a submission once. Some things are serious though. I’m not going to make fun of the Ebola epidemic, or Aids or gun deaths, but I can still approach it in this friendly way that draws people in. It’s an emotion but I have to be careful, as soon as you say the word “emotion” in this business, you’re in deep water because this is supposed to be cool and outside (indeed in journalism too) and I got a lot of criticism from people who said I was putting my opinion into my work.”
Using friendliness as an attitude is not chartjunk, Tufte. “I don’t think all humour is international. A sense of good humour is, and a kind of friendliness in approach helps the audience relax. If you can get them into that state, then they’ll read on. My first measure of its effectiveness is that at Time Magazine I used to get a ton of fan mail. They would say, “Now I see” and by “see” they meant understand as well as to see with their eyes. And that’s the nicest compliment that I can get. Immediately that excludes any of what Edward Tufte has called “chartjunk” [a term coined by Edward Tufte which refers to all visual elements in
charts and graphs that distract the user from the main information]. In fact it says you did it, now I get it. Tufte’s first book came out just before my first book in 1982, but in his second book he was just short of libellous about my work. And he reproduced some of my work from Time Magazine, did not credit it and made some unsavoury comments about my piece. Twenty years later, a team at the Ben-Gurion University did a study in which they took one of my pieces from Time Magazine, not one I was particularly proud of or would ever do nowadays, and analysed it next to a graphic that used Tufte’s principles, two simple very straightforward pie charts. They pitted my piece against the plain one (the Tufte-like one), got people to look at it and come back some weeks later. They asked about their preferences for the pieces – what do you remember, how accurate was it, did you like it, did the piece represent the facts correctly? In every single item, my piece was ahead. I did not care at that stage, but I thought it was interesting to see those results. I agree though with a lot of what Tufte has said, but he will not speak to me, I have tried many times. We have jointly been invited to talks in the past, but he has cancelled at the point that he finds out that I will be there. I’ve given up, but I think it would be very interesting to have a talk and see our different points of view. Towards my end at Time Magazine, and I worked there 16 years, I began to think perhaps some of the things I did were over the top – too illustrated and which probably got in the way of the information, and so I’ve now calmed down quite a bit in my own work. People still say to me, “So where’s the fun?” especially magazines.”
Storytelling is not new, but it is essential in data visualisation. “Quite a bit of data visualisation now is beautiful, undoubtedly beautiful to look at, but coming from a journalistic background in which you’re trying to tell people a story doesn’t do very much. My approach is to call it information graphics, which means that it is making information graphical, not making data visualised. Data visualised is just still data. It doesn’t tell you anything until you get into it and work it out. And the key (missing) bit is the editor. Because a lot of academic work is not being edited today, you’re just doing it with software and printing it out (the rude word is “data dumping”), and there seems to be a lot of that now. I find it interesting that many who develop software for data visualisers like Tableau now are saying the new thing is “storytelling” and I answer, “You just realised that?”
Your style might change, but the facts don’t. It comes back to audience. Is there an academic audience, or is it commercial art, to be crude, that’s what I do, and because there are different audiences, you do things differently. Now, I always content that all the graphics I have done were always correct. Nothing could have left Time Magazine if it had not been thoroughly checked. I’ve also done graphics for the New Yorker or National Geographic for example, which take six months to check the facts. So all the facts are there, it’s just the way you present them that is different.”
Holmes’s fine tuned process and discipline. “My process is almost always the same, which certainly is about drawing. But it starts with writing. My sketchbooks are filled more with writing than with drawings. With a complex process, I believe the best explanations cannot be done visually alone. They form a wonderful kind of marriage of words and pictures (and a good marriage at that). The way to understand complex things is to start with writing. For example, one of the most complex pieces I’ve done was for Northwestern University to explain the workings of an advanced proton scanner. So I do research to try to understand what I have to explain. You then have to break down the process into a sequence of steps (top 9) and edit it to “compress” and “simplify” the text. And by these terms, I mean taking out what’s not necessary. You do need some detail so at one point, I look at the graphic and see if there’s enough, but not too much. You also need to know your audience well (will they be alumni or scientists?) with an objective get them to look at your graphic (if it is set to be a lead page in a magazine for example), then to stop at the counter and buy it. Think also
about the shape of where your graphic will be produced (a magazine, the dimensions, head page, colour, typefaces, etc). Then, sketch out the steps by hand and pass these on to a computer program equivalent to Illustrator. Get it to a point in which it’s still in draft-form, but include your full text (not dummy text), and send it to the editor for feedback. So communicate quickly and economically. Ask yourself, who am I doing this for, and what does what you communicate mean? Don’t just do graphics for art.”
Inspiration for Holmes comes from many places. “My process was inspired by Austrian Otto Neurath from the 1920s, a social scientist who was trying to explain to the (everyday) people of Vienna what the housing situation was like there. He put together a team people, which included himself as a thought leader, Marie Reidemeister who became his wife, and a brilliant artist Gerd Arntz. These three invented a way of explaining and charting called “ISOTYPES”. Many thought Neurath was trying to invent a visual language, but he said he was trying to invent a “helping language” instead, one where the pictures were helping the words, but where the words were more important. I have always believed that the way to explain things is not just to get people to look at pretty pictures, but for them to read something as well because both go together. I also get inspiration from the electrical engineer, Harry Beck, who produced what has been the basis for the current map of the London Underground; the English photographer Eadweard Muybridge, father of motion picture; and Thelonious Monk, a pianist whose work I loved from an early age. At age 11, his was the first album that I bought. He was not afraid of making mistakes with his big hands and was considered quirky and an eccentric. And finally another big influence was Eric Gill, a controversial figure given his personal life, but known for his work as a typographer, wood engraver, a writer and the inventor of the Gill Sans typography.”
—————————————
Stephen Few
With over 30 years of experience in business intelligence and information design, Stephen has become a leading thinker in the areas of data visualization for sense-making and communication. Stephen is author of several books, including Now You See It and the quarterly Visual Business Intelligence Newsletter from Perceptual Edge as founder and principal since 2003. He also speaks and teaches internationally, and provides consulting services to clients like Cisco, Apple, Microsoft, Intel, Facebook, Cognos, eBay, Hewlett- Packard, Swiss Statistical Society, Stanford Research Institute, NASA, the US Army, the US Navy, Netherlands Statistics Bureau, Government of South Australia, UNESCO, Time Warner Cable, British Telecomm, Fidelity Investments, World Health Organization, Boston Scientific and many others.
www.perceptualedge.com
John Grimwade
The very personable and approachable John has brought the world closer to our grasp as Director of Infographics at Condé Nast Traveller magazine and also works from his own information graphics business. He has produced infographics at over 30 influential magazines around the world. Before moving to the United States, he worked in major newspapers in London, including The Times. In his free time and for the past 20 years, he’s been a cornerstone figure at the infographics workshop “Show, Don’t Tell” in Pamplona, Spain and has been a great force in establishing the prestige of its annual Malofiej event. He is also professor at the School of Visual Arts in Manhattan.
www.johngrimwade.com
Nigel Holmes
Known for his sparkle and sense of humour perceived throughout his work, Nigel sees the world through a friendly lens, and you can always expect to see him in his characteristic blue gear. Nigel, a well-known British graphic designer, lecturer and leading theorist, focuses on “explanation graphics” and is an author of several books, his last one being the awarded The Book of Everything. After many years in Time Magazine, he now owns an information graphics company, and has worked with clients as diverse as Apple, Fortune, Nike, The Smithsonian Institution, Sony, United Healthcare, US Airways and Visa, The New York Observer, The New York Times among others.
www.nigelholmes.com