Make your data Flourish
We interviewed Duncan Clark, one half of the Startups for News winning team and cofounder of Flourish.
If data visualisations are made for the masses, interaction turns them into personal tools. And when interaction is done well, it resembles a one-on-one conversation with an expert who is patient enough to explain everything, wrote Dominikus Baur, a data visualisation and interaction designer.
But creating successful interactive content takes time and talent — it might be inefficient to produce in large newsrooms and too big a task for smaller newsrooms. Without the help of developers, non-coding journalists sometimes struggle to produce anything except for maybe a simple chart.
Duncan Clark and Robin Houston, winners of last year’s Startups for News competition, are looking to change that. They are creating Flourish, a platform that aims to make it easier for journalists to produce and publish high quality interactive visuals and stories based on a number of different templates. Down the line, newsroom developers will also be able to produce their own templates for in-house use or to share with the rest of the visual creating world. The aim: ‘creating powerful, beautiful, easy data visualisation’.
Hang on, aren’t interactives dead?
‘Yes and no’, said Clark. After Gregor Aisch, a former New York Times graphics editor, revealed that only 10 to 15 percent of people click on New York Times interactives, some people, according to Clark, inferred that interactives are no longer a desirable format for newsrooms. For example, Martin Stabe, interactive editor at the Financial Times, wrote in 2016 that ‘the promise of cool, whizzy visuals is seductive, but they can often turn out to be an expensive mistake that does little to improve communication’.
Aisch, however, does not actually think that interactive graphics are dead. In a March 2017 blog post, he clarified his position and gave reasons for creating interactive content even if a small share of the audience will click on it. ‘Knowing that the majority of readers doesn’t click buttons does not mean you shouldn’t use any buttons. Knowing that many many people will ignore your tooltips doesn’t mean you shouldn’t use any tooltips. All it means is that you should not hide important content behind interactions. If some information is crucial, don’t make the user click or hover to see it (unless you really want to). But not everything is crucial and 15 percent of readers isn’t nobody’.
While the Times still produces some interactives, Archie Tse, deputy graphics director at the paper, pointed out that the bar is now very high. According to him, readers prefer scrolling, but if you want them to click on anything then ‘something spectacular has to happen’.
Adapting to a mobile audience
If anything, interactives have to adapt to an increasingly mobile audience to survive.
‘‘When we first started making interactives five or six years ago we could do them at 940 pixels and the rest was straightforward, but those happy days are long gone’, Clark said.
The key points to consider now are how an interactive can adapt to different screen sizes and what level of detail is appropriate for the various formats: mobile users will want a different experience to desktop users.
While the smaller mobile format calls for simpler and perhaps less detailed infographics, Clark points out that this doesn’t mean that in-depth, focused reading is no longer happening on desktop. A lot of news media are therefore doing a combination of light, touch-basic charts for mobile as well as more rich interactives that can be consumed on desktop. Flourish templates take care of responsiveness by themselves: some templates scale things down to fit to mobile, and others offer a simpler user interface or fewer graphical elements. The Financial Times is beginning to introduce more responsive interactivity into their basic charts. In August 2017 the publication used Flourish to visualise their analysis of travel habits of chinese tourists: find the interactive behind a paywall here.
Show us something extraordinary
While there is no golden rule about the type of topic or interactive that will drive engagement, Clark says that if it is really interesting and original it will generate traffic. If it’s a commodity and unremarkable, it will be less likely to — just like any other types of content. While most energy is put into visualising things related to business and climate change due to the sheer amount of data available, Clark says there is often just as much rich data available on less obvious topics. The Pudding, for example, creates pop culture content through a data lens, resulting in a number of powerful interactives and visualisations.
John Burn-Murdoch, data journalist at the Financial Times, said in an interview with Datenjournalismus II Werkstatt that ‘the key to a good data story is the story itself. […] No matter how clever the math, or how innovative the graphic, it’s all worthless if the reader is left thinking “so what?”’
A visualisation should convey ‘one extraordinary fact’, says Clark. And in 2016, Kiln (the data visualisation studio that gave birth to Flourish) did exactly that by creating an interactive map of the movements of all ships in the global merchant fleet during 2012. The one extraordinary fact? The world’s ocean is alive with shipping.
‘[Here] is a huge and rich part of the world that we don’t normally see’, quipped Clark. The map was shared widely precisely because it had an immediate pull. And those who wanted to get into the nitty-gritty were able to explore the map further.
This two-track concentration span is something that Flourish have been exploring more deeply.
‘When you make a story with Flourish, the whole idea is that you can guide your audience through a bunch of different views to explain a complex topic’, Clark said. ‘The idea of making everything available and just guiding people through the key points is something that we have found works well with data visualisations’.
The ship map, for example, contains multiple views — the user can simply sit back, press play, and get a 90-second overview of the data. Clark explained that when there is an immediate story layer, almost everyone who arrives on the page will press play.
‘As long as you keep the introductory experience down to about one minute, most people will make it to the end as they see how short it is. Two thirds of people will then drop out as they feel that they have experienced the story and the graphic fully. The rest might then spend around 15 minutes digging around the data, see what else is interesting, and explore the other features’.
An efficient model for interactives is therefore providing a simple version and making the exploratory version available.
Personalisation: bringing a story to life
According to Clark, the main role of an interactive is to bring a story to life and to make it easier, quicker, and more enjoyable to digest. In the case of a rich data set, the best thing to do is making the entire set explorable so that people can find their own stories. For example, if you publish a dataset that has detailed information about every country in the world, it should be put in a form where everybody has access to the overall narrative of the data. Beyond that, it is useful for the reader to be able to enter their own geographical position so that they can access details about their own county and the information that is relevant for them. Opening the data this way also leads to increased transparency, making interactives a good driver of trust.
Less is more
While giving options is important, it is even more important not to confuse the reader with too much interactivity.
‘Our early efforts at mapping the rise of ISIS in Syria, for example, involved an interactive map which we updated with the latest reports detailing the tactical situation. As the conflict grew more complex, however, the map grew unwieldy. Users had to spend a lot of time interpreting it by clicking on markers. Our later maps about the Syrian conflict sacrificed interactivity, but not data or detail’, wrote Martin Stabe on the Financial Times.
The data-ink ratio introduced by renowned data visulisation expert Edward Tufte serves as a good guideline to get rid of the gratuitous. Data-ink is the ink, or amount of pixels, a visualisation uses to present data. If data-ink is removed, the visualisation loses all content and meaning. Anything apart from that is non-data-ink — the ‘ink’ or ‘chartjunk’ that doesn’t communicate facts but is used for labels or the clickable markers in interactive maps.
The data-ink ratio is therefore the proportion of data-ink (read: the essential ink) that is used to present actual data compared to the total amount of ink used in the entire display. The ratio should be kept at around 1.0. According to Tufte, good graphics should only contain data-ink to stop the reader from being distracted by irrelevant non-data ink elements.
Then why the focus on beautiful data visualisations?
Clark agrees that design and interactive features should be kept simple. He adds that there are some data visualisations that resemble an experiment in graphic design at the expense of communicating data successfully.
Yet a certain amount of beauty remains important.
‘If you do make a visualisation beautiful, people are more likely to engage with it. However serious and important your content it, putting it in a form that is attractive and intuitive and offers a good user experience increases the number of people who would want to engage with it’.
Alberto Cairo, the Knight Chair in Visual Journalism at the University of Miami, calls data visualisation ‘the functional art’. While the words ‘emotion’ and ‘data’ are not put together very often, beautiful visualisations are an effective way of creating compassion.
Cairo called an interactive project by Periscopic on gun deaths in the US in 2013 a ‘triumph of emotion made functional’. The interactive scales the path and length of a person’s life through an orange line that arcs over a life expectancy axis. The orange line turns grey at the year a person has been shot dead, and then continues to show the amount of years that person could have lived. The interactive ends by showing a mesh of orange and grey that represents the thousands of people killed by a gun and the number of years stolen from them.
Cairo said the ‘the figures and the somber, sad beauty of this interactive graphic will chill you to the bone’.
Duncan Clark is co-founder and director of Kiln & Flourish, with Robin Houston. He is consultant editor at the Guardian and has written a book called The Burning Question.
