Why I think data visualisation is all about the story

A defence for storytelling being used in the data viz world

I saw a tweet from Alberto Cairo recently which, as you can see, got a fair few people in the data community talking about whether we should be talking about stories when we create data viz. Storytelling is something I talk about a lot when I critique other people’s visualisations and when I think about and talk about designing dashboards, so I thought I’d share my thoughts on the matter.

For me, a story is the organising of information into some kind of structure so that it can easily be communicated to the user/viewer/reader.

That story does not need to be linear — the story can be the pointing out of gaps and presentation of missing data or something that doesn’t quite make sense, something that makes you ask “why?”.

And in this instance, I do not think the story always needs an ending. Just like every Hollywood blockbuster leaves you with enough loose ends to come back for the sequel, I think data visualisation often does the same thing. It can lead you down a path to explore a data set but it does not always need to present some final conclusion or have all the answers. Instead, it can be the basis for further questions and curiosity into the matter.

Alberto Cairo mentions Jon Snow’s Cholera map as an example of data viz that is not a story at all. In one sense, he is right — there is no beginning, middle, or end to this communication of information. But the placing of shapes on a map does tell a story. The map outlines where you are, a legend or title will tell you what you are seeing, and there are some hints to the why and the how when you see the concentration around the water pump. It is these hints that sparked the further investigation into the causes of the Cholera outbreak.

It is the design, combination of shapes, and presentation of information that creates or at least begins the story.

It’s this idea that we can ask better questions by telling stories that I try to implement when working with clients in Business Intelligence settings on a day to day basis. And quite often this involves telling incomplete stories where they can ask things like, is there a way we can improve sales? Or do we need more data to know what’s really going on here? Is there another department that can help us with this?