What can't you visualize? (Hint: nothing)

Some things seem obvious candidates for data visualization—how many different flavors of ice cream exist, for example. But what about things without numbers? How do I quantify just how much better French vanilla ice cream is than all of the other flavors? You can't make a chart of that, can you?

Maybe not directly. But when you're faced with a question that isn't immediately quantifiable—and thus not readily made into a killer infographic—the trick is to change your thinking, and look for a quantitative metric that aligns with the qualitative metric you're actually interested in.

Okay, so you can't visualize how much better French vanilla is than all the other flavors. I mean, it's so much better that it's, like, not even a comparison, right?

Image courtesy stu_spivack, 2007, via Wikimedia commons.
Image courtesy stu_spivack, 2007, via Wikimedia commons.

But maybe you can find data that would indirectly shed light on this issue. What about finding and visualizing how many gallons of different flavors of ice cream are produced each year? Surely that reflects sales, with companies producing more of flavors that sell. Or maybe you could find sales data itself—how many gallons of each flavor of ice cream are actually sold each year? There's a metric that's almost certainly related to popularity.

Even if you can't find data out in the world, you might be able to create your own, quickly and easily. You could conduct a short survey among employees at your company, or perhaps even among your clients, asking what their favorite flavor is, and visualize that data.

So, when faced with something that seems uncountable and unquantifiable, rethink your question in terms of something that you can count, and you've got the seeds of a solid approach that can easily lend itself to visualization.

Unless of course we're talking about the superiority of French vanilla ice cream, in which case, we already know the answer, don't we?

For more tips on how to measure anything, see the aptly titled How to Measure Anything by Douglas W. Hubbard.