Cialdini’s penny experiment: more donors, not double the donations

Analysis & application

An invitation can leave people unsure whether their small contribution is worth making. The useful design question is whether a truthful statement of a low acceptable threshold changes participation—not whether a familiar slogan can guarantee growth.

Working with the local American Cancer Society branch, researchers approached 84 adults in a middle-income suburban area. Each condition involved 42 people. One group heard the standard donation request; the other also heard “Even a penny will help.” The added phrase made a small gift acceptable without asking everyone to give exactly one cent. ^[1]

In the control group, 12 of 42 people donated (28.6%); in the added-phrase group, 21 of 42 donated (50%). Among those who gave, the study detected no significant difference in individual gift size. Total funds were $18.55 and $30.34 respectively. Participation, average gift and total revenue are different measures; this was not a doubling of money raised. ^[1]

The authors proposed that legitimizing a small gift made refusal harder to justify. That is an explanation to examine, not proof of every donor’s motive. In product design, reducing uncertainty about a useful first action is a hypothesis derived from the case, not an online conversion result measured by this experiment. ^[1]

A 2008 meta-analysis included 19 studies with 2,730 participants. It identified important differences between face-to-face requests, mailed requests and pledges versus immediate contributions. Only one included experiment used a non-face-to-face request, limiting conclusions about that channel. The review does not establish a universal uplift for modern websites or Apps. ^[2]

Analysis & application

For your own test, define the smallest action that genuinely helps and describe it honestly. Keep a clear option to decline. Compare otherwise equivalent invitations and track completed actions, contribution size, total value and complaints separately. Set the observation period before examining results; do not stop when a favourable number first appears. Treat the outcome as evidence for that audience and task, not permission to promise the same gain elsewhere.