Here, we will introduce a process to plan out visualizations, which we will call the Hierarchy of Questions method. Instead of starting with requirements and mockups, you start with mapping out the thought process of the audience.

After this, you pull together the data needed to answer the questions identified in the mapped thought process. The data should be in a form that can be queried efficiently and in multiple different ways. We will discuss ways to do this in a later chapter.

Finally, visuals are created which are aligned to the mapped thought process. The same data will probably be used multiple times in different ways. The benefit of following this process is that it puts the needs of the audience first, without anchoring to a starting visualization - which is likely to be whatever was easiest to create. The tendency, in that case, is to do what is simplest for the developer (you) to create and maintain, as opposed to what is best for the audience. Freeing yourself from a set starting place for a visualization allows you to be open to innovative ways of interacting with the analytics.

Remember, you are one person while the audience is many; and, in the case of dashboards, each person will have to use it many times over. The sum benefit of getting it right can be large, especially for customer-facing dashboards, which could have thousands of users.