Simon Chadwick speaks to David McCandless about the art of data visualisation.
David, tell me exactly what it is that you do?
Good question. I appear to be a data journalist and information designer. That means that I take all kinds of data and information – numbers, statistics, ideas – and turn them into images and designs that anyone can understand.
How did you coin the term ‘data journalist’? What are you doing that makes you a data journalist?
It’s a two-fold approach. On one level I work with data: I use data as a medium to create stories. I rove it for patterns or interesting correlations or unexpected connections. On a second level I use the tools that a programmer or a coder might use. I do a lot of scraping: running scripts to pull in data from disparate websites. I do a lot of mining: using tools to examine a word-frequency analysis or look for statistical patterns in data.
You’re a storyteller whose raw material is data.
That sounds about right. Someone defined it the other day as a new kind of photojournalism. We’re all surrounded by this ocean of data that’s swamping us or overwhelming us. Data journalists may be seen as people who go out into this realm and come back with images and stories – new perspectives and new ways of looking at the world – constructed out of data and information.
Looking at your lecture at TED, which I thought was very good, I really liked your Billion Dollar-O-gram. Tell us a little bit about that.
I feel quite affectionate about it, because it’s the first infographic I ever created, back in June 2007. It’s a tree-map, a tessellation of a lot of squares and rectangles fitted together, and it forms a single pattern. That image arose out of my personal frustration with not being able to envisage what billion-amounts were. They’re repeated in the news a lot – 500 billion for this war, 50 billion for this pipeline. It’s hard to comprehend, they’re mind-boggling numbers. So I set about scraping numbers from various news stories and collecting them all in a spreadsheet, and then I visualised them. I scaled a set of boxes according to the amount, and then coloured the boxes according to the motivation behind the money. It enabled me to create a colour pattern rippling through the data. And the cool thing about it is that when you visualise information in this way, you can have a different kind of relationship to quite abstract numbers. You can literally see them. Even cooler, you can start to see patterns and connections and relationships between numbers that you’d otherwise never see together. You can really see the scale of something, like the expense of the Iraq war compared to the amount of money we give to charity every year. You can see the contrast, it’s very striking when you do that.
One of the things I saw in there that made my eyes bulge a bit was the personal charitable giving of Americans compared to the amounts set aside for foreign aid by the world’s biggest economies.
That’s right. The American people are very generous, individually. Combined they give about 300 billion dollars a year to charitable causes. The top seventeen industrialised nations give a combined amount of about 100 billion in comparison. It’s one of those contrasts. I didn’t set that up, it’s just in there, it’s something for you to discover. That’s what excites me about visualisation. It’s non-linear and you’re able to find your own stories.
I think you referred to visualisation as knowledge compression?
I did. This is one of the potentials of data visualisation, and one that I find quite annoying. I wrote a book of infographics, Information is Beautiful, and I had six or eight months to create this book. I did a graphic that took me about a month – a graphic about the evidence for whether nutritional supplements worked or not. It took me about a month to complete and two weeks to design. So after six weeks I finished this graphic, but it only filled two pages of my book. Only two hundred forty-eight pages to go! It was very frustrating, but at the same time very exhilarating, because you could really see that you can condense and compress huge amounts of information into a very small space without overwhelming your viewer. That’s another exciting thing about visualisations: it’s actually enjoyable to look at them, it’s beautiful. If you have beautiful objects made out of meaningful information it becomes enjoyable to learn and understand.
That’s one of the things that really struck me. The title of your book is Information is Beautiful – except, I think, in the US it’s The Visual Miscellaneum?
Pity, that. But there is an excitement to what you do. It is borne out of the beauty of the stories that emerge, and is akin in many ways to what the best researches do: teasing out the story of a brand, or the story of what’s going on in a market. I was struck by this idea of excitement and beauty going together at the same time.
I love that, and everybody loves a story. There’s really something about a tightly-told story beautifully expressed – it unveils a mystery that you haven’t seen before. I find a photographic metaphor quite useful. Take an aerial view of the subject, or a satellite photo of the subject. Or you can zoom in to a little statistic or a little detail about that subject. You can make a story out of a tiny statistic, you can make a story out of the longer view – either can be one of the scenes in your story as you move your camera through the data set that you have.
You’re using the language of the eye – the camera analogy – and then combining that with the language of the mind.
The eye is just so exquisitely tuned to detecting variation in colour and pattern and shape. Those things are the language or the vocabulary of the eye. Combine that with the language of the mind – concepts and ideas and numbers – and overlay them in a single design, a single approach. The idea is you can start to speak two languages simultaneously, the language of the eye and the language of the mind, each enhancing the other. It’s a richer, more exciting, more vivid and vibrant way of engaging with content.
That goes back to the old cliché, I guess: a picture is worth a thousand words. Today, a picture is worth ten thousand data points. How did you get into all of this?
I was a print journalist and a writer. In twenty years I’d written quite a lot and I was looking for a new challenge. I’d gotten into writing and working online a lot, and obviously, on-line, design and information go hand in hand – you can’t really work online without some kind of design sense, and often the impact of what you do is related to how well it is designed. That started to influence me. When I came to research articles, I often found myself scribbling and drawing up diagrams to help me keep track of different ideas and the perspectives. That became more and more of a regular thing, until I had an epiphany and I thought – why do I do all this research and end up writing a huge article? Why not do all the research and, instead of writing the article, make these diagrams I’m drawing into the article itself? That was the moment I realised I could just visualise these ideas.
It seems that what you’re doing is rather out there in front of everybody, that people haven’t really cottoned onto this before. Is that fair?
I think I have done a bit to popularise it. I’ve in no way invented it. Infographics and data visualisation of sorts have been around for a hundred years, but it hasn’t broken through. I think now the time feels right, or ripe even. Because of the internet, I think. Because the internet is a great medium for spreading visual information. The internet is training us all to be visualisers. There’s a visual literacy and an appetite for information that’s beautiful in this high-speed age. It’s just good timing on my part.
The archetypal – and probably not too unfair – image of the market researcher is the dry purveyor of objective truth, a hundred-page power point deck filled with pie charts and bar charts and god knows what. Today more and more people, particularly clients, don’t have time for that. They’re looking for the story, the killer chart that tells it all. A lot of researchers are asking, “Can I learn data visualisation? What’s necessary for me to be able to come into the mindset that you occupy?”
I’ve started doing workshops, working with people and training them. From the outside this approach could be seen as a form of design. But for me, really, it’s a form of journalism. A lot of my images start with research, with concepts, with story. Only very late in the chain do I start to do any proper design. I’m structuring the data, I’m structuring information, I’m trying to find the story, I’m trying to find the concept that will guide me or hold it together. And then I start sketching and drawing schematics or visual ideas. It’s about 20% design and 80% research. I would say to people that refining the journalistic side of this form, using story-telling to draw people’s attention, is a good place to start – and then maybe turn to the design side. I don’t think there are any magic bullets, I don’t think there are any amazing new forms that can suddenly solve all these problems. But I think there’s a hunger for story, there’s a hunger for tight concepts. Nobody has much time, least of all clients. So if you can get that side done, you’ve done at least half of the job.