The Freakonomics blog recently held a contest that challenged readers to guess the outcome of a direct-mail fundraising appeal. (Winners of Heart + Mind Donations Contest; read the post to see how people did in the contest.)
The test was a direct mail piece for Freedom from Hunger that pitted a standard "emotional" appeal against one that used scientific evidence to make the case.
The results:
- Small prior donors (up to $100) reduced giving when presented with scientific evidence by 0.9 percentage points.
- Large prior donors ($100 and up) increased giving when presented with scientific evidence by 3.54 percentage points.
My experience with tests like this has been similar, but with less difference between the small and large donors. I've found that "scientific evidence," has been harmful to response from both groups, but less harmful with the larger donors.
Here's a hypothesis that may explain both my experience and this test:
- Some donors are only lightly involved with the cause. They need to be "re-acquired" over and over again. The emotional approach to fundraising is the only way to reach that kind of donor. Science will turn them away almost every time.
- On the other side, there's a usually smaller group of donors who are connected and committed to the cause. Science doesn't turn these donors away. They've already had their emotional "conversion." A lot of facts may even be persuasive to them.
- There's a higher portion of low-involvement donors among lower-dollar donors; there are more high-involvement donors among the high-dollar donors. Of course, there are both kinds of donors in both groups, but the difference is enough to create differences in testing.
It's a hypothesis.
I've found that as you climb the donor-value ladder, you find more and more donors who are more donors who are more persuaded by information. By the time you reach donors who give $1,000 or more a year, an "investment prospectus" approach to fundraising is very effective.
The other thing to carry away from this (or any) test: One test is always just one test. You never learn the whole truth from one test; just a piece of the truth.
