My family loves making Christmas cookies! This last Christmas to prep the grocery list I quickly put the recipes on a spreadsheet by ingredient. My business intelligence kicked in and I noticed some patterns in the cookies. I made note of the common sugars used – white sugar, brown sugar, molasses, and powdered sugar – and how they are similar to what we call “grain” in BI.

For convenience in recipes, we typically use the highest grain available for ingredients to save time unless we feel the quality of the recipe will be sacrificed. For example, brown sugar is a mixture of white sugar and molasses. But I don’t typically make my own brown sugar. I also do not pulverize granulated sugar to create powdered sugar. I will say, in everyday cooking, I do tend to use less refined ingredients to get more natural flavors, but these are Christmas cookies. It’s ok to not crush the sugar cane by hand.

With BI the opposite is true – always use natural ingredients. Your models should be designed at the lowest grain possible. There may be insights that are lost within data that has been massaged into a higher grain. You may not be able to answer questions that come up during analysis. Also, as you add more business processes to your platform you may need to connect them together at a lower grain. You can always rollup data to a higher grain for analysis if needed. You can never take the grain lower without using assumptions – and assumptions are bad for making decisions. The one exception to this rule is if your system cannot handle the workload and it is cost-prohibitive to upgrade.

We would love to help you setup a BI platform with the correct grain. Contact us today!