Raphael Epperson, Johannes Diederich and Timo Goeschl ran the cleanest test of this I've seen, published in Management Science (online in 2024, in print in 2025). About 8,700 people decided how much of a $90 stake to give to a relief charity that provides therapeutic food for malnourished children. Some were asked for money, the normal way. Some were asked to fund daily rations at $1 each. Some were asked to fund monthly rations at $30 each.

Same kids. Same food. Same charity. Here's what came in per person asked (the detailed figures are from the authors' open working paper):

Read that twice. The $1 unit pulled in a few more donors and less money. The $30 unit scared off a few donors and still raised about 57% more, because the people who said yes gave a lot more. Small units buy participation. Bigger units buy gift size.

The team didn't stop there. They ran seven versions of the ask to pull apart what a unit actually does. Adding the cost per unit pulled gifts down by roughly $3 to $5, because donors anchored on the price. Framing the gift as units pushed gifts up by roughly $3 to $18, because people could see what their money did. Turning the ask into a grid of choices barely mattered. So a unit is two levers taped together, one pushing up and one pulling down, and the size of the unit decides which one wins.

Notice where the winning unit landed: $30, almost exactly where average giving under the plain money ask already sat. One unit equaled one normal gift. The $1 unit turned that same gift into "twenty-seven of something tiny." That's my read, not a result they tested, but it's the design rule I'd bet on.

The limits: online participants, a 1-in-100 chance the choice was paid out for real, one cause, no relationship with the charity. Your file isn't their sample, which is why it's worth testing.

A second finding points the same way. Christopher Hsee and colleagues found in Psychological Science in 2013 that asking people what they'd give to help one person, before asking what they'd give for the whole group, raised donations in real fundraising campaigns. They call it unit asking. And yes, it replicates. Karlsson and colleagues reran it with 1,039 people in 2020, across children, animals and the environment, and gifts came in about 1.8 times higher.

Now the twist from outside our sector. Hal Hershfield, Stephen Shu and Shlomo Benartzi tested a savings app in Marketing Science in 2020. Describing the same deposit as $5 a day instead of $150 a month quadrupled sign-ups, and the gap between high and low earners disappeared. Same lesson from a different sector: small units are fantastic at getting people in the door. They just aren't built to make the gift bigger.

In Neurogiving I spend a fair bit of time on anchoring, the brain's habit of grabbing the first number it sees and adjusting from there, never quite far enough. A unit price is an anchor wearing a halo. It tells the donor what one unit of helping costs, and they build their gift out of those units. Twenty-five meals feels like a lot. It's twenty-five dollars.

The second piece is scope. Our feelings don't multiply. When Alexander Genevsky and colleagues showed donors an orphan's photo instead of a silhouette, activity in the nucleus accumbens, a small region deep in the brain that anticipates reward, tracked the jump in giving (Journal of Neuroscience, 2013). The brain responds to one. It doesn't respond twenty times harder to twenty. That's why unit asking works: let the donor feel the one, then let the math catch up. Hsee's team describes donors as initially scope insensitive and subsequently scope consistent. Translation: start with the heart, then show the math.

This week's move is a unit audit. It takes about an hour.

Pull every "$X provides Y" line your shop uses, from appeal letters to the gala script, onto one page. Next to each, write the job that ask is supposed to do. Get more people to say yes? Or get the people who say yes to give more?

Then look for mismatches. A $1 unit in a year-end letter to mid-level donors is doing the wrong job. A $500 unit on a first-time donor page probably is too. Small units belong where participation is the goal: acquisition, monthly giving sign-ups, peer-to-peer pages. Bigger units, sized closer to what your donors already give, belong where gift size is the goal.

And if you sit across the table from donors, try the unit ask out loud. Before you describe the campaign's 400 scholarships, ask what one would mean to them. Let them answer. Then talk about the 400.

Research Sources

In Lab Notes Pro

Pro follows the unit into the rest of your program: the study where a child's face made donors feel more and give no more, a field test where the numbers beat the story, and three ways organizations get units wrong (one of them I’ve called The Paper Goat). Plus a unit test you can run on your year-end appeal.