Contributed by Bruce Gregoire, President, WiseGuys Marketing
Adjunct Professor/Marketing, Carey School of Business, Johns Hopkins Univ.
Introduction: In Part 1 of 2: “Customer Scoring in MOM using R-F-M”, we described the “Why” of using your R-F-M utility (Recency-Frequency-Monetary Value) in M.O.M. Basically, to segment your customer base. Segmenting is a fundamental marketing concept to differentiate your customers. Your R-F-M metrics supports this concept, by measuring buying behavior.
In this posting, we will describe the “How” of using R-F-M. As a reminder from Part 1, you may recall the following:
- Recency, as you might guess, is a measure of recent customer activity.
- Look in your Cust table to find the”odr_date” field, which stores the most recent order date. Recent customers will get a higher score – “higher” is “better”.
- Frequency is simply the count of orders, over the lifetime of the customer.
- Look for the “ord_freq” field in your Cust table. But note that the ord_freq value is not the same as your R-F-M Frequency score: more on this later.
- Monetary Value is the sum of the dollar value of those orders.
- Look for the “gross” field in the Cust table, which stores the dollar value. Obviously, “big spenders” get a higher score.
Steps to Implement R-F-M in M.O.M:
The M.O.M. cust table contains a field “RFM”, which displays the sum of the 3 R-F-M individual scores. In our scheme, we will use a range of 0-100 for this field.
Start: Open your List Management module, then the RFM utility.
1. Assign Recency scores: We recommend using 5 scores, representing 6
month intervals. Thus, establish your settings as follows:
- Range from 1 to 26 weeks: Assign a score of 50
- 27-52 weeks: assign a score of 35
- 53-78 weeks: assign a score of 25
- 79- 104 weeks: assign a score of 20
- 105-999 weeks: assign a score of 15
Note that the descending scores do not drop in a linear fashion: they fall more radically in the 1-26 week and 27-52 week periods, than in later ranges. This is meant to highly favor recent buyers: see footnote below for this reasoning. Also, if you have a large database (500K+ customers), it may be worthwhile to statistically test these scoring assignments. See 2nd footnote
- Assigning Frequency scores: we recommend the following assignment scheme:
- Range from Freq = 3+: Assign a score of 25
- Freq = 2: Assign a score of 15
- Freq = 1: Assign a score of 5
- Assigning Monetary Value scores: Monetary Value varies widely for each MOM user. If you can estimate the $ threshold for each weight assignment, 25 down to 5, that would be the simplest approach.
For instance, if your “Big spenders” tend to spend $800-$1,000 over their lifetime, Assign a score of 25 for $800 – $9999. Then continue as follows:
- Score 20 for $600-$799
- Score 15 for $400-$599,
- Score 10 for $200-$399,
- Score 5 for $1.00-$199
Note that If you want to be more precise with Monetary Value assignments, contact our WiseGuys Marketing staff for assistance.
Give R-F-M a try and see if it works for your business model. There is no risk with M.O.M. – populating your RFM values in the Cust table does not touch other functions. Tell us about any success stories you may have! R-F-M is routinely used for selecting customers for catalog/direct mail mailings – but you may have your own uses that advance your customer marketing. One M.O.M. user has found R-F-M valuable to select VIP customers for holiday greetings at year end!
If you want to jump-start your R-F-M scoring with M.O.M., contact the WiseGuys Marketing staff at 703-941-8109, or via email at BruceGregoire@WiseGuysMarketing.com. Mr. Gregoire is currently a M.O.M. Consulting Partner with Freestyle, and a M.O.M. consultant since 1999.
 *Recency scores: To explain the non-linear drop in recency scores, consider the following. Brand new buyers, who just found your enterprise, have not had the luxury of ordering with great frequency, or with high cumulative monetary value, from your business. They are in their “honeymoon” phase with you! The challenge this presents is that you won’t want to score them – comparatively – with long time buyers. They have not had enough time to love you, with multiple orders, as long-time buyers. The solution is to score the new/recent buyers very high, as shown above, to compare them on an even playing field with veteran buyers.
 Statistical testing: contact WiseGuys Marketing staff for help with multiple regression analysis. This analysis may confirm whether the scores we recommend are indeed accurate for your business model.