RFM Scoring: Who are your best customers or donors?

The Recency report:
Customers are scored based on how long ago they last purchased.
RecencyReport.pdf

The Frequency report:
Customers are scored based on how often they purchase.
FrequencyReport.pdf

The Monetary report:
Customers are scored based on how much they have spent.
MonetaryReport.pdf

The Summary report:
Sorted descending by RFM Total – shows various statistics.
summary.pdf

The Breakout reports:

Sorted descending by RFM Total (permutations are broken out).
BreakoutReport_SortedBy_RFMtotal.pdf

Sorted descending by # of customers scored per RFM permunation.
BreakoutReport_SortedByTotalCustomers.pdf

Sorted descending by maximum difference between RFM scores.
BreakoutReport_SortedBy_MaxDiff.pdf

If you’re planning a marketing campaign, the RFM scores can be extremely useful in choosing your targets.
Each record you receive back will have been scored as to (R)ecency, (F)reqency and (M)onetary with a value of 1-10.

You can use these scores to be very selective about tailoring your offer to a particular audience. As an example:

• If it’s an RFM for a fundraiser, then donors who have a high (F)requency who are not yet monthly donors may be convertible to monthly donors.
• If it’s an RFM for a business, customers with a high (M)onetary and or high (F)requency but a lower (R)ecency score shows a valuable client who may be shopping elsewhere. The correct offer might just be what’s needed to get these clients back.
• If it’s an RFM for a business, customers with a high (R)ecency score but a low (F)requency score are new clients. You could use this opportunity to send them a “thank you” or a “welcome” package.

For a basic selection based on the RFM codes; Here are some points to consider.

The (R)ecency score is the most important indicator. Take a look at the dates on the recency report and judge, keeping in mind your business cycle, whom you think are unlikely to respond to your marketing efforts.
That is where you should draw the line in terms of choosing records scored by (R)ecency… except for a few gems that can be found in the (F)requency and (M)onetary groups.

Looking at the (F)requency scores, it might be a good idea to also choose targets whose (F)requency scores are at a level where, if they’ve purchased or donated so many times, they might just do that again. Therefore, even if a (R)ecency score might be on the low end, a high (F)requency score indicates that in the past this person or organization has shown to be a willing client or donor.

Looking at the (M)onetary score, the high value donors and clients with a lower (R)ecency score may produce results if approached in the right way. If the value is high enough, it might be time to personally visit them.

For more ideas on using the RFM score, look at the MAXIMUM DIFFERENCE field. The MAXIMUM DIFFERENCE is a number that’s been calculated by subtracting the highest of RFM from the lowest of RFM. High values here can indicate opportunities. It’s not always the case but it’s worth exploring.
Take a look at the breakout reports sorted in various ways to discover opportunities.

Over time, each time you run RFM, you should store those values on the donor or client record. That way, you’ll be able to tell if that persons score is increasing or decreasing. Another factor for your selection.

Using the RFM score is not rocket science. Use your imagination. You can do it!

RFM scoring looks at these 3 areas of your data:
R – Recency: How long ago was the last activity?
F – Frequency: How often is there activity?
M – Monetary: How much is spent?

Before beginning the data mining process, we will choose a timeframe of activity to analyze. Typically we recommend 3 – 5 years, but we will make customized recommendations based on your database, organization and purpose for the RFM scoring.

Your data may be scored into anywhere between 3 to 10 slices using a numerical scale from 1 to however many slices you’ve chosen. The more slices you choose, the more granular the scoring becomes.

For this example I will use 3 as the number of slices for simplicity sake. Each client or donor will receive a number of 1 to 3 for Recency, Frequency and Monetary.

Here are some examples of how data can score during the RFM scoring analysis (one number for each of the R/F/M in that order):
ie: 3/3/3 indicates the top of the scale for recency, frequency and monetary
ie: 3/2/1 indicates most recent, average frequency, and lowest on spending
ie: 2/2/3 indicates it’s been a while since the last activity, average frequency, and biggest spender.

Once the RFM scoring of your database is completed, it can be used to further analyze your information. For example, let’s say you have diligently collected the source of your customers or donors and recorded it in your database. You can now select all the high performing spenders, the “3’s”, and analyze what their source was. Where did the name come from? By being able to answer this question, you have a valuable edge in targeting your marketing efforts. This is just one example of gaining insight and we’ll consult with you to mine the right data for the right insights.

If you are planning on doing a direct mail campaign, you may wish to target your “Triple 3” list (who scored high in all areas) on a regular basis, while you may want to send an enticing re-activation mailer to your “1/3/3” list (it has been a while since their last transaction, but they used to spend frequently and were a big spender).

Also, if your database can store the RFM code transactionally (meaning you can store multiples over time, each stamped with the date of when it was done) you will have the ability to know who is increasing/decreasing their business with you. This is very valuable information!

Before we perform the RFM scoring, we want to ensure your data is clean of duplicates as this would distort the results. Our de-duplication process should be applied prior to RFM scoring.

We’ll work with you to make RFM scoring a successful tool for your organization.

We supply you with 1 data file (Tab delimited TXT), which contains the scored data and several report files (as CSV’s).

It will have 1 record for each client or donor with these fields:

• Client ID
• Most Recent Transaction Date
• Number of transactions (ie: purchases or donations)
• Total Amount
• Recency Score
• Frequency Score
• Monetary Score
• Maximum difference from one score to another (shows disparity – possible opportunities)

Results_CompanyRFM.pdf

> Go to view sample reports

We need a file with 3 data fields:

The 3 fields are:

• The client ID – Unique to each client
• The transaction date
• The Dollar Amount

Before you submit the data to us for scoring, it is a good idea to remove high value donors as they will skew the results. For instance, if you have a donor who regularly donates $5000, you may not want to approach this person with your regular marketing efforts. Instead, you might want to phone or visit that kind of a donor and; may not want them to affect the scoring of the others. It’s up to you.

A COMMA or TAB delimited file is best:
Sample for RFM Scoring.pdf

The main data file is returned as a Tab delimited TXT file.
The report files are supplied as “.CSV’s” which can auto open in Excel or Open Office/Libra Office.
Hint: You may need to drag open the first row on the reports to view the column headers properly.
The sample PDF’s have been converted and formatted for easy viewing.

These samples are from an RFM scoring performed for a company.
The RFM scoring works equally well for donor type organizations.
Perhaps you’ll think of new ways to use it.