2 TXT OR NT 2 TXT?

Text messaging is a growing field for nonprofit communications.  M+R Benchmarks found that last year text messaging audiences grew by 26%, while email lists shrunk by 2%.  Metrics are also strong with a 4.2% click-through rate for fundraising messages (with .56% for fundraising emails).

On the flip side, it’s still a small audience — less than eight percent of the average email lists, so it’s expected that it will grow more in the same way I expect my 11-year-old will grow by more than I will this year.  This also means that the increased results could just be because it’s your die-hards who are using it rather than strength of the medium.  And with increasing text volume, unsubscribe rates to texting programs increased by 54% year-over-year.

Because lists are smaller, there’s precious little data on whether texts really lift response rates; it’s tough to have a list large enough to make significant conclusions. 

Unless you look at blood donors, which several Danish researchers did for us.



They took 20,365 blood donors.  A random half of the people received a text message when their donation was used about the impact of their gift; the other half didn’t.  They also randomized the time of day the messages were sent.

The results?  These messages increased subsequent donations by 3.6%, with evening texts doing the best at 6.5% lift.  So let’s go text wild!

Except…

The difference was only among plasma donors; there was no significant difference for whole blood donors. 

Interpreting these results requires a bit of knowledge of the psychology of blood donors.  In a study of different types of blood donors, plasma donors were more likely to say:

  • “Helping people is in my nature”
  • “I think there is a strong need for blood products”
  • “It gives me a sense of pride”
  • “I like to have goals”
  • “I receive telephone reminders”

Whereas whole blood donors were more likely to say:

  • “It’s a positive thing to do and requires little effort”
  • “I give thinking that a member of my family or a close friend could need blood someday”
  • “I give when a blood drive is being held near where I live or near my workplace/place of study”
  • “I give when I see posters and advertising”

So, plasma donors are your gung-ho people.  Giving generally and giving blood specifically is part of their nature and identity.  They are filling a need and that gives them pride; they seek out that opportunity.

Whereas whole blood donors (and note we are painting with a broad brush here; individuals in each group will vary) are convenience donors.  They’ll do it if it’s easy or nearby.  And they do it because it will help them and theirs.

To summarize: telling people about their impact… increases retention among the people who care about their impact.

So in an ideal world, you’d text plasma donors when their blood is being used and text whole blood donors when there is a drive near them where they can get a free T-shirt, right?

Wrong.

First, if we lived in a perfect world, most of us would be happily out of a job because our mission was done. 

But skipping over that, in an ideal world, we wouldn’t go exclusively off this group identity.  There are almost certainly whole blood donors who care about their impact and plasma donors who want convenience.  We’d be able to ask about, or model for, what message would be most effective with each individual person.

So:

  • Level 0: No segmentation — everyone gets the same message
  • Level 1: Bucket segmentation — messaging is based on a single factor
  • Level 2: Bucket-free segmentation — everyone’s messaging varies based on what they want or the closest approximation of this.

All of this to say, this is why we are holding a webinar on September 29 called Death to buckets! Toward more advanced and profitable segmentation.  We’ll discuss how multiple data points are better than one or one type and how RFM segmentation and models based only on transactions are as good as the segmentation between whole blood donors and plasma donors: good, but we can do better.

Sign up for Moore updates