They were also able to bring in additional sources to help refine insights. For example, when examining two stores that showed diverse results, yet had similar patterns across sales, menu, time of day, etc., they could apply broader demographics, such as area income or ethnicity.
They might break out data not just against their metrics, but against data such as city, metro, or rural locations, locations to schools, military bases, or public gathering places, age groups and income levels, etc. They can also inject additional parameters to predict future performance, such as which side of the street they are on or whether the route is on the way to work or the way home. This data not only helps to better forecast store performance, but also helps decide where expansion will be most successful.