this idea is based on how tableau treats axis and the ability to split an axis by an attribute.

example:

- i have 3 companies and i want to see the running total of widgets they purchased from me over time

- if i set up the x-axis with date and the y-axis with widgets purchased and running total i can see the total widgets purchased over time, great

- now to see it per company, this is not possible

- in tableau for reference, if i were to drop an attribute on the y-axis ahead of the # of widgets it would split each attribute into a new graph with the y-axis # of widgets repeating

- the x-axis date would still be the same for each graph so you can see differences in start/end of purchases per customer in relationship to the other customers

i have attached an example.

why this is powerful? 

- without this feature you have to create a graph for each company and the x-axis will be very hard to align since different companies may have bought widgets at different start/end dates

- data density, without being able to do this if you made a dashboard with 20 different graphs representing 20 companies the white space, miss aligned x-axis etc make it pretty useless

- management of 20 different graphs if you decide you need to make a formatting change etc is prohibitive, if they are all contained in on graph this is easy

this is the main thing i really miss about tableau, once users begin to use a feature like this it becomes indispensable to get data dense plots and see relationships or differences between attributes.