24/2/2012

Always walk before you run: Datawrapper breaks down barriers in visualising and publishing data

 

There are many tools available to turn data into graphics, whether you want to show trends or explain and improve the understanding of complex issues. Not many of these tools are designed to respond to the specific needs of journalists who work on a deadline and are not trained as visual designers. The recently launched data visualisation tool Datawrapper aims to meet these needs.

Datawrapper is built by journalists for journalists. It was created by Mirko Lorenz and Nicolas Kayser-Bril for ABZV, a German training institution for newspaper journalists. It is available in German, English and French. The website received over 3.000 visitors in the first two days alone and newspapers have already started to use the tool.

With the first beta version of this open source tool, users can register for free, copy-paste their data into the platform, create simple, static charts in no time and embed them in their articles. The types of charts available at the moment are: Line, Bar, Pie, Table and Streamgraph. All charts created with the tool link to the data source so that anyone can download and further explore the data used to generate the chart, as you can see below:

When asked about how the idea for this project came up, Lorenz replied: “In early 2011 I started thinking about what we might actually be missing in the data journalism workflow, from finding data to publishing it, even with all the tools that are already out there. The obvious answer was: no external logo, high privacy, flexibility and, above all, ease of use.”

Datawrapper builds on modern HTML5 chart libraries like Highcharts or D3.js. These are currently gaining popularity due to a combination of low bandwidth need and versatile use in all types of media, including print. “There is a barrier. To use HTML5 charts you need to be able to code. Here is where Datawrapper can help. It does what the name implies: It provides an interface to push data into HTML5,” explained Lorenz.

In the following Q&A section we asked Mirko Lorenz (M.L.) how journalists can make the best out of this tool:

Q: What does Datawrapper offer to journalists that other tools, such as Tableau, don't?

M.L.: There are already many great tools and more to come. Datawrapper aims to fill a gap on the lower end. It is best at producing an accurate, embeddable chart in only a few minutes.

While Tableau, RGephi and others are great for technically competent users, most traditional journalists don’t have any knowledge of coding or design. You cannot expect that all journalists will immediately sit down, dive into technical manuals at night, and get themselves at the level of those who have years of technical experience. Some do that, but most are starting to feel like outsiders in their own profession. Using terms like “scraping”, “APIs”, and “D3” is like talking in tongues to most of today's journalists. With Datawrapper we wanted to lower the barriers so as to enable the majority of journalists to turn data into graphics every day, even under the pressure of a deadline, without needing the help of a designer or developer.

Q: What is Datawrapper best at?

M.L.: Datawrapper is a fast and simple tool to use, but at the same time it is limited to static visualisations. The time someone takes to create and publish a basic chart is reduced from hours to minutes. Would we want the whole magic of D3.js in Datawrapper? Sure. But we want to get journalists at that level step by step. In the coming months we will test the tool with German newspapers to understand more about their needs and how to help them integrate it in their workflow. Based on this feedback we will decide which features to provide next.

Q: How do you see data journalism evolving in the next few years?

M.L.: Right now, the pace of development is phenomenal, but it is a bit scary too. There is something new almost every day in data journalism. Have you seen 'Obama's 2011 budget proposal' visualisation from the New York Times? How about the 'Reading the riots' Twitter analysis from the Guardian?

One effect is a widening gap between the newsrooms who practise data journalism and those who don't. This is not the right course of development. New York Times' editor Aron Pilhofer was recently quoted saying: “... when you’re covering a small school or when you’re covering a local government or whatever, how do you not throw their annual budget into a spreadsheet to make sure that everything adds up?” 

Another aspect that should be taken into account is that visualisation should serve a purpose, not just add to the noise, as some infographics do. Andy Kirk of VisualisingData.com recently wrote that the visualisations he loves are “those that exhibit a very simple design execution but unlock seemingly disproportionately rich stories.”

This is what I believe in as well. For data journalism to flourish we don't only need the right tools and access to the right data, but journalists with a nose for stories, the knowledge to ask the right questions to the data and a lot of practice. After all, you must learn to walk before you can run.

 

 

Comments