16 Dec 10

How Data Visualisation Can Make Stuff Better, And Make Better Stuff

by Alex Kirk

For those of a less geeky persuasion than me and who don’t lurk in the more technical corners of the internet (i.e. almost everyone), you may not be aware that the last twelve months have been a banner year for data analysis. Now, before you instantly doze off (and I'll admit, that sounded a lot more exciting in my head than on the page) this is exciting and important stuff - if only because it'll mean that all our meetings can now become a lot more fun.

I'll admit data analysis doesn't exactly have the reputation of being the world's most exciting subject, but 2010 saw the data geeks finally realise that Photoshop, Illustrator and all the fancy design tools that their better-dressed web designer cousins were using could actually be used for - shock, horror - data analysis. Not only this, but that anyone who could combine smart analysis, visually impressive design and (most importantly) a darn good story could quickly become a bit of an internet rock star.

The best example of this sort of work comes from David McCandless, he of the website http://www.informationisbeautiful.net   data star for hire and general face of the cult of data viz. McCandless has probably done the most to prove that information and data can - in fact - be beautiful, and not just dry and dusty charts.

The image I've used here (click here http://www.informationisbeautiful.net/visualizations/the-billion-dollar-gram/ for its full-sized glory is an old favourite from his back catalogue, called "The Billion Dollar Gram", an attempt to put some context around the news media's reporting of the vast billions being pumped into economies during the financial crisis. All boxes are to scale, and it hammers home not only the titanic impact of the financial crisis in a way that words can’t quite get to, but also some other not-very-funny titbits, such as the fact if no-one ever bribed a Russian official, you could pay off the entire debt owed to the West by Africa, and still have more than $100bn kicking around to buy yachts and football clubs with. Compelling stuff, and a lot more impressive than just the numbers alone.

McCandless is just one of hundreds of similar folk (disclaimer - I'm one of them) who've started their own blogs and sites on data visualisation across 2010, and some of the work being produced - sometimes at a hobbyist level - is absolutely stunning. For some easy examples, there's a nice Top Ten of the year's best at http://bit.ly/fREuAp] Flowing Data, or for this year's ne plus ultra (for me anyway) of what can be done with data, you should go and see  http://projects.nytimes.com/census/2010/explorer what the New York Times has done with the US Census Data, which will blow the brains clean from your head with its awesomeness.

It's more than just pretty pictures though - the knock-on effect is that journalism itself is having to course-correct for this shift. Tim Berners-Lee (he invented the internet, so he's pretty clever),  http://www.guardian.co.uk/media/2010/nov/22/data-analysis-tim-berners-lee recently said that journalism's future lies in data, and that journalists had better (data joke coming, sorry) know their http://www.r-project.org/ from their elbow or they'll be completely redundant in a future where the difference between haves and have-nots comes down to who's got both the better information and the skills to analyse it.

This (belatedly) is where the interest for agencies and marketers comes in, and specifically for media agencies. Used as we are to crunching vast reams of hard data, we've never been as good as we could be at the design and the story-telling that should go along with it, stories being the traditional preserve of our creative cousins, leaving us to occasionally sniff about "style" versus "substance". With the statistical expertise inherent in media agencies though - see your friendly neighbourhood econometrician for details - we are like the data geeks at the start of 2010, realising that we've got our own compelling stories to tell, and that adding some style can actually enhance the substance of any numbers-based hypothesis.

So, I'd like to suggest one industry-wide New Year's resolution - let's kick the bad PowerPoint habits, and ditch the old-fashioned notions of how to present data. Let's never again stand up and present hundreds of monochrome line charts, pretending that it's in some way interesting - let's try and put as much effort into designing the data as we do into collecting and analysing it. Doing so will not only enable us to tell better stories, it'll allow our clients to understand our thinking better, it'll give our ideas a more natural canvas on which to sit, and it'll make our conclusions from the data stickier and more memorable.

Best of all, it'll just make meetings more fun.

Some further reading on Data visualisation:
http://www.informationisbeautiful.net/  the mother lode of data visualisation
http://www.ted.com/talks/david_mccandless_the_beauty_of_data_visualization.html Video: David McCandless TED talk  - a must-watch for anyone who uses data
http://flowingdata.com/] - one of the many excellent data visualisation blogs out there
http://ilovecharts.tumblr.com/]  - a much more playful data blog, full of hand-drawn charts and graphics
http://www.flickr.com/photos/walkingsf/sets/72157624209158632/] Eric Fischer’s astonishing city maps - overlaid with tourist vs local photo locations
http://projects.nytimes.com/census/2010/explorer] New York Times US Census Data - Just. Plain. Awesome.

Click http://bit.ly/gpNOpb] here for a handy, bundle of all links contained in this post