Cohorting Analysis is a common analytical pattern we see our customers using at Omni. In our documentation page, we have an example of this pattern in practice. In this example, we create a cohort analysis based at cohort of users’ first order dates. In order to calculate the first order date for each user, we needed to create a query view.
With our new released level of detail (LoD) calculations, you can skip the create query view step. To define a new LoD calculation you can find the field you would like to aggregate (Order Date in this case) and then go to the three dots on the field > Modeling > New Level of Detail field.
In the editing pane, you will want to use the Fixed LoD Grouping to group by User ID since we want to identify each user’s first order. This calculation aggregates Order Date by User ID and ignores other fields in the query that might otherwise affect the result.
Now without needing to create a query view or perform any joins, we can bring our new Level of Detail field, User’s First Order Date, Order Count and Order Date into a table to see how cohorts have bought products over time.
