How to find attitudinally derived segments across your entire house file with the “ABA Sandwich”

August 6, 2012      Kevin Schulman, Founder, DonorVoice and DVCanvass

For direct marketers (and researchers for that matter) the Achilles heel of attitudinal data is not having answers for everybody, only our sample.  This tends to relegate findings to interesting, maybe even useful but never mission critical – at least in the eyes of fundraisers and direct marketers.

What if you could know, with reasonable certainty (as reasonable as a predictive model scoring likelihood to renew, etc..), attitudinal insight at the constituent level and have it applied across the file?

In geeky research land or maybe just geeky DonorVoice land we refer to this as making the data addressable or findable on the donor file and…it can be done.

It is worth spending a quick minute reviewing why this is worth doing; what is the value of using attitudinal data, what does it provide that behavior only information does not?  To put it succinctly, it is the difference between prediction and explanation.   Behavior only insight misses causality, motive and intent and preference and needs are, at best, inferred, and more often ignored.  If we understand what makes donors tick and group them accordingly (presuming there are different segments with different needs, preferences and as importantly, prospective financial value) then we can be in the “creation” business – the business of making more good donors and increasing the relationship strength and Commitment of the already “good” ones.

Returning to process now we introduce the ABA sandwich – the Attitude, Behavior, Attitude sandwich (we so wanted the acronym to include an extra “B” to drop in a gratuitous  Swedish rock band reference – oh wait, we just did anyway).

 

The graphic is hopefully mostly self-explanatory but since it is high concept and we know there are doubters out there, we also included a sufficiently geeky function and visual to show a real example of doing this to identify two attitudinally defined segments – High Commitment and Low Commitment based entirely on strength of relationship, which we define entirely by attitudinal measures.