Note: Images may be inserted into your messages by uploading a file attachment (see "Manage Attachments"). Even though it doesn't appear when previewed, the image will appear at the end of your message once it is posted.
Register Latest Topics

  Author   Comment  

Posts: 2
Reply with quote  #1 
I have recently revisited a graph I put together to illustrate a specific problem and, based on the wealth of information available on this site, concerned that I am barking up the wrong tree.

The scenario is billing related.

At the top level customers can be grouped as either offline (paper billing) or online (eBilling).

Having said that, I was aware of some bad behaviour regarding a subset of the online (eBilling) customer base: those that had not signed on within a certain period. So I broke the online (eBilling) customers into 2 sub-categories, engaged and not engaged.

The KPI I was looking at was the average amount of debt associated with each customer type.

My analysis of the situation showed that the bad behaviour was actually not the customer's but the incorrect creation of the online account in the 1st place.

So to illustrate the point that online (eBilling) customers were the best and the worst type of customer, I published this:

The sectors show the scale of the grouping and the radius showed the scale of the average debt.

 I was looking to generate the follow up question of "what makes someone red" so that the underlying problem could be discussed and resources allocated to resolve it.

Am I trying to combine too much into the one graph?


Posts: 852
Reply with quote  #2 
Your graph is a variation of a Nightingale Rose (a.k.a., Coxcomb or Polar Area) chart. It differs only in that you've hollowed out the center of the pie, transforming it into a donut. Perhaps the most significant problem with this chart is that we perceive objects that vary on two dimensions, in this case arc (i.e., distance around the circumference) and radius (i.e., distance from the center), in a combined manner as areas, but the areas in this case are meaningless because they combine two quantitative variables (percentage and average debt of customers) that cannot be summed. Percentage and average debt should be perceived independently. The following combination of side-by-side bar graphs displays the data in a way that does this in a way that is easy to read. It also clearly features the problem with disengaged online customers.

Rose Redesign.png 

Stephen Few

Posts: 2
Reply with quote  #3 
Thanks Stephen, very helpful.

Loving Big Data, Big Dupe!
Previous Topic | Next Topic

Quick Navigation:

Easily create a Forum Website with Website Toolbox.