I’ve been kicking around this notion of exploratory vs. explanatory data visualization and I’m starting to like it. Despite what most BI tools might have you think, not all data viz is created equal. The charts that we cobble together often have different goals.

On the exploratory side of the spectrum, you are visualizing data in different ways in order to discover new ideas or insights within the data. The goal is to learn something new. With explanatory visualizations, this looks a bit different. You already know your hypothesis is proven to some extent, and you are trying to best communicate that story to an audience. Despite how divergent these modes of visualization sound, it’s surprisingly easy to conflate them.

Picking the right mode

The interesting thing about these different cases is that different details matter to each. When exploring data through visualizations, it’s all about fast feedback loops. “I want to see these views and I want to do it as quickly as possible.” There’s no need for polish. See something interesting and then pick a branch from that to explore further. Drill-down into specific series to better understand them. It’s all about moving fast and letting your creative, analytical mind flow. It’s fun, and ideally a fairly chill experience.

On the other hand, the process of creating an explanatory visualization feels a lot different. It’s no longer about fast feedback loops. It’s about how much context you can effectively pack into your visualization and surrounding elements in order to tell a compelling story. Chart type, series groupings, colors, titles, axes, legends, and presentation format all begin to matter much more. Should there be an accompanying write-up or is the visualization sufficient on it’s own? This isn’t as “quick twitch” or "flowy" as the aforementioned exploration workflow. It’s significantly less chill and overall, less intuitive.

Practically speaking

So what does this mean? Know which mode you’re in so that you can optimize your workflow for the one that is right for your use case. I find that being intentional about the goals of the viz leads to far better results.

If you are in exploratory mode then don’t get bogged down in polish. Be explicit about your hypotheses up front and explore as much surface area as you can. Document findings and revisit them. In explanatory mode, put yourself in the audience’s shoes. It seems so simple, but it can make or break a project. Storytelling matters, so present to others ahead of time. Think about it from other perspectives. Make sure the story is clear, and ideally, leads to some cool outcomes.

Thanks for reading! If you enjoyed this post, you can subscribe to get future ones like it delivered straight to your inbox. If Twitter is more your speed, you can follow me there as well. 🔥