A Repeat Customer Insights customer was asking about how the Frequency scores are calculated:
How are your coming up with your frequency scores of 1-5? I get the recency and grading methodology but would like to better understand how those are being calculated.
All three of the RFM scores are based on five groups of 20% of your customer base (called quintiles). Frequency is based on how many orders a customer has placed.
Frequency is using modified version from the regular RFM algorithm to account for how ecommerce customers behave when shopping online. Online the majority of customers tend to place only one order which can make the regular RFM algorithm harder to use.
The modified version I use works better.
Instead of splitting your entire customer base into 5 groups of 20% like the default RFM algorithm, a score of 1 is used for customers with only 1 order. Then scores 2-5 are split into 4 groups of 25% of the rest of your customers.
2's are the bottom 25% of your customer base (e.g. ordered 2 times), 3's are the 25% above them (e.g. ordered 3-5 times), and so on.
This lets you easily compare all of the one-time customers (score of 1) against the returning customers (scores of 2-5) which have four groups based on their purchase frequency.
It's a more realistic view of their behavior which means you can do a better job of targeting customers and their needs.
If you haven't installed Repeat Customer Insights yet, it's an easy way to get a detailed look at your customer behavior. This RFM analysis is included in all accounts and it comes with a 14-day free trial so you can see how it works before paying.
Eric Davis
Market to your customer's timing
Figure out how long customers wait in-between purchases and you have a key component for your marketing timing. This is the basis of the Average Latency metric and Order Sequence Report in Repeat Customer Insights.