Data journalists need to meet designers MORE than half way.
If they want to create great stories.
Data journalism is a relatively new (at least in name) field of reporting. The amount of data created in our lives and work is increasing exponentially, so it makes sense to have journalists dedicated to making sense of it all.
But a journalist + data != a data journalist, or at least not a good one. Great data journalists have another essential quality: a sense of design.
At Visual.ly, we use a lot of data journalists and pair them with data designers to get projects done. Most of our journalists only meet the designer halfway by handing off a text-based outline that the design is then supposed to run up field with. This usually results in good projects, but not great projects.
A great data journalist, on the other hand, can elevate the outcome to greatness by meeting the designer more than half way — even if the designer is mediocre.
The first step to being a great data journalist is to familiarize yourself with the wide array of data viz styles that do NOT come out of Excel. They often have strange names like cloropleth, sankey, radial tree, and parallel coordinates. But don’t let the names trip you up. At some point, all of these only existed in the mind some creative individual.
It’s a rapidly changing field with plenty of unexplored territory. The above visualization doesn’t have an official name, so I’m calling it “The Hairball”.
This chart (above) is one of the holiest relics of data visualization. It was published in 1869, well before any of the modern tools people use today — like Tableau, Processing, R, or Excel — existed.
So don’t let tools hold you back, either. All you need to get started in visualizing data is paper and a pen, and maybe a ruler. The tools will help you deliver something end-to-end — that is, from data to design — but if you are working with a designer, let them do the final stretch.
See no data, design no data
Once you have familiarized yourself with these design techniques you have to learn to see them in the data you are researching. This is almost a Neo-Matrix like ability, but comes with practice.
You, the journalist are the closest person to the data. You have the widest view and can spot the unseen angles and can capture the tone. If all you are presenting to the designer is some bullet points and summarized articles, than that myopic view is all the designer has to work with. A great designer may be able to make some magic with that but great data designers are just as rare as great data journalists.
When you are doing your research and looking at a spreadsheet, try to think of possible visualization methods.
As an exercise, try to avoid using bar and pie charts. These are the tourist traps of data viz and will distract you from the real wonders down the road. Sometimes the data may fit well into a sankey diagram, or perhaps its a tree map, or stream graph. All of these types can convey a ton of information and carry a story, and if you are not looking for them, you will miss the opportunity for telling a great story with data.
I recognize that journalists are not designers and I do not expect them to start pushing pixels or vectors around in Adobe’s creative suite. This is the designer’s job and what they are good at—but meeting the designer more than half way means providing some visual structure that the design can then run with.
That can be something as simple as a sketch in a notebook. If all you are delivering is a spreadsheet or bullet points, the design may miss the opportunities in the data simply because you haven’t presented a wide enough picture.
But even if you are not as visually oriented to come up with a circular-type chart, you can accomplish the same thing with bullet points (and a few arrows).
At Visual.ly, we work hard to maintain and improve our stock of qualified data journalists and we turn out a lot of good projects. But to turn out a great one, I implore data journalists to get out of their comfort zone and take a walk on the design side. We, designers, will owe you a huge high-five (and that is not a data-viz term… yet).
