Most books about data visualization teach you how to build a chart. Which chart type for which data type, which colors, which scales, which labels. The assumption is that the difference between a good chart and a bad one is a design decision.
It isn't. The difference between a good chart and a bad one is almost always made earlier — in the thinking that happens before the chart exists. What question are you actually trying to answer? What level is that question at? Is the data fit for it? What transformations does it need? What does the number actually mean once you have it?
By the time you're choosing a chart type, ninety percent of the work is already done, for better or worse. This book is about that ninety percent.
Ten chapters walk through ten decisions that separate a chart people look at from a chart people actually understand. Each chapter stands alone as a reference, and each decision is illustrated with a working example. The book is written for professionals who have to produce charts as part of their actual job — not designers, not data journalists, but the much larger group of people for whom data visualization is a tool in service of a real decision.
A complete decision path: framing the question, climbing the question ladder, trusting the data, transforming it, analyzing it, interpreting it, and finally choosing how to communicate the answer.
A method for moving up and down levels of abstraction until your question is actually answerable with the data you have.
The handful of operations that reshape data so a question has a clear answer — and why most people skip them.
The discipline of turning "here is a number" into "here is what the number means" before anyone sees a chart.
Most of the bad charts you have seen in your professional life were not bad because someone picked the wrong chart type. They were bad because someone tried to answer a question they had not actually thought through, with data that was not actually fit for the question, in a way that did not actually communicate the answer.
The chart was just the visible part. The damage was done much earlier.
This book is about the part that doesn't show up in the final image — the thinking that has to happen before the chart exists. If you get that part right, the chart almost picks itself.
The Question Comes First is in development. Join the notification list and you'll hear from Kevin when a release date is set.