Are Business Shifts Positive?

Recall how our customer file evolved over the past six years.


There are a reasonably constant number of customers at X = -1 / Y = 0.5. These customers buy about 1.3 times a year, and only buy from the primary product category - generally buying via email marketing.

There are a reasonably constant number of customers at X = -0.5 / Y = -1.2.  These customers buy only one (inexpensive) item from the primary category via online. This probably isn't a segment that has a high repurchase rate.

I created a GIF of repurchase rates over the past five years. How does this GIF compare to the GIF above (if you cannot see the GIFs, please visit http://blog.minethatdata.com)?

Notice that the largest file count areas have poor repurchase rates.

High repurchase rate cohorts are reasonably consistent over time ... customers around X = -0.25 / Y = -0.5 tend to have the highest repurchase rates. Who are these customers?
  • They order about three times per year.
  • 80% of their dollars come from the primary product category.
  • Nearly all of their dollars are online - only 15% come from online marketing channels.
So the most loyal customers purchase every four months ... generally buying from the most popular product category ... and they're so loyal that they don't need marketing to drive their purchases.

This creates interesting challenges, when looking to the future. File growth generally comes from low-repurchase segments. We need to see if "some" of these customers will migrate to high value in the future, don't we? Let's take a look at that tomorrow.


"Forecasting outcomes are the sum of all analytics and marketing knowledge possessed by your company."