Pat Peppler, President    August , 2006

Directions Customer Scoring

 Having researched several of the industry methods of scoring or ranking customers in a direct marketing database it is obvious that there is no “one” best method. Customers need to be scored on many attributes and we do not have all the required information on every customer from the fulfillment history. 

In the future, we will be getting more information about our customers from them during each contact and from other public sources.  For now, we will score customers on the three classic attributes: recency, frequency, and monetary values for behavior.

Every customer will have three base scores from 5, for the best, to 1 for the least. The best customers will have scores of 5 for most recent, 5 for most frequent, and 5 for highest value. These scores will then be weighted to give each customer a single value from 100 to down to 20.  Customers will be re-scored every month and just before any mailing or promotional activity.

Recency  (John Wirth, PhD.)
Customers will be scored based on their most recent purchase.  Customers who have purchased in the last 90 days are scored 5. Buyers 4 to 6 months old are 4, 7 to 12 months are 3, 13 to 24 months are 2, and the rest are 1.  Most direct marketers consider recency to be the primary indicator of “propensity to buy.” 

Frequency  (Ted Miglautsch)
Frequency scoring is based on behavioral quintile scoring or segmenting customers into five groups.  Customers with only one purchase are given a score of 1.  The remaining customers’ total number of purchases is averaged and all those customers having a number of purchases below the average are scored 2. This process is repeated with the remaining customers two more times until all customers are scored. 

Monetary (John Wirth, PhD.)
Monetary scoring is also based on segmenting customers into five groups.  Total customers’ sales are divided by 5 to give the total sales for each segment.  Customers are then ranked by total sales.  Starting with the most valuable customer, customers are scored as 5 until the total customers sales equals that segment total, the next customers are scored as 4 to fill that segment until the last customers are scored as 1.  As in the previous attributes, this gives us 5 groups of customers and each group has similar behavior for that attribute. 

Weighting RFM 

Every customer now has three scores.  The best customers are 5-5-5.  They have purchased in the last 90 days, most frequently, and highest total sales. The least productive customers are 1-1-1. 
We could just add these scores to give every customer a rank from 15 to 3.  Most direct marketers give recency much more value than either frequency or monetary.  The most recent customers are most valuable.  Frequency is also considered more valuable than monetary. Customers that order more often are more likely to order again.
Weighting recency, frequency, and monetary is accomplished by multiplying the scores by 9.9, 6.6, and 3.3.  This gives our best customers (5 x 9.9) + (5 x 6.6) + (5 x 3.3) a score of 99 down to 19.8 for our least productive customers.

 What do you do with the scores now? 

Now we market to our customers.  You can run reports listing the numbers of customers with each score with a cumulative total.  To produce a mailing file you just select customers with scores through the quantity you want to mail. 

This scoring seamlessly works in conjunction with all other selection criteria such as zip code or product category selections.

Sources: 

Libey, Donald R.  “Five Defining Trends in the Catalog Industry” in The Library at www.libey.com/five_defining_1.htm Dec. 14, 2000

Libey, Donald R.  “Libey on RFM value” at www.e-rfm.com/top.html Dec. 14, 2000

Libey, Donald R.  “The New Direct Marketing Management Structure: Evolving Trends From The E-Com Revolution” in The Library at www.libey.com/new_direct_1.htm Dec. 14, 2000

Libey, Donald R.  “Too Few, Too Little, Too Much: Why Direct Marketers Fail…  And Ten Ways To Make Sure You Don’t” in The Library at www.libey.com/too_few_1.htm Dec. 14, 2000

Miglautsch, John.  “Thoughts on RFM Scoring” at http://migmar.com Dec. 10, 2000.