By John Landsman, Director, Strategy and Analytics

We were at a professional conference recently, where a marketing executive from a well-known retail brand admitted, “We have so much data — we don’t know what to do with it.  How do we even get started?”  It seems like a surprising question to hear in the 21st century, but it’s really not all that uncommon to be considering how best to leverage your customer data asset.  

Let’s look at one simple approach.  It can serve many purposes, and doesn’t cost a lot to do.

We’ve talked often in this space about the critical importance of segmentation — the classification of and targeting of customers based on who they are and how they behave.   The concept is vitally important given the value that is added to the productivity of email messaging (and lots of other marketing effort) when it’s precisely targeted.

The ‘who’ part (e.g., gender; location; other demographics) is hardly trivial.  But the “behave” part is even more important, because the best predictor of behavior is  . . .  behavior.  And the behavior of special interest is primarily related to transaction-level purchasing; that is, based on a detailed record of who bought what, where, when, and for how much.  Browse-level data are also extremely useful to us, but the primary focus of today’s Nugget is the classification of purchase activity.

“R.F.M.” is a Customer Relationship Marketing (CRM) term associated with purchase activity.  The term has been in use since the beginning of CRM as a specialized field.  It’s familiar to most, but here’s a quick refresher.

RFM metrics are generally considered within a 36-48 month time frame.  They are:

  • RECENCY:  Number of days since the customer’s last purchase
  • FREQUENCY:  How many purchases has the customer made in the analysis timeframe
  • MONETARY:  How much has the customer spent in that timeframe?

Each of these measures can be shown in simple distributions that reflect the ranking of customers on that measure.  For example:

  • RECENCY:  Quintiled (five equally size segments) from most to least recent purchase, stated in days
  • FREQUENCY:  Divided into two categories:  1 visit; 2+ visits
  • MONETARY:  Quintiled from highest to lowest total spending

Of course, each of these distributions can be considered and applied on its own.  But the real power of RFM comes from combining the three measures into one view.  This view can then be subdivided to delineate several actionable RFM segments that can be named and ranked based on their strength or potential, and then treated accordingly.  Segment names can be anything descriptive you choose.  And there doesn’t need to be an airtight scientific basis for making the segment delineations, although they can and should be validated.  

What’s really important is identifying the segments and then creating marketing objectives and effort around them.

The analysis can look like this:

Screen Shot 2015-09-24 at 10.54.56 AM

 

Segmenting customers in this way identifies many opportunities to further analyze, and then develop and deploy special programming targeted at customers based on their RFM status.  Examples:

 

  • Segment status becomes a field on the customer database, and the basis for analyzing and targeting customers.  Segment status is refreshed periodically, allowing for the tracking of customer migration between segments over time.
  • Clearly identifying ‘best’ customers begs the questions of whether they’re being properly recognized and retained — and how to migrate other customers to ‘best’ status.
  • There is huge potential in migrating to multi-purchase status what may be a very large number of single-purchase customers.  Every business has these.
  • Pinpointing the significant percentage of the customer base that’s been purchase inactive for 12-24 months allows for targeting tactics to reengage or reactivate these customers
  • Each of these groups can be further analyzed to determine the extent of their email engagement, and mailability. This analysis allows for identifying which email inactives may be worth reactivation efforts, and which should no longer be emailed.
  • These RFM segments can also be profiled demographically, adding considerable insight into who and where they are, and how they can best be approached.

 

Given access to the required data, RFM analysis is a quick and easy way to profile your customer base.  Results are clear, highly actionable, and — when truly acted upon — they can have big, positive impact on customer development, audience optimization, list health and inbox reputation.

 

 

Comments are closed.