Design Choices

Back in the 90s, I took a photography class. I loved it. Exposure, f-stops, focal length, mixing chemicals, dodge and burn, zone system, doing things in the dark. Which is weird, considering that photography is all about the light. With all the science and math involved in great photography, guess what the best tip to a good picture was? Frame your shot, then take one step closer. It simplifies your shot and keeps what’s important in the picture.

What's this have to do with dashboarding*? It seems we have a parallel path to better communicate the importance of our data. We have an immense amount of information at our fingertips, and the temptation is to show as much as possible.

Is that what we want to show? What questions would this chart answer? What action would your client take based on the chart? This stacked bar chart attempts to show a count of case types over time. There are 76 different case types in this chart. Past the largest one or two, could you identify any of them?

Another problem that we run into is too little information in a chart.

Scratch that. Too little usable information. Above, we see four data points and a plethora of labels. Don’t label your y-axis and put labels on your data points. It’s redundant. When you’re close to $100 million, do you really want to have your precision down to the dollar? Better to display it in millions -- $99.5M vs. $99,537,488. Now, tell me if these sales are good or bad. One more data point, possibly a goal or quota, would tell me the answer.

So, back to the analogy. Our profession is built on science and math. We use them to extract conclusions and recommendations from a lot of data. Take a step back and only present what’s important and actionable in your design.

It’s a simple as that.


* Somebody please help me come up with a different term than dashboarding! It’s not really what we do. We have elements of dashboarding, but also of reporting. We are a hybrid and looking at it as a dashboard limits our designs. Help!