The Basics of RFM Segmentation - RFM Enhancements

This section briefly describes three major factors that can be used in segmenting customers: product history, industry code, and company size. While the specifics vary for each mailer, here are some guidelines that I follow.

Product Purchase History

The buyers of different types of products act very differently from each other. If you sell several distinctive product lines in one catalog, you should look at RFM and retention reports for the different product buyers.

If you see greatly different behavior, you may want to construct different mail plans for each type of buyer, including specialized pieces and even catalogs. You should explore the idea of personalized reorder letters for product lines where that makes sense.

You can prepare RFM Dynamics reports that dramatically show the growth and decline of each product line. See the Reports section for an example.

Industry Code, Employee Size and Prospecting

For customers, RFM and product history are far more valuable than SIC/NAICS and employee size. For prospecting, industry and employee size are very valuable since they are one of the few facts consistently available.

You have to understand your existing customers in order to gain new ones. You may not use the information for house mailing decisions, but it is critical for prospecting.

You start by adding and maintaining industry code and employee size to your existing customers. This is a project all by itself and an ongoing list maintenance chore.

Create a special RFM structure with just a few cells so you can have a mini-RFM report for each industry and employee size combination of interest. Make sure it fits on one page! A possibile structure is recency of 5 year groups (1-4 and 5+), frequency of 1, 2 and 3+ times and monetary of high, medium and low average order size.

Start at the two digit SIC level and use just two employee sizes: 0-4 and 5+ employees. Run the RFM for each industry/size group. Scan through the 200 pages looking for items of interest. Focus on those areas for more detailed analysis. A business intelligence tool such as Cognos Powerplay would be very useful. Any areas that look interesting can be explored in more detail at the industry and size dimensions.

Get the universe totals for your groups so you can compute market penetration. Look closely where you have high penetration. Why do these folks like you? Look where you should have penetration but don't. Why is that? Have you ignored them? Are you missing a product they need? Would a special offer solo pitched at them be appropriate?

As you learn more about your customers, you will discover prospecting opportunities. A very nice part of this approach is that the names of interest can be rented cheaply from compilers such as InfoUSA. If the compiled names work you can mail the entire universe at low list cost. If the compiled names do not work well enough but show some potential then you can explore renting better names from controlled circulation publications, public databases, and specific response lists.

This is a great way to gain knowledge about your customer base and to leverage that knowledge to improve customer acquisition.


Next section: RFM Reports

All sections: