I’ve been preaching testing donor identities as ways of segmentation your file even in my pre-Agitator days.  But here’s a secret: not all donor identities are created equal.

Remember that the goal of a good segmentation – including an identity segmentation – is to minimize difference among group members and to maximize difference among different groups.

So, for example, if all men behaved like all other men, and all women behaved like all  other women, and those two groups really were from Mars and Venus, that would be a perfect segmentation.

(Respectively: they don’t, they don’t, they aren’t, and it isn’t.)

There’s an additional and very essential criterion: segmenting and customizing by that identity must increase your results.

Not all identities meet these two criteria.  For example, you may have a segment of advocates on your file.  They love to sign petitions and have contacted their member of Congress so much, the Member is considering a cease and desist letter.  (Perhaps you think members of Congress wouldn’t do this.  You are mistaken.)

But if advocates are worth the same as non-advocates, reply to the same pieces as non-advocates, and don’t respond any better when you add a petition to the mail package, then — to use a Josh-Whichard-ism–“the juice isn’t worth the squeeze.”

In the next post I’ll cover how to test to make sure these variables are worthwhile.

But first, we need to get a list of potential identities to test (after all, if not all identities work, we may go through a few before we hit the best option).  Here are some — non-exhaustive– thoughts for brainstorming identities:

• Direct impact (e.g., I have the disease; I have been poor myself) versus indirect (I love someone who has the disease/has been poor) versus none
• Receiving services from your organization versus received services from your organization versus no services
• For religious organizations: “I give to live out my X religious identity” versus “I am not a X, but you do good work”
• Religion, even for non-religious organizations. You may laugh, but some animal organizations have reported that they have a segment who respond differently to subtle cues like “helping all God’s creatures.”
• Interacted with your organization directly versus not. e.g., behaviors of people who have adopted an animal from your shelter are often different  from those who haven’t.
• Interactions with your mission. Similarly, someone who adopts an animal from a shelter may respond differently for an unrelated animal organization.
• Volunteers versus non-volunteers.
• Relate to volunteers versus relate to those served. Let’s say you are Habitat for Humanity – you may have a segment of people who wish they could be out there swinging a hammer, but who are doing so vicariously through your volunteers.  A traditional appeal emphasizing the plight of the homeless may not work as well for this group.
• I’ve used cats versus dogs repeatedly as an example because 1) it’s simple and 2) it works.  What’s your organization’s equivalent?
• Especially for cultural organizations, have they been to the museum/opera/library/monument/park or not? One such organization reports that the average lifetime value of a donor is cut in half once you get more than 50 miles away from their site (one of the few relevancies of location).
• Content patterns. What program(s) do you tune in to your PBS stations to watch? (Answer: for me, British people killing each other: Sherlock, Poirot, Miss Marple, etc.; for my wife, British people not killing each other: Downton Abbey, Victoria, Pride and Prejudice, etc.)  Is there a difference in donors by what content they consume?
• For environmental/nature organizations, do they like to experience nature outdoor or indoor? I’m in the former group – I like nature if and only if I don’t get any on me.
• An interesting quirk of crime victim organizations is a small subset of donors who may have committed the crime or an antecedent in their past.
• Disaster versus non-disaster donors. Identity in these types of donors is usually based on how they came into the organization rather than by self-identification.  Sometimes a person may have donated to a mail package because it was in front of them when a disaster hit and they are actually disaster-only donors.  Or someone who donated to a disaster, but is willing to support other efforts
• Globalists versus “here firsts.” If you serve both populations, you will likely have some donors who want to go as far as to restrict their donation to their home country… or to anything but their home country.  Clearly different behavior…
• Parents versus non-parents. We’ve also seen a variant on this where parents who involved their kids in their philanthropy are different from parents who don’t.

To any of these ideas, add in anything that you are seeing as a pattern in open-ended feedback.  You are actively soliciting feedback, right?  A simple “why did you donate today?” question will not allow you to create an identity then and there (you’ll get answers like “because you mailed me,” guaranteed).  But it will give you ideas and patterns from which to work.

Other identities you have seen?

Nick