sponsored by

D ATA V I S UA L I S AT I O N

TM

ONLINE DATA VISUALISATION

Stunning interactive dashboards for your market research data

For more information visit: www.e-tabs.com/dv [email protected] Europe +44 (0)20 8205 4665 US +1 888 823 8227 Asia Pacic +64 9973 5310

INTRODUCTION

Seeing is believing
Most people will remember that one research presentation they had to sit through or the research report they had to read the one with 100-plus charts and a word count to rival a short novel. Its probably more accurate to say that theyll recall the experience, rather than the substance of what they read or heard. Chances are there was too much stu there to make anything stick in the mind. If researchers want to deliver insights that have real business impact, its clear that work needs to be done on the delivery mechanism. This is where data visualisation comes in. Finding new ways to share business-critical data in a manner that captures the attention of senior executives is essential. In this supplement weve gathered together experts to share with you the principles, skills and tools you need to make the most out of your data, while researchers and marketing executives discuss the ways they are putting data visualisation into practice. Thanks to our sponsor Kantar. Brian Tarran, editor

Editor Brian Tarran 020 7566 1864 [email protected] Senior writer Joe Fernandez Advertising Guy Oliver 020 7566 1843 [email protected] Design Peter Drake Production editor Stefan Leszczuk Printer the colourhouse Research Magazine 15 Northburgh Street London EC1V 0JR Published by the MRS.

2012 DATA VISUALISATION 3

BRIEFING

A still from an animation produced by City University professor Jo Wood showing the rst ve million journeys taken using Londons public bicycle hire scheme. The three clusters are (from left to right) Hyde Park, Kings Cross/St Pancras and journeys to and from Waterloo Station to the City

Telling tales
There are many good reasons for researchers to embrace data visualisation. One of them is to make people pay attention to the story you want to tell
4 DATA VISUALISATION 2012

BRIEFING

Words Brian Tarran

ood design can be many dierent things to many dierent people. There are rules of course, dos and donts to follow the kind they teach in college. But ultimately, the question of whether a piece of work is good or not is down to the individual. Its almost entirely subjective. Thats true of data visualisation too. Whether a data visualisation does what its meant to do really depends on whos looking at it and what they take from it. Most would agree on the denition put forward by Vitaly Friedman, founding editor of Smashing Magazine, who in a 2008 article stated that: The main goal of data visualisation is its ability to visualise data, communicating information clearly and eectively. But, as Friedman noted, clearly doesnt have to mean simply, nor does eectively mean immediate. To convey ideas eectively, both aesthetic form and functionality need to go hand in hand, he wrote. Yet designers often tend to discard the balance between design and function, creating gorgeous data visualisations which fail to serve their main purpose to communicate information. Market researchers make similar mistakes, but for them the balance skews the other way. If designers are sometimes guilty of prioritising what something looks like over its ability to convey a story, researchers are often guilty of trying to cram too much information into visualisations to the detriment of the narrative. Most other businesses are very organised and have a thorough understanding of how you present yourself and how you present data, and how important it is, but it struck me that the research industry generally didnt, says Martin Lambie-Nairn, founder of Lambie-Nairn & Company and the man responsible for the original Channel 4 logo and the 1997 rebrand of the BBC.

In July last year Lambie-Nairn was hired by research group TNS to become its rst creative director. He saw the job as a challenge, he said. Essentially, youre asking an organisation and the industry generally to have an appreciation of a particular process that has not been traditional to them. What we had to do was create a [visual] language, show how it worked and change people from the inside.

Drawn to scale
Visualising data is not new. Line charts, pie charts and bar graphs are traditional ways of telling a story, based on numbers, in visual form. But data visualisation tends to imply something more impressive. For instance, network maps of the blogosphere (bit.ly/PzaW), renderings of the social media connections between people in dierent parts of the world (on.fb.me/eW5es9) or David McCandlesss Billion Pound-o-Gram (bit.ly/57FakF) comparing dierent aspects of public expenditure. Each of these has something in common. They are ways of dealing with huge numbers and masses of data points and presenting them in a way that makes it easy for people to comprehend things that would otherwise be dicult to think about without a visual aid. City Universitys Jo Wood is the man behind a popular recent visualisation showing the journeys taken by users of Londons Boris bikes, the public bike hire scheme. Wood is professor of visual analytics, a discipline that has emerged out of the US home security and defence sectors. The reason its arisen in those sectors is that they have the challenge of huge volumes of data and they need to pick out particular patterns whether its terrorist activities or unusual behaviour or whatever, says Wood. Essentially, visual analytics is designed to address a class of problems that says, Ive got a massive

2012 DATA VISUALISATION 5

BRIEFING

Provoking a reaction
For people like Stan Sthanunathan, vice president of marketing strategy and insights at The Coca-Cola Company, impact is everything. What we want information to do is inspire and provoke people. Data visualisation is just one way to achieve this, he says. Sthanunathan thinks researchers should think about storytelling a little more broadly, but there are very good reasons why clients are wanting more from their agencies deliverables. People are getting increasingly tired of busy charts, he says. People dont have time to look at them. Its not only researchers that are guilty of this, he says, but most people get confused between a fact-based presentation and a fact-lled presentation. You have to go through all the due diligence to gure out what the story is, says Sthanunathan. But your story will only require some of the data that you have in your study. When youve chosen the data that you want to use, spend some time thinking about how to bring it to life. Within Coca-Cola, Sthanunathan says the process is to identify two or three killer facts that the insight team wants the senior executive team to remember. Within General Mills, meanwhile, the focus is on coming up with the killer chart, says Jerey Hunter, a former consumer insights director at the rm, now a consultant. This is the chart that makes an argument in a unique, clear and compelling way. Hunter says: I generally think of data visualisation as something we do with very complex or big data. General Mills abhors complexity (and jargon) and data visualisation of complex, big data or statistical outputs would of necessity be reduced to something very essential. Suppliers wont be surprised to hear that time is of the essence. Asked whether hed be prepared to wait for his agency to deliver a piece of data visualisation, Sthanunathan says: Fancy graphics are nice but if its going to take an extra week then I would prefer to just have the data earlier. For big companies like TNS, then, the focus is on where to look or what to look for. David McCandless, a journalist turned information designer, agrees that, Using a visual mechanism makes it easier to spot trends, apprehend patterns and see outliers. I imagine very experienced researchers can see that kind of stu in spreadsheets, he says, but the fact that its easier visually makes it more widely accessible. If youre interested in telling a story, then obviously a visual language has a broader appeal: it draws the eye, is attractive and impactful and cuts through the blizzard of information were often navigating through each day.

Stock Check, by David McCandlesss Information is Beautiful Studio, shows the estimated remaining supplies of non-renewable resources and how many years until they run out. Smartphone owners beware indium-dependent touchscreens seem to have a very limited shelf-life

dataset. Buried in there somewhere is something really important and I need to nd it, Wood says. Traditionally that sort of analysis has been done through data mining and the like. But visual analytics has emerged because there was a realisation that you need much more human input into that process of spotting the unusual or the important. Were dealing with datasets that are much more complex and much larger than they ever used to be. There are so many ways in which you could potentially explore a dataset that you need some kind of guidance We need this visual navigation to tell us
6 DATA VISUALISATION 2012

Were dealing with datasets that are much larger than they ever used to be. There are so many ways you could potentially explore a dataset that you need this visual navigation to tell you where to look
Jo Wood Professor of visual analytics, City University

BRIEFING

industrialising the data visualisation process. We have built a system that works in such a way that it allows people to produce PowerPoint decks in a matter of seconds, says Roeland Nieuwenhuis, a senior consultant who has been working with Martin Lambie-Nairn on what the agency calls Project Everest a name that speaks of the scale of the task facing researchers, whether technical or cultural. The automatic charting outputs are really quite wonderful, says Aziz Cami, creative director for Kantar, TNSs parent company. Theyre bespoke pieces of work which are linked to databases. Of course, the outputs at

the moment are fundamentally PowerPoint-based, but were also looking at more web-based opportunities because we feel thats where the future lies. (For more of Camis thoughts, see his piece on page 8.) Time to play Everybody interviewed for this supplement imagined a world where data visualisations become much more interactive and live with real-time information feeding in to them. How soon we get there, though, is a question of technology. Despite the availability of a host of free and paid for tools (see page 14) these sorts

These images from Stamens Eric Rodenbeck show a minutes worth of trading on the Nasdaq exchange. In the top image, a single trader (in lime green) is responsible for the majority of shares. Below that, the yellow dots in a row show trades in a single stock at a single price

of interactive portals still take a hell of a lot of time and money to put together, says Ian Jarvis, who leads the data visualisation working party at Ipsos Mori. At the moment if we do these, even with the reporting portals weve got, it generally takes a month or so to set them up because youve got massive datasets that have to be formatted and organised properly and run through to make sure everythings perfect. Then youve got to design the portal and make sure the data links in properly. Meanwhile Jarvis questions how much freedom people should get to work with real-time or massive datasets. The danger is if you open up entire datasets to people they can do their own analysis and get it completely wrong. And yet giving people the freedom to play with data is important in helping them get comfortable with it. Eric Rodenbeck, founder of design studio Stamen, says the aim of his work is to entertain by informing and inform by entertaining. Much of what he does is for public consumption Im not super-interested in tools for CEOs, he says. I denitely want people to engage

Theres so much that you can learn by using your eyes. For me, data visualisation is not just about answering the questions you already have, its about nding better ways to ask new questions
Eric Rodenbeck Founder, Stamen
with our work and learn something I just want them to have fun rst. Rodenbecks belief in the fun potential of data visualisations positions him at the opposite end of the scale from the research buyers and suppliers who are looking for more in the way of function. As such, he rejects the idea that viewers need to be able to understand everything a visualisation is trying to tell them right away. The whole point of doing this work is to make something thats visually robust or visually seductive enough to engage people initially, but which gives them more the more time they spend with it. Its not something that reveals itself all at once. I dont want to go on record as being intentionally obscure, but I do think theres a lot of room in this space for work that is richly interactive, not just immediately apparent. Theres so much that you can learn by using your eyes. For me, data visualisation is not just about answering the questions you already have, its about nding better ways to ask new questions.
2012 DATA VISUALISATION 7

COMMENT

Visual communication skills are alien to some in the research industry, but they neednt be. Data visualisation can become part of the research process through smart hiring, skills training and expert partnerships

Embedding creativity
Words Aziz Cami
CNN took top honours in the Kantar-sponsored Information is Beautiful Awards earlier this year with its visualisation (on opposite page) connecting US casualties in Afghanistan and Iraq with the towns they came from back home. In doing so, they also connected back in time with the rst great wartime data visualiser Florence Nightingale. Lamp-carrying activities aside, Nightingales Diagram of the Causes of Mortality (reproduced below) played a crucial role in her campaign to show that many more soldiers died of preventable disease than of their wounds. A keen statistician since childhood, she recognised that while words could build an argument, a single visual could convey the essence of that argument in an instant. That is precisely the opportunity that data visualisation oers the insight and research business today. As data piles up around us in ever-growing quantities visualisation has been hailed as a way to cut
Nightingales diagram shows war deaths by injury (red), disease (light blue) and other causes (grey) from April 1854March 1855 (right) and April 1855March 1856
E JUN
JULY

through the clutter and shine a light on the insights that matter. And of course thats precisely the kind of clarity that time-poor, data-rich clients are craving. But we should be careful not to over-claim for this stu. Data visualisation should not be regarded as an end in itself. Its not just turning data into pretty pictures. The real point to data visualisation the value that it brings to research buyers and suppliers is as an aid to storytelling. Its about seeing the patterns in the data that ush out a story and then help you to start telling that story. Only by doing that can you move data o the spreadsheet and out into the real world of consumer behaviour and preferences. The data visualisation helps. But it doesnt get you there. Its all too easy to fall into the old clich of show rather than tell. What were really talking about is, showing in order to tell it better.

The whole truth

The best analogy, I think, and the one I use most frequently, is with journalism. Its no surprise at all that Kantars data visualisation partner David McCandless was a journalist. And its no surprise either that many great examples of data visualisation come from the publishing and media sectors. Journalists face the same challenge that we do of sifting large amounts of often conicting data to arrive at a truth (or an insight, in our case). But if they dont present that truth within a narrative that hooks the reader and imbues trust in their ndings, they wont nd much of an audience. So we need to think of data visualisation as an aid to creating compelling stories that move our clients to act. Its about arriving at a point of view. And this has huge implications for the way we recruit and train our people. Visual communication is frankly alien to some people in the insights industry. Concerned correctly, of course with the integrity of data, many researchers are suspicious of the introduction of right-brain thinking. They worry that visualisation risks trivialisation. And yet it is perfectly possible to encourage people, through training, to discover creative talents in both visualisation and storytelling that they had no idea they had. At Kantar weve put about 250 people through workshops with McCandless and its been a joy to observe people realise that not only do they get it, but they are also perfectly capable of doing it. Such is the hunger to get involved with data visualisation that a group of our people have set up a global network of champions across all our companies. They are now holding monthly virtual meetings and webinars to share their latest outputs and best practice. However, it seems certain that in terms of hiring policy, the industry will have to swing to the right

AU

TEMBER SEP

ST GU

APRIL 185 5

MA Y

855 RCH 1 MA

SEPT.
OCT.

OCTOB ER

RY UA

8 DATA VISUALISATION 2012

R BE

ER EMB DEC

JANUA RY

FE BR

. OV

DEC.

JAN.

FE B.

NO VE M

COMMENT

CNN and Stamens Home & Away is an interactive web-based visualisation that plots US war deaths in Afghanistan and Iraq and links them to a map of the US showing the towns the soldiers came from bit.ly/TloHRZ

(brain) if it is to capitalise on the potential of data visualisation and other creative tools. It is a rare recruit who exhibits equally left- and right-brained traits. Anyone who can deliver the desired fusion of numbers, words and visuals is going to nd themselves in serious demand.

The Visualisation Trifecta

Horse racing fans will know that the trifecta is the ultimate bet. Correctly pick which horses will nish rst, second and third and you are guaranteed exceptional returns. Thats how I feel about data vis if we can win on three counts, we win big. 1. We want to visually signal change and illustrate the value of global consistency by delivering a beautiful visual story everywhere in the world. The work needs to be worthy of the c-suite. 2. We want to use visuals to reinforce our brand. By creating a signature look and feel to our deliverables we remind our clients of our brand expertise. We also take some of the cues from the visualisations as the backbone of our client communications. 3. We want to inspire our employees and model better visual storytelling. By working with great design partners, we provide our teams with a learning experience and the opportunity to deliver work of which they can be proud. Data visualisation is about more than just illustrating numeric concepts. Its symbolic of the change going on in our industry. Its a manifestation of the growing importance of design and artistry in brand building both for our clients brands and for our own. Eileen Campbell is global CEO of Millward Brown

Part of the process

Overall I think its a moot point whether we will deliver data visualisation outputs through in-house resources (as a couple of our companies do) or pursue partnerships with the agencies that specialise in the eld (as many more of our companies do). But as with any other creative product, you have to understand it to commission it. The great challenge down the line is how data visualisation skills can be embedded throughout the research process. Theres a great deal of software now available the so-called charting engines which generate credible data visualisations that are certainly better than the average PowerPoint eort (see pages 1415). These packages can lend consistency and facilitate more clarity and engagement. And they are helping to take data visualisation out of the creative ghetto. Theres a democratising inuence, much like the early days of desktop publishing. But I suspect most of the bigger research companies will want a look and feel in their outputs that is properly connected to their brand and product portfolios, so will opt for a more bespoke approach that addresses client needs more exactly. And the sheer number of datasets that many companies are now trying to integrate means that visualisation packages are more likely to be custom-built.

This requires investment of course. But the automation of many outputs will by denition reduce the number of man-hours required so ROI should not be too hard to prove. Money, though, may not be the only investment required. The eective application of data visualisation brings with it some big issues around work process and planning. Companies may have to adopt a much more agency-like approach to scheduling and time-budgeting, rather than treating data visualisation as an add-on applied at the end of a project, which I suspect is often the case today. This could bring with it an element of positive disruption. Its only two years since Kantar partnered with McCandless and a year since we co-created the Information is Beautiful Awards. The pace of development in data visualisation has been unrelenting, especially in the area of interactive visualisations. This is clearly the dominant trend, accounting for a surprisingly high number of awards entries. Interactivity opens new vistas of discovery in enabling clients and researchers to nd the stories themselves and will bring with it huge benets in terms of engagement and partnership. Ultimately well enter the age of live, active data visualisation brought to us through always-on feeds. When do I think this will happen? Hard to say, but itll be a lot less time than the 150-year gap between Florence Nightingale and CNNs work on the human cost of war. Aziz Cami is creative director of Kantar and before that co-founded brand strategy, design and innovation agency The Partners
2012 DATA VISUALISATION 9

GALLERY

Michell Zappa won silver for Envisioning emerging technology. Each node is sized according to the importance of the technology it represents. The bigger the outline, the larger the consumer impact. bit.ly/so0hFw

Michele Mauri won gold in the Information is Beautiful (IIB) Awards Infographic category for Cover Mania, which shows the most covered artists of each year, based on the WhoSampled.com database. bit.ly/N9Ab95

10 DATA VISUALISATION 2012

GALLERY

Peter Orntoft won gold in the IIB Data Visualisation category. His work is based on opinion poll data measuring Danish attitudes to the Muslim headscarf in various public occupations. bit.ly/NgG890

Paulo Estriga won bronze for his visual CV. The main image represents his two career stages to date, with each segment colour coded according to the type of experience. bit.ly/O3ukoF

2012 DATA VISUALISATION 11

GALLERY

Dimitre Lima won bronze in the IIB Data Visualisation category for his lunar calendar, charting the phases of the moon for 2012. His 2013 edition is now available. dmtr.org/lunarcalendar

Notabilia was IIBs Interactive Visualisation winner. Its a way of exploring the Articles for Deletion discussions among Wikipedia editors. Straight lines are controversial debates while curly lines are unanimous. notabilia.net

12 DATA VISUALISATION 2012

GALLERY

Interactive bronze medal winner The Evolution of the Web plots the various di erent browsers on a timeline. The colour bands represent the interaction between web technologies and browsers. evolutionofweb.appspot.com

Omid Kashans Timeline of the Universe won IIBs Student Award. Its complex print orders come with instructions but theres no denying its visual appeal. bit.ly/UaQpT6

2012 DATA VISUALISATION 13

SOFTWARE REVIEW

Practical applications
There isnt a one-size- ts-all tool for data visualisation work. DataMarkets Hjalmar Gislason reviews what is available to help researchers nd the best solution for their needs
There is no single correct answer to the question, What is the best tool to visualise data? It depends on the task at hand and what you want to achieve. So heres an attempt to categorise those tasks and point to some of the tools Ive found useful. well-crafted visualisations on top of data that can be imported from all of the most common le formats. Common charting in Tableau is straightforward, while some of the more advanced functionality may be less so. Then again, Tableau enables you to create elaborate interactive data applications that can be published online and work on all common browser types, including tablets and mobile handsets. For the non-programmer who sees data visualisation as an important part of their job, Tableau is probably the tool to use. Tableausoftware.com DataGraph is a little-known tool that deserves a lot more attention. A very dierent beast, DataGraph is a Mac-only application ($90 on the AppStore) originally designed to create proper charts for scientic publications. Nothing weve tested comes close to DataGraph when creating crystal-clear, beautiful charts that are also done right as far as most of the

information visualisation literature is concerned. The workow and interface may take a while to get to grips with, and some of the more advanced functionality may lie hidden even from an avid user, but a wide range of samples, aggressive development and an active user community make DataGraph a really interesting solution for professional charting. visualdatatools.com/DataGraph R is an open-source programming environment for statistical computing and graphics. Its a superpowerful tool that takes some programming skills to get started. However it is becoming a standard tool for any self-respecting data scientist. R does a lot more than graphics within its interpreted and commandline-controlled environment. It enables all sorts of crunching and statistical computing, even with big datasets. In fact, the graphics are a bit of a weak spot for R. Most outputs needs polishing in other software such as Adobe Illustrator to be ready for publication. r-project.org

Simple one-o charts

The most common tool for simple charting is Microsoft Excel, where it is possible to make nearperfect charts of most types if you know what youre doing. But many Excel defaults are sub-optimal: some of the chart types they oer are simply for show and have no practical application. 3D cone shaped bars anyone? And Excel makes no attempt at guiding a novice user to the best chart for what they want to achieve. Here are three alternatives. Tableau is fast becoming the number one tool for many data visualisation professionals. Its client software (Windows only) thats available for $999 and gives you a user-friendly way to create

This image was created using Transversal Lines, an interactive digital art project by Joao Martinho Mourha (jmartinho.net) built using Processing

Videos and custom high-resolution graphics

If you are creating data visualisation videos or highresolution data graphics, Processing is your tool. Processing is an open source integrated development environment (IDE) which uses a simplied version of Java as its programming language and is especially geared towards developing visual applications. Processing is great for rapid development of custom data visualisation applications that can be run directly from the IDE, compiled into standalone applications or published as Java Applets for publishing on the web. Java Applets are less than optimal for web publication, but a complementary open-source project Processing.js has ported Processing to JavaScript using the Canvas element for rendering the visuals (Canvas is a way to render and control bitmap rendering in modern web browsers using JavaScript). This is a far superior way to take Processing work online. The area where we have found that Processing really

$50 million

Tale of 100 Entrepreneurs, created in Tableau, shows in ation adjusted revenue growth for 100 of the largest public software companies

14 DATA VISUALISATION 2012

SOFTWARE REVIEW

shines as a data visualisation tool is in creating videos. It comes with a video class called MovieMaker that allows you to compose videos frame-by-frame. Each frame may well require some serious crunching and take a long time to calculate before it is appended to a growing video le, but the results can be stunning. Processing.org

Special requirements and custom visualisations

If you want full control of the look, feel and interactivity of your charts, or if you want to create a custom data visualisation for the web from scratch, the out-of-the box libraries mentioned previously will not suce. In fact youll be surprised how soon you run into limitations that force you to compromise on design. So if you want to take it up a notch and follow the lead of some of the wonderful and engaging data journalism happening at the likes of the New York Times and The Guardian, the tool for you will probably be one of the following:

Charts for the web

There are dozens if not hundreds of programming libraries that allow you to add charts to your websites. Most of them are rubbish. We believe we have tested most of the libraries out there, and there are only two we feel comfortable recommending: Highcharts is a JavaScript charting library which renders vector-based, interactive charts in SVG (or VML for older versions of Internet Explorer). It is free for non-commercial use and commercial licenses start at $80. It is a exible and well-designed library that includes all the most common chart types with plenty of customisation and interactivity options. Interestingly, even though Highcharts is a commercial product the source code is available to developers who want to make their own modications or additions. With plenty of examples, good documentation and active user forums, Highcharts is a great choice for most development projects that need charting. highcharts.com

A Protovis streamgraph, described as a generalisation of stacked area graphs where the baseline is free. See more examples at bit.ly/czvYSB

rendering, making it an option for older versions of Internet Explorer that still account for a signicant proportion of trac on the web. mbostock.github.com/protovis D3.js is in many ways the successor to Protovis, building on many of the same concepts. The main dierence is that instead of having an intermediate representation that separates the rendering of the SVG (or HTML) from the programming interface, D3 binds the data directly to the DOM representation. If you dont understand what that means dont worry, you dont have to. But it has a couple of consequences that may or may not make D3 more attractive to your needs. The rst one is that it makes rendering faster therefore animations and smooth transitions become more feasible. The second is that it will only work on browsers that support SVG, so you will be leaving behind users of Internet Explorer versions 7 and 8. d3js.org

This polar clock was created using Raphal. The lines (from outside in) represent seconds, minutes, hours, days and months

A wind rose for the South Shore Met Station in Oregon, created using Highcharts

Raphal, gRaphals big brother. It is a powerful JavaScript library which works with vector graphics. It renders SVG graphics for modern browsers and falls back to VML for Internet Explorer 6, 7 and 8. It comes with a range of good-looking samples and decent documentation. Raphal is open source, and any developer should be able to hit the ground running with it. We dont recommend Raphal for the advanced charting part, but for entirely custom data visualisations or small data apps it may very well be the right tool for the task. raphaeljs.com Protovis is an open source JavaScript visualisation toolkit. Rather than simply controlling at a low level the lines and areas that are to be drawn, Protovis allows the developer to specify how data should be encoded in marks such as bars, dots and lines to represent it. This approach allows inheritance and scales that enable a developer to construct custom charts types and layouts which can easily take in new data without the need to write any additional code. Protovis natively uses SVG to render graphics, but a couple of eorts have been made to enable VML
A chord diagram, produced in D3. These show directed relationships among a group of entities

gRaphal is another JavaScript charting library built on top of Raphal (see next column). Like HighCharts, gRaphal renders SVG graphics on modern browsers, falling back to VML for IE versions before 9. Its not a very mature library with limited capabilities, few chart types, even fewer examples and pretty much non-existent documentation. It is, however, available under proper open source licences and could serve as a basis for great things. g.raphaeljs.com

After thorough research of the available options, we chose Protovis as the base for building out DataMarkets visualisation capabilities with an eye on D3 as our future solution when modern browsers nally saturate the market. We see that happening about two years from now. Hjalmar Gislason is the founder of DataMarket.com
2012 DATA VISUALISATION 15

THE CLIENT VIEW

Making eye contact with customers

Internal data-sharing is one thing but brands also have to communicate information to consumers. Betfair, Ford and Absolut discuss how they use data visualisation to engage the public
Words Joe Fernandez

Alasdair Wright, global head of customer marketing and loyalty, Betfair

Data visualisations are a great way of driving marketing eorts, which enable customers to get the best out of their relationship with Betfair. In the bookmakers environment we are able to use complex event processing (CEP) technology to monitor and react to customer behaviour about major events on our website and on social media sites. As our odds change, we use graphs to show the peaks and troughs of the day and explain all the signicant happenings that are changing the value of a bet. We are able to visualise this by using the CEP to analyse events (xtures information and user events such as bet placement and funds withdrawal) to detect patterns and extract data from them. We can then translate this into an illustration which is clear and simple. It works for us because it highlights our dierences from our competitors, particularly in the value proposition. These visualisations directly play into our USP namely that Betfair is one of the few companies to allow gamblers to bet at odds set by other users rather than by the bookmaker. This is about bringing in the right customer who will be valuable for the long term. Were not looking for one-o punters, but if we get people in at the start of the tournament they are more likely to stay with us longer. They enjoy these real-time visualisations, because they are empowering them to be in control and understand when is the right time to have a utter. It also enables us to spot any detail that would otherwise go missing such as the hottest parts of the website including banners and oers with high degrees of traction. The ability to create, identify and curate content that is going to be interesting and carry across multiple channels is quite useful. As the internet and webbased content continue to grow at such a rapid pace, numeracy and the ability for a marketer to deal with huge amounts of data are vital. The ability to sort

16 DATA VISUALISATION 2012

THE CLIENT VIEW

through all that and nd the one thing that is going to make the dierence, the one small change you make that could have a massive eect on the bottom line, is crucial.

Raj Nair, group vice president, global product development, Ford

A great example of our use of data visualisation is our latest prototype dashboard concept developed with the Smart Design agency. The SmartGauge dash gives drivers a wealth of information in an enticing format without confusing or worse distracting them. We take the wealth of information coming from the vehicle and translate it for the user so that their drive becomes seamless. Complex graphs or fuel usage timelines were ignored in favour of iconographic alternatives: fuel gauges, for instance, are represented by an image of a tank lled with an oil-like yellow liquid. User research discovered how some drivers get obsessed with achieving a high score while driving that is the lowest per-mile fuel usage. So fuel eciency is represented by a rendering of curling vines blooming with green leaves. Its more than a decorative element, though. Its a data visualisation tool intended to change the way people drive. If a driver wastes gas by aggressively accelerating or slamming on the brakes, for example, the vine withers and leaves disappear. More leaves appear if individuals drive more economically. What we need to remember is that data in its purest form is dull and not altogether useful to the everyday person. By visualising things in a dierent way, we are helping drivers get the most out of their vehicles and we are simultaneously learning new behavioural trends that will impact the next product development cycle. Its win-win when its done neatly and is supported by the best data scientists.

Absolut Vodkas bottle designs are interpretations of the data the company gathers about its customer segments. Our challenge to designers is to take this pile of facts and nd a way to reinvent our look and feel, says global marketing director Mark Hamilton

Mark Hamilton, global marketing director, The Absolut Company

There is something very compelling about data visualisation as a vehicle for helping take your brand beyond being just one of a number of options in a competitive marketplace. Done right, it oers stakeholders an instant visual representation of your brand ethos. Our approach has seen us challenge artists to showcase a dierent side to our brand. For me, this is the most important role that data visualisation must address. When we commission such projects, the brief is always to identify something that will make a consumer stop and consider what Absolut is all about and why it might be relevant to them. It doesnt necessarily have to shout numbers or brand values, but it does have to have a distinguishing mark that makes people appreciate why they pay above average prices for our product. We are constantly gathering intelligence from our consumers and suppliers to identify the dierent things our brand stands for within dierent customer segments. Our challenge to designers or digital

Fords SmartGauge, by Smart Design, makes greater use of icons than the typical dashboard display. Fuel e ciency, for instance, is represented as green leaves on a vine which wither when per-mile fuel use increases

enthusiasts is to take this pile of facts and nd a way to reinvent Absoluts look and feel to strive to ensure that we continue to establish ourselves as a global brand icon. This is what all our stakeholders expect from us and oering them this instant visual read is what they want to see. All the words in the world wouldnt have the same eect as an image that can be observed and understood. By taking account of data and giving the artists our bottles as a blank canvas, we are inspiring people throughout the world to collaborate and ll them with creativity. We brought together creative collaborators from a variety of disciplines and watched the journey from pure white canvas to exceptional pieces of art. What we get is not only concepts that have attention-grabbing cut-out imagery, colourful and playful graphic designs and bold illustrations imprinted on the bottle, but a visualisation that reects what Absolut wants to be known as to our target audience. Its much more interesting than sales data or company reports would ever be.
2012 DATA VISUALISATION 17

TOP TIPS

Research still rules Data visualisation is 80% research, 20% design, according to Information is Beautifuls David McCandless. If you skip on the early research stage and you move onto the design prematurely, youll be ailing around trying to make something work, to make the story surface. You will lose so much time doing that.

1.

Whats the story? The most eective visualisations know what story they want to tell, and let the story dictate the visuals. Clever new visualisation styles might look nice, says DataMarkets Hjalmar Gislason, but Im going to bet that 90% of the time one of the existing, well-known ones like pie charts or bar graphs will do a better job of conveying the story.

2.

Know your audience Data visualisation means dierent things to dierent people. Like Maria Popova, editor of Brain Pickings. In a Businessweek article she wrote: Data visualisation has nothing to do with pie charts and bar graphs Its a dierent way to look at and think about data. So make sure your deliverables are aligned with expectations.

3.

Less is more As Coca-Colas Stan Sthanunathan notes on page 6, theres a world of dierence between a fact-based presentation and a fact-lled one. Resist the temptation to cram as many elements as you can into your visualisation. If it doesnt add to the story, leave it out.

4.

Top tips
Our expert contributors oer rules to live by for researchers wanting to brush up on their data design skills

That applies to text too Some people forget that in a data visualisation, the text itself is a graphical element, says McCandless. That means it should be pared down to be as short as possible and optimised so that every word contributes to the overall meaning of the piece.

5.

Keep in line If you look at any well designed piece of work, its always built on a good layout, says Ipsos Moris Ian Jarvis. It makes such a dierence to have things nicely lined up and well spaced out. Design grids and guides are scaolding there to help build something beautiful rather than constrain creativity.

6.

Be consistent Creativity is nothing without discipline, says TNS creative director Martin Lambie-Nairn. Pick your colour palettes and your fonts and stick to them. Font size and style and colour clashes can destroy any good design youve got, warns Jarvis.

7.

You cant do everything Its rare to nd one person with the design, data and technology skills needed to produce stunning data visualisations. But each of these skillsets are needed, says City Universitys Jo Wood. The focus, then, should be on building strong teams in each area and getting them to collaborate.

8.

18 DATA VISUALISATION 2012

Strengthen Strengthen your integrity your integrity Gain business Gain business advantage advantage
MRS Company Partner Service With MRS support and accreditation we can work together to make evidence matter MRS Company Partner Service Contact on [email protected] With MRS support andus accreditation we can work or 0207 566 1877 together to make evidence matter Contact us on [email protected] or 0207 566 1877

CPS ad Archana.indd 1

24/10/2012 10:30

CPS ad Archana.indd 1

24/10/2012 10:30

Inspiration for a Beautiful Partnership