I’ve been a writer for almost two years now at Kontinentalist, a data-driven editorial studio in Singapore. In that time, I’ve interviewed my fair share of prospective interns, writers, and editors.
Almost always, when asked why they’re interested in the position, our interviewees answer something along the lines of, “to learn how Kontinentalist merges data and storytelling”.
I always feel slightly sheepish when this happens, because it’s not like we run a masterclass on data storytelling. Nor should we, because it’s an evolving form. And even though I’ve written a few such stories, I’ve learned, and continue to learn, about data storytelling on the job itself.
So, this isn’t some decisive guide on how to write data-driven stories. Rather, it’s a candid account of my methods and thought process based on my experience working with the medium.
Steps 1–2: Find the story angle / Get the data
This is where data-driven stories are unique.
For other story formats, the process usually goes like this: You start with a topic or story angle in mind, do further research, conduct interviews (if it’s a journalistic piece), and write the story. But as the name suggests, a data-driven story should be driven by data rather than have data play a supporting role.
That means making sure that your chosen data allows you to extract interesting takeaways and weave these key takeaways into a story. I have been excited by a story idea countless times, only to shelve it because of insufficient data (even after collecting data myself) or because the data were too thin to support a story.
But where do you start? Nowadays, data is everywhere and easily accessible. You’ll quickly drown in data and its endless possibilities if you don’t set some parameters. So I usually brainstorm story angles in two ways.