The other day, Douglass Turner asked me a good question. Here’s what he asked:
“One thing unexplained in your book is how to interpret relative length of the bars in the mental model diagram. In fact the vertical axis is (oddly) unlabeled. My Tufte instincts recoil in horror and disbelief in an otherwise lovely book. Could you please say something about the meaning of length? “
He’s right, I never explain it in the book. The height of the bars is governed entirely by the number of tasks that appear in the tower, which has no significance other than “people mentioned this a lot.” The model focuses on how lower boxes support the upper towers, but you wouldn’t treat a supporting idea differently if it supported a tower with just one task in it as opposed to nine tasks in it. There might be an implication of importance in the towers with more tasks in them, but your business strategy shouldn’t be tied to the size of the towers directly. You will want to focus on the supporting boxes to assign weight and priorities.
The script I use to produce the model from the data actually limits the height of the tower to seven tasks, and if there are more than seven tasks, it makes the tower wider and adds the additional tasks in a new column within that same tower. The reason for the height limit is so that you can print this thing out on regular sheets of paper and not rely on a plotter. You can tweak the script to remove this limit if you want.
Thanks for asking the question, Douglass, and thanks for letting me blog about it!