How to Make a Fundraising Video When You Can’t Film the People You Serve
Most advice about nonprofit video assumes you can film the people you help. Find the right person, sit them down, let them tell their story. It is good advice. It…
Every nonprofit has numbers it is proud of. Meals served, students enrolled, patients seen, years in operation. When it comes time to make a fundraising video, those numbers usually go in early, because they feel like the proof. There is a well-known line of research suggesting that this is exactly the wrong place for them. […]
Every nonprofit has numbers it is proud of. Meals served, students enrolled, patients seen, years in operation. When it comes time to make a fundraising video, those numbers usually go in early, because they feel like the proof.
There is a well-known line of research suggesting that this is exactly the wrong place for them. It is worth understanding what that research found, what it did not find, and what has happened to it since, because it is the most useful thing anyone has learned about how fundraising films work.
In 2007, Deborah Small of the University of Pennsylvania, George Loewenstein of Carnegie Mellon, and Paul Slovic of Decision Research published a paper in Organizational Behavior and Human Decision Processes called “Sympathy and Callousness.” It reported four experiments. All of them worked the same way: people on a university campus were paid five dollars in one-dollar bills for filling out a survey, then handed a letter asking them to donate any of it to Save the Children, and left alone with an envelope.
What changed between groups was the letter.
Some people read statistics: food shortages affecting more than three million children in Malawi, four million Angolans displaced, eleven million Ethiopians needing food aid. Others read about one child, a seven-year-old girl from Mali named Rokia, with her photograph, and were told their money would go to her.
In the third experiment, which is the one most often quoted, 159 people were split three ways. Those who read the statistics gave an average of $1.14. Those who read about Rokia gave $2.38. And those who read about Rokia and the statistics, on the same page, gave $1.43.
Adding the numbers to the story cut giving by roughly 40 percent. The group that had the most information gave almost exactly what the statistics-only group gave. The story’s effect was, in the authors’ words, dampened.
The paper’s explanation rests on a distinction psychologists had been developing for years: people have two ways of responding to what is in front of them. One is fast, emotional, and triggered by something specific and vivid. The other is slower, deliberate, and triggered by anything abstract. A single child with a name and a face sets off the first. A table of casualty figures sets off the second.
The problem is that the second mode does not just sit alongside the first. It overrides it. Once people are thinking analytically, they discount the sympathy they felt a moment earlier, and analytical thinking does not generate sympathy of its own to replace it.
The other experiments in the paper bear this out. In the first, simply telling people about the research (that humans give more to identifiable victims than to statistical ones) cut donations to Rokia from $2.83 to $1.26, while doing nothing for donations to the statistical victims. In the fourth, people who had just answered five arithmetic questions before reading about Rokia gave $1.54; people who had just answered five questions about their feelings gave $2.34. The arithmetic did not change how much anyone gave to the statistics. It only took the story away.
The paper’s abstract puts it in one sentence: when thinking deliberatively, people discount sympathy toward identifiable victims but fail to generate sympathy toward statistical victims.
This finding has been cited thousands of times and it has also been challenged, and any nonprofit relying on it should know both.
A 2016 meta-analysis by Lee and Feeley found the identifiable victim effect to be real but smaller than the early papers suggested. In 2023, a pre-registered replication of the first and third experiments, published in Collabra: Psychology by Maier, Wong, and Feldman, found evidence of publication bias in the literature and reported weaker effects than the 2007 original. The direction held. The size did not.
The honest summary is this: the original numbers are from one campus, one charity, and small groups of people with five dollars in their hands. They should not be treated as a law of donor behavior. But the underlying pattern, that abstract information moves people into a frame of mind where they give less to a specific person and no more to the general cause, has held up across a lot of attempts to knock it down. For someone deciding how to build a three-minute film, that is enough to act on.
The research does not say statistics are bad. It says that put next to a story, at the same time, they compete with it and win. Which leads to a set of practical rules.
Do not open with the numbers. The first ninety seconds of a fundraising film have one job: get the audience to care about a specific person. Anything that invites the viewer to evaluate rather than feel, during that window, works against the film.
Let the story finish before the scale arrives. The numbers prove that the person on screen is not an exception. That is a legitimate and important thing to establish. It belongs after the emotional peak, not before it, and it belongs in the ask rather than the film when there is a live ask to make.
Treat the ask as the analytical moment. In a gala room, the film ends and a person stands up. That person can carry the scale, the budget, the match, the goal. The audience has already decided how they feel. Now they are deciding what to do, and numbers help with that decision in a way they never help with the first one.
One person, not six programs. The study’s entire effect comes from one named child versus a population. A film that introduces six beneficiaries to represent six programs has quietly turned its subjects back into statistics.
Do not explain the technique to the audience. The first experiment is the strangest and most useful. Telling people how the effect works switched it off. A film that says “we could show you numbers, but we would rather show you one person” has done exactly that.
This finding is the foundation of our Story + Heart = Action framework, and it is why every film we make for a nonprofit follows one real person and settles the story before anyone turns on a camera. It is also why our clients’ statistics still appear in their films. They appear late, once the audience already cares, doing the job they are good at: proving the problem is bigger than one person.
The method is on our nonprofit video production page. The full argument, with the research walked through in more depth, is in Stories That Move Mountains, our guide to fundraising video for nonprofits. Pay what you can afford, including nothing.
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. Full text
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). Article
Lee, S., & Feeley, T. H. (2016). The identifiable victim effect: A meta-analytic review. Social Influence, 11(3), 199–215.