Writing a donation request that works: what research has measured

A donation request is often a letter or email written on a weeknight by a parent volunteer, based on last year’s. There is plenty of advice going around: tell a story, mention a matching gift, suggest an amount. Some of it has been tested on real donors, with randomly assigned groups. The results are sometimes clear, and sometimes more nuanced than the rule people drew from them.

Here are five of those experiments, what they measured and their limits. After that you will find a sample letter for a school or club, where each sentence is tied to the finding it applies.

Five measured results

Announcing that a gift will be matched

In an experiment published in 2007, Dean Karlan and John List worked with an American civil liberties organization. The organization wrote to 50,083 people who had given to it before. One-third received the usual letter. The other two-thirds received the same letter with a matching offer: an anonymous donor would add $1, $2 or $3 for every dollar given.

Simply announcing a match increased the likelihood of giving by 22% and the money raised per letter by 19%. However, a more generous match added nothing: “$3 for $1” did no better than “$1 for $1.” The effect also varied a lot by region. It came almost entirely from states that had voted Republican in 2004, where money raised per letter rose by 55%. Elsewhere, the effect was weak.

Another experiment adds nuance to this result. With the British Columbia chapter of the Sierra Club of Canada, Daniel Rondeau and John List compared a match with a lead gift already pledged, of the same amount. The announced lead gift brought in $1,620, compared with $1,235 for the match. The match did not do significantly better than letters with no offer at all.

Takeaway: if a sponsor wants to help, a simple matching offer (1 for 1) can help. There is no need to negotiate a higher ratio. An announced lead gift seems at least as effective.

Telling one person’s story

This is the most cited study on the subject. In 2007, Deborah Small, George Loewenstein and Paul Slovic gave $5 to participants at a Pennsylvania university, followed by a letter from Save the Children. In one of their experiments, 159 people received one of three letters: a portrait of Rokia, a 7-year-old girl in Mali; statistics on hunger in Africa; or Rokia’s portrait along with the statistics. The average gift was $2.38 for Rokia’s portrait alone, compared with $1.14 and $1.43 for the other two letters. The authors found no significant difference between those last two: adding statistics to the story erased the portrait’s advantage.

This result is now debated. In 2024, Maximilian Maier, Yik Chun Wong and Gilad Feldman redid the experiment with a pre-registered protocol. They found no identifiable victim effect. Their reanalysis of an earlier meta-analysis also suggests that studies with no result were published less often. Their own limit: their participants reported hypothetical gifts, with no real money.

Takeaway: a concrete story does no harm, but it guarantees nothing. Don’t pile a heap of numbers next to it.

Saying exactly what the gift will pay for

In 2013, Cynthia Cryder, George Loewenstein and Richard Scheines proposed another reading of the previous effect: what drives people to give may mainly be concrete detail, whatever form it takes. In their first experiment, 119 passersby in a city in the northeastern United States received $2 and could give part of it to Oxfam. Half read a general description (a large international aid organization). The other half read a specific example: Oxfam helps villages in West Africa get access to clean water. The average gift was $0.88 with the specific example, compared with $0.48 with the general description.

A second experiment with 94 people gave the same result with an organization that distributes mosquito nets ($0.74 compared with $0.40). According to the authors, the specific example gives people the sense that their gift will have a real effect.

The amounts are small, and the participants were passersby paid for five minutes of their time. Still, the direction of the result is easy to apply.

Takeaway: “Your gift pays for part of the bus to the May tournament” works better than “Your gift supports our activities.”

Mentioning what others gave

Jen Shang and Rachel Croson ran their experiment during on-air pledge drives at an American public radio station, in June and September 2003. The researchers answered the phones as volunteers. Before asking “How much would you like to give today?”, they sometimes read a sentence like: “Another member gave $75” (or $180, or $300). Other callers heard nothing.

Mentioning $300, an amount that only 10% of donors usually exceeded, increased gifts by 12% on average, or $13 more. The effect applied to new members, not to those renewing. A year later, people who had heard an amount were also more likely to renew (23% to 32%, compared with 12% in the control group).

Context matters: these were calls from people who had already decided to give and were only asking how much.

Takeaway: a true sentence about what other families gave can help, especially with first-time donors. The amount you cite must be real, and high enough without being exceptional.

Offering a default amount

On an online form, you can pre-select an amount. Steffen Altmann, Armin Falk and their colleagues tested this on betterplace.org, Germany’s largest online donation platform. About 680,000 visitors were randomly shown a pre-selected amount of €10, €20 or €50, or an empty field.

The pre-selected amount strongly influenced choices: in each group, the most frequent gift was exactly that amount. Yet total giving did not change. The suggested amount pushed some people to give more and others to give less than they would have. At €50, more visitors left without giving anything, which cancelled out the gain.

In the Karlan and List study cited above, the example amount printed in the letter (equal to 1, 1.25 or 1.5 times the person’s largest past gift) also had no notable effect.

Takeaway: a suggested amount steers gifts, but it does not raise the total on its own. Avoid a pre-selected amount that is too high. Offering three choices and an open field is a reasonable way to leave the door open to everyone.

What the research doesn’t say

• The settings are mostly American, with one case in British Columbia and one in Germany. None of these studies looks at a school or sports club in Quebec or elsewhere in Canada.
• Response rates are low everywhere. In the Karlan and List study, about 2% of people gave. A 22% increase on 2% is real, but it still means a small number of extra gifts.
• These techniques were tested one at a time. Nobody has measured what a letter combining all of them produces. There is no reason to believe the effects simply add up.
• Some results have held up poorly in newer studies, such as the identifiable victim effect. The most solid ones here come from large field experiments with real gifts: the match, the announced lead gift, the pre-selected amount.

Annotated sample letter

The letter below is a fictional example for an imaginary soccer club. The numbers refer to the notes that follow.

Subject: We’re $1,400 short for the bus to the provincial tournament

Hello,

Our 18 U13 players have qualified for the provincial tournament in June. To get there, the club has to rent a bus and book two nights of lodging. (1)

The total budget is $4,000. Thanks to two local businesses, $2,600 has already been raised. (2)

A parent on the team has committed to doubling every gift received by May 15, up to $500. (3)

In concrete terms, $40 pays for one player’s round-trip transportation. $120 pays for her share of the lodging. (4)

Many families have chosen to give $50. (5)

You can give $25, $50, $100 or the amount of your choice, online or by handing your gift to the coach. (6)

We’ll send you news from the tournament, with photos of the team.

Thank you for your support,

The fundraising committee

Notes:

1. A specific, time-limited project. This is the principle tested by Cryder, Loewenstein and Scheines: a concrete detail rather than a general description.
2. An announced lead gift. The two experiments on this topic (List and Lucking-Reiley, Rondeau and List) show that saying a large share is already raised helps. Putting the amount left in the email subject line also puts the focus on how close the goal is.
3. A 1-for-1 match, with a deadline. Karlan and List: the presence of the match matters, not its ratio. Only announce a match if the commitment is real and in writing.
4. What the gift pays for. Amounts tied to real expenses. Do the math with your actual costs.
5. Information about other donors. Shang and Croson. This sentence must be true. If this is your first campaign, remove it.
6. Suggested amounts and an open choice. Altmann and colleagues: a suggested amount steers gifts without guaranteeing a higher total. The open choice leaves the door open to small gifts.

You’ll notice there is no portrait of any one player. If you add one, first get the parents’ written consent, and keep it short.

A note on tax receipts

Many donors ask whether they will get a receipt. According to the Canada Revenue Agency, as of October 2026, only qualified donees, such as registered charities, can issue official donation receipts: “Unlike non-profit organizations, qualified donees can issue official donation receipts”. A club or parent council that is not registered therefore cannot issue them. Check your organization’s status before writing anything about receipts in your letter.

Before sending your request

To check	Done
The letter names a specific project and a date	☐
The amount already raised and the amount left are shown	☐
The match, if there is one, is confirmed in writing and has a cap	☐
At least one sentence says what a gift of a given amount pays for	☐
Any mention of what others gave is true	☐
Suggested amounts stay modest and an open choice is offered	☐
What you say about receipts matches the organization’s actual status	☐
The letter fits on one page or one phone screen	☐

In short

• Announcing a match increased giving in an experiment with 50,000 donors, but a more generous ratio added nothing.
• Saying exactly what the gift would pay for almost doubled small gifts in two experiments.
• The “one person’s story” effect is now contested: a concrete story can help, with no guarantee.
• A true sentence about what others gave raised new donors’ gifts by 12% at a public radio station.
• A suggested amount steers choices, but does not raise the total on its own. Keep it reasonable and offer an open choice.