Only emailing when you need moneyThe appeal works because of the eleven months around it. A list that hears from you three times a year, all asks, produces falling returns and rising unsubscribes. Send impact reports, send the thing that went wrong, send a thank-you with no link in it. The appeal is a withdrawal from an account you have to pay into.
The statistic next to the storyIt feels like the story plus the scale is stronger than either. It is not — putting an aggregate number alongside an individual's story has been repeatedly found to reduce giving compared with the story alone, apparently because the scale prompts a mode of thinking in which one gift feels futile. Put the numbers on the donation page, not in the appeal.
Not suppressing people who already gaveAsking a donor for money four days after they gave is the fastest way to make a generous person feel like a record in a database. Wire the donation event to an exit condition, test it with a real gift, and send donors a thank-you path instead. This is worth checking every year — it breaks quietly when a payment provider changes.
Vague amounts'Whatever you can give' produces less than three specific amounts tied to three specific outcomes, because most donors have no idea what is useful and are afraid of being wrong. Anchor them.
Overstating the emergencyIf this December is described as the most critical moment in the organisation's history, and so was last December, readers stop believing the fifth one. Reserve genuine crisis language for a genuine crisis. A calm, specific gap is more credible and, across years, raises more.
Giving tax adviceDeadlines are facts and you can state them. Whether a donation reduces a particular person's tax depends on their circumstances and whether they itemise. Say when the deadline is, say gifts are receipted, and leave the rest to their accountant.