A few Rules of Thumb to Make You Dangerous
Chances are if you’re reading this is you’re a product manager or in some way a contributor to a product team and would like to give yourself a leg up when it comes to understanding the data that is coming your way.
I’m going to give you a few basic rules that are going to allow you to get up and running making visualizations in no time, but first things first.. a data types primer
The visualizations that you choose are going to depend on the types of variables you have access to. What I want to focus on here are numeric, categorical, & time data.
Variables represented on a continuous scale. This could be dollars, clicks, calls, and so forth
Variables represented by a discreet list of values. This could be geographical region, gender, income range, business type, industry and so forth
Time is exactly what it sounds like. Could be the moment a click occurred, a deal was closed, and so..
Get Your Hands Dirty with Numeric Data
single variable – numeric:
Box & Whisker Plot: Great for assessing variable distribution
Histogram: Also great for assessing distribution
Two variables – numeric & numeric:
Scatter plot: Simple way to see how two numerics move together.
Single variable – Categorical:
Bar Chart: Here we group by our categorical & take an aggregation– in this case… counts.
Two Variables – Categorical & Categorical:
Heatmaps: Great way to identify how two categorical variables work together– a two dimensional table can sometimes be difficult to consume or take in all at once.
Bar chart: broken out by a second categorical variable
Two variables: Time & Numeric
Line Graph: Great way to indicate the change or movement of a given value through time. When time is a variable you want to include, this very frequently will be the case.
I hope you found this helpful as you crack open data visualization at your organization. If this was helpful and you’d like more detail, in the linked post I go into far greater depth on this topic and how to potentially include three or four variables in a given visualization:
Happy Data Science-ing!