What Is Data Storytelling? A Q&A with Datadoodle’s Ted Cuzzillo

That’s a really good question, and probably the main one for this trend as it matures. I know what people seem to assume: a parade of visualizations, perhaps with narration. Data first, story if there’s time.

I say a data story is primarily a story, one that has data in its spine. If it’s not a story first, you might as well just call it “data.”

The main question, then, is what’s a story? Of all the definitions I’ve seen, I like the one by Kendall Haven the most. In his book “Story Proof: The Science Behind the Startling Power of Story,” he says it’s a “detailed, character-based narration of a character’s struggles to overcome obstacles and reach an important goal.”

I think you can’t be too strict. For example, I think it’s OK if there’s no explicit protagonist. One can be inferred. The audience always fills in.

The New York Times often has simple data stories in its Upshot column. A recent one asked how the U.S. economy is doing. The author asked that question, worked through the data, and arrived at answers. It’s a simple data story.

A data story dramatically broadens the understanding of any data analysis. People can relate to it. They understand and discuss it. Good data stories might even be retold.

Data industry surveys estimate that only about 20 or 25 percent of business users use data. Why? There are different theories. The main ones seem to be that many people aren’t interested enough and doubt their ability to understand it. So they just don’t try. They may also doubt its validity.

Some just don’t have time. Several upper-level executives have told me they prefer a brief summary of the data if they see data at all.

But everyone does seem to have time to tell each other stories or to hear good stories. It’s everyone’s native tongue. Data storytelling is easy to learn, and the need is clear and urgent. So why not use it to give data traction and let the insights it bestows infuse decision-making?

The good data storytellers are like good storytellers: They don’t make a show of themselves. They don’t stand up and say, “Everyone quiet down because I’m going to tell you a data story!” Groan. No, they just tell the story.

Some of the best are delivered in words, like one I came across a few days ago. My external hard drive had failed. It was about the fourth one in only five years. So I wanted to know if there were any brands that were more durable than others. I went looking for reviews or, better, actual road tests. The cloud backup service Backblaze has a data story that describes the guiding question—which disk drives endure the best?—and walks through the process and the data.

Also, take a look at The Upshot and FiveThirtyEight. The presidential election is spawning tons of data stories.

Visualization is useful, but who says it’s crucial? This a big misconception. But let’s talk about tools anyway.

I’ve always liked Tableau. I also like Qlik Sense. Both have people who have research going on about storytelling, and I hope to see story-friendly features continue to evolve in both products. Other story-friendly vendors are Yellowfin and Tibco Spotfire.

Many tools work. Hans Rosling, famous for his pre-Tableau visualizations, now available on YouTube, actually told a data story on parking lot pavement with rocks.

Plain old pen and paper will do it. Words work well, as demonstrated in news media. If you sing opera and your audience can take it, try that.

It became recognized. Tom Davenport, the co-founder of the International Institute of Analytics, might have given the spark when he said it was a crucial skill for data scientists. Data storytelling has been around for a long time, and storytelling has been part of business even longer.

Study storytelling and study data. Watch for data in stories to see how it’s portrayed. Above all, though, study your own reactions. You know yourself best: What do you like or dislike? Observe your own reactions to decide what works and what doesn’t.

There’s nothing new about storytelling in business. In fact, storytelling in business is as old as business itself. So is data. In fact, data and stories have worked together for a long time. It’s just become recognized, that’s all.

Data storytelling is at risk of dying just as so many other trends have. People try out the bright shiny new thing, and they soon decide that it’s too limited.

In this case, they could be right. If this is really just visualization on parade, what’s the point? Let’s just drop “story” and call it viz.

I suppose I’m one of the leaders at the moment. I write blog posts and articles, and I teach data storytelling at TDWI conferences and wherever else I can. I also have a short book coming out in the next six months or so on data story genres.

But I’m just one voice, and data storytelling needs a full choir. The Tapestry Conference is good, and various individuals in the data industry lead. Bree Baich at SAS, Robert Kosara at Tableau, and James Richardson at Qlik are all good voices in this thing.

About Ted:

Ted Cuzzillo believes that data storytelling — or whatever it’s called eventually — will lead the next great trend in business data. He helps organizations use this new tool to make data better understood and used. He is a data-industry journalist and researcher. He writes for TDWI, Information Management, and for his weblog, Datadoodle. Have you got a story or thoughts about data storytelling? What’s a better label for “data storytelling”? Email tcuzzillo@gmail.com.

Start Earning Your Master’s Online

Now is the time to advance your education and career through the iSchool.

Explore Our Programs