Help
* The top 5 customers represent 60% of all revenue
* The top 3 sales people account for 50% of all new sales
* The top 10 programming books account for 75% of all units
Why do I like this? First, it's memorable. Second, people are usually interested in the top values, and this formulation tells you what you want to know pretty quickly. Third, these factoids can be quite dramatic when the underlying data set has a "long tail" distribution, which is often the case for business data. (Basically, this is the 80/20 rule.) Finally, it's really simple to compute these. Here's how.
I'll use the following data, which represents the frequency with which technical topics appear on a popular programming site, to illustrate the technique :
The process is as follow:
* Step 1. Sort the data into descending order so that the largest value is at the top. In my example, this is Column C.
* Step 2. Compute the percentage of the total that the item represents. This is Column D.
* Step 3. Drag the first item's percentage into a new cell. This is cell E1. This cell will have a running total (also called a cumulative percentage).
* Step 4. Now complete the running total on the next row by adding cell D2 (the percentage of the next item) to cell E1. Then, drag this cell down for all the other cells.
Then, scan for interesting breakpoints that best represent the data. You can focus either on the rank (top 3, 5, or 10) or on the percentage (50%, 75%, or 80%). So, on this site, the top 10 topics represent 20% of all posts. (As an aside, the top one half of one percent represent 50% of all posts.) Fascinating...
The top 10 analysts agree --this format works in almost 80% of all cases.
What tips or techniques have you used to help people better understand complex information?
