Execute in-depth cohort analysis with BIME

When working with your CRM data, you may want to perform cohort analysis. A cohort is simply a group of people who share a common characteristic over a period of time, to give a simple example, visitors to the BIME website in November might be our ‘November cohort’. Cohort analysis helps isolate engagement metrics from growth metrics therefore it can be useful to track customer retention, brand loyalty, the effectiveness of particular campaigns, etc. This is very easy to do with BIME, as data can be grouped, segmented, and otherwise organised, using our powerful calculation engine.

At a basic level, for a date cohort, you can create a calculated attribute that picks up the level of date you want to use - in our dashboard, we have used the month in which a customer subscribed. The ‘formula’ could not be simpler - just pick up the field from the list, and you have organised all your customers into the right cohort!  

For more complex groupings, you can use the same approach, but concatenate several elements, for example the month, and the year, and the region - and our grouping function means that you can even reorganize your reorganisations!

You can then track their activity over time based on other date fields (here ‘order date’) or compare key performance indicators between them.  Your cohorts will be dynamic and react to your customers’ other attributes, such as location, sales channel, etc.  As BIME allows you to select the aggregator of each metric used, you can also easily see the quality of the members of different cohorts, even if they differ greatly in size - switch to see the average or median of your KPIs to get a deeper look at each group.

If you need to use more specific date ranges, or factor in value, you can do that to!  The IF...THEN... function in Bime allows for the application of sophisticated conditionals to be applied to organise data into very targeted groups to ensure actionable results.

When working with multiple date fields (such as ‘first order’ and ‘last order’) you can also use the DATE_DIFF function to calculate loyalty at whatever level you need, as it will report the number of periods (from years all the way down to seconds!) between your date markers, to allow you to target customers who may be at risk of being lost, plan periodic contact, and inform your forecasting.

View an example and explanation on our example dashboard below.