Our friends over at Bloomerang have a good post up here that argues we shouldn’t put all our donors into one bucket, sending “the file” undifferentiated communications.  Absolutely right on.

They also list out 20 segments that, for starters, you can use to differentiate your file.  Great!  And it’s only a start.

And there’s been some excellent commentary from our friends at the Agitator, who point out that there are some things like experience, donor feedback (or lack thereof), customer service (ditto), and ignoring preferences that will put holes in all your buckets.

I wanted to look at a different aspect of the multi-bucket universe: you can’t segment your file and have 20 fully different versions of each communication you create unless you are a huge organization and/or have a desire to go stark raving mad.  To focus on everything is to focus on nothing.

So the million dollar question: what goes to the top of your list?

Some would say, like Jeff Brooks does well here, that the buckets should be lifecycle categories – that it’s most important that you segment by whether someone is, for example, a lapsed donor or not.

I’d argue two other things should be at the top of your segmentation analysis: why a donor gives (identity) and how committed that person is to your organization.  (And this latter point isn’t measured by when someone last gave, but what they feel about you.)

Here’s a thought experiment as to why.  Let’s assume you work to raise money to combat a disease.  You have four people in your file: which sets of two logically belong together:

1. Someone who has the disease and last donated to you 23 months ago
2. Someone who has the disease and last donated to you 25 months ago
3. Someone who doesn’t have the disease, doesn’t know anyone who has the disease, and last donated to you 23 months ago
4. Someone who doesn’t have the disease, doesn’t know anyone who has the disease, and last donated to you 25 months ago

For me, it’s fairly obvious that A & B make sense together and C & D belong together.  However, a lifecycle segmentation approach would say that A&C and B&D belong together.

The goal is to put people in segments where they are the most like the other people in the segment.  Attributes are the best way to assess this.

Commitment works the same way.  When someone says you are their absolute favorite charity and you can do no wrong in their eyes, you want to be treating them differently from someone who is thinks you have lost your way and, unless you change, this is their last donation.  This is true even if both people are in the 4-6 month, $50-$99, multidonor segment.

One might be interested in monthly or multigiving or an event or volunteer opportunities.  The other is someone who you want to see if you can fix the relationship or save your costs by not communicating with them.

You should still do other types of segmentation and customization – the more specific a communication can be to a person, the better it is (I’m sure there are exceptions to this, but can’t think of any off-hand).  But these are the ones that will make the biggest differences in what, when, and how you communicate with your donors and thus can be at the top of your decisions tree.

The challenge is that many organizations won’t know the different answers people have about why they donate to a nonprofit and/or won’t know commitment levels of their donors.  So transactional is all they have.

We’d love to talk through this type of analysis with you; let us know your thoughts!