The fundraising study everyone cites couldn't be replicated

If you've sat through a fundraising workshop, there's a good chance you've heard about Rokia. She was seven and lived in Mali. In a 2007 experiment, a letter about Rokia raised more money than a letter describing food shortages affecting millions of people across several African countries. In a version where researchers added those statistics to Rokia's letter, people gave less [1][2].

The study, by Deborah Small, George Loewenstein, and Paul Slovic, became the backing for a rule most development teams follow without realizing it: 1) lead with one person, and 2) keep the stats away from the ask.

In 2023, Maximilian Maier, Yik Chun Wong, and Gilad Feldman ran a replication of those experiments, this time asking people about hypothetical donations. They found next to no correlation between the use of numbers and the amount donated [3].

That doesn't settle it, and I don't want to write as if it does. After all, the 2007 participants were deciding what to do with cash they had just been paid. The participants in 2023 were asked to imagine what they'd give. People say one thing about imaginary money and do another with their own.

What interests me more is a question the debate mostly skips. If focusing on one person in a letter does help, why?

The 2007 article's answer was sympathy: numbers push readers into a calculating frame of mind, and calculation dampens feeling [1]. If that's right, the practical rule is to protect the emotional moment. One person, no statistics nearby.

There's another possibility that doesn't depend on sympathy at all. A concrete scene may simply be easier to take in than a number. "We served 4,200 families" asks the reader to picture something really hard to imagine. A boy reading his first chapter book out loud to his little sister doesn't ask anything of them. If that's the mechanism, the rule changes. The thing to avoid would be abstraction, whether it arrives as a number or as a story. So, a statistic made concrete could sit comfortably next to a person, and a vague story could fail as badly as a vague total.

I lean toward the second explanation, though I'm aware that's partly because my own work is about clarity.

You don't need to settle that debate to improve your letter, though, because the two explanations agree on one thing: a vague total is the weakest line you can write. Before your appeal goes out, read every number in it and ask whether a donor could picture it. "4,200 families" is hard to see, while "enough tutoring for every child on our waitlist" is easier, and one boy reading to his little sister is easier still.

The scientist in me wants to point out that if your mailing list is large enough, you can also find out which approach works for your own donors. Send half of them the letter that opens with a person and the other half the one that opens with a number, and compare what comes back. I do understand that this requires more work than a team with limited resources can knock out. One can only dream.

Sources

1. Small, D. A., Loewenstein, G., & Slovic, P. (2007). Sympathy and callousness: The impact of deliberative thought on donations to identifiable and statistical victims. Organizational Behavior and Human Decision Processes, 102(2), 143-153. https://doi.org/10.1016/j.obhdp.2006.01.005
2. Stanford Graduate School of Business. (2007). Increase charitable donations: Appeal to the heart, not the head. https://www.gsb.stanford.edu/insights/increase-charitable-donations-appeal-heart-not-head
3. Maier, M., Wong, Y. C., & Feldman, G. (2023). Revisiting and rethinking the identifiable victim effect: Replication and extension of Small, Loewenstein, and Slovic (2007). Collabra: Psychology, 9(1). https://doi.org/10.1525/collabra.90203

Featured photo by Justice Fleming Attachie on Unsplash