The Three Core Steps to Data Storytelling
Back in 2016, Brent Dykes penned an article for Forbes magazine that introduced a representation of the three critical elements of data storytelling – data, visuals, and narrative.
When I first read Dykes’ article, I punched my hands into the air in excitement. Finally, someone was defining the concept of data storytelling the way I had been!
At the time, I was working with a group of senior marketing leaders on how to mentor for a data-based culture and shared Dykes’ image. We had a rousing conversation about what goes into each element, but the value of the concept was clinched when I pointed out that it illustrates that representing data is, in fact, a sequential process: first data, then narrative, and finally visualization.
This is where so many of us get caught. When leaders complain to me about the quality of insight they see from the data, most often it traces back to the same problem – getting stuck in the cross-hairs of data and visual and assuming that a great visual will carry the story.
It looks like this: pull data and make a chart; pull data and create a chart. Pull and chart, pull and chart, until the data source has been exhausted. Finish the deck and only then figure out what to say.
Unfortunately, that process is backwards. You need to define your narrative before you visualize your data.
That’s because visualization tools are designed to make information more accessible, not actionable. A number is just a number until a human applies judgement. That’s when you have insight.
For example, a photo of a goalie saving a puck is a striking image. But what makes it a compelling story is the selection of relevant data points shared by the storyteller: it was a shoot-out, it was the winning save, it was the tournament final, it was the 14th shot in the shoot-out, and the goalie was 10-years old.
Narrative provides a filter of relevance between the data and visual to make a story. That’s why a story told by a friend that hasn’t first been filtered through narrative relevance will ramble. Can you envision what a data story that hasn’t been filtered through narrative relevance looks like?
A data dump. (Shudder.)
Creating a narrative last is, almost without exception, where I find the source of tension between those who need to action the data and those who discover the data. Resolving the tension means getting the framework right:
What have I learned to answer or address the issue? (Data)
What does my audience need to understand to take action? (Narrative)
How do I need to present the relevant data to instill confidence in the insight and action plan? (Visualization)
Thinking through this sequence will make a world of difference. Next time you’re building your story, stop and consciously think about the process, and make sure you’ve thought about your narrative before you develop your supporting information. I’d love to hear how it goes.