The first draft of anything is shit. -Ernest Hemingway
When designing a dashboard, you’re never going to get it right the first time. Like writing, or perhaps sculpting is the apt analogy, the real work comes when chiseling away at the morass to create your masterpiece. Here are a few tips I’ve found helpful when editing your dashboard.
Clarity Trumps Brevity
There are most likely clear naming and abbreviation conventions for the majority of metrics you’ll want in your dashboard. But most business are going to have at least a few that don’t have such a clear and concise label. It’s tempting to create a naming convention that is as short as possible — it’s easier to lay out in the dashboard, and easier to refer to in conversation. Except when it’s not.
If your naming conventions come at the expense of clarity, it often causes more confusion than it’s worth. If you have to explain the metric every third time you’re discussing it with someone — it’s not clearly labeled. It’s even worse if you don’t find yourself explaining the metric, but you have people misinterpreting it in silence. A colleague could go weeks or months with incorrect assumptions because they forgot that an “Activated” account actually assumes two or more actions on site, instead of a simple registration.
You Need a Glossary Even When You Don’t
Of course there is no getting around complex metrics with complex definitions. Depending on the size of your team, and the audience for your dashboard, it’s possible that everybody is on the same page with the definitions of each metric.
Even so, I recommend having a glossary available for your metrics. At some point your team will grow, the audience will change, or someone will simply forget how you’re calculating something. You don’t want to clutter the main view with definitions, but making them clearly available will save both you and your colleagues a lot of time if they can refer to the basics quickly and easily.
It’s also important to be clear in your definitions here. For example, there’s a big difference between a rolling 7-day metric and a week-ending metric, and those types of differences should be spelled out in plain language.
Deductive Reasoning Still Requires Reasoning
The purpose of editing is to reduce the unnecessary. Taken to the logical extreme, this can cause people to think that if you can already calculate a metric based off of other metrics in the dashboard, it should be removed. And maybe it should! But that shouldn’t be a hard rule.
The purpose of editing is to remove the unnecessary, but the purpose of a dashboard is to provide valuable business information in an easily digestable format. If you constantly want to know a particular metric, it’s not easily digestible if you have to back into it each time with the other data provided. Especially if you’re looking at sparklines or other trended data. Bottom line is if it’s important, show it. It doesn’t matter if it’s overlapping with other metrics.
Trend Consistency
Not all metrics should be going up and to the right. If you have metrics you’re trying to move in the other direction — time to service resolution for example — it’s often best to group those together. At a glance, it’s difficult to see deviations from the norm if it’s not clear whether a given metric should be moving up or down. This is especially true if you’re looking at a year over year or other time based comparison. (E.g. why is this row highlighted green as a -20% YoY change, but the next row is highlighted in red as -20% YoY change?)
There are many other factors to consider when grouping and laying out your metrics, but this is easy to overlook and can really impact the readability of your dashboard.