AI in Nonprofit Fundraising and Donor Management Quiz

Answers and explanations appear as you take the quiz.

1. What does RFM stand for in donor segmentation?
   • Revenue, forecast, margin
   • Recency, frequency, monetary value
   • Reach, focus, messaging
   • Retention, fundraising, membership
2. Why is wealth-screening capacity alone not enough to identify good major-gift prospects?
   • Wealth data is always illegal to use
   • Wealthy people never give to charity
   • Capacity data is too expensive for any charity
   • Capacity is not inclination; affinity for the cause also matters
3. Why did the UK Information Commissioner's Office fine charities such as the RSPCA and British Heart Foundation?
   • Partly for wealth screening donors without telling them
   • For refusing online donations
   • For using RFM segmentation
   • For sending too few thank-you letters
4. What did the 2020 Blackbaud incident demonstrate?
   • That AI models cannot predict lapse
   • That printed appeals outperform email
   • That concentrated donor data creates concentrated risk, after a ransomware attack exposed many clients' data
   • That wealth screening is always accurate
5. Which is an example of data leakage in a propensity model?
   • Using gift history from before the mail date
   • Including a 'last gift date' field updated after the outcome
   • Using logistic regression
   • Testing on a holdout group
6. What does a lift or gain chart show?
   • How fast the CRM loads
   • The number of emails sent per day
   • Staff hours saved by automation
   • How much of eventual income top-scored donors produced compared with random selection
7. Why use a holdout group of high-scoring donors who do not receive the appeal?
   • Because a high score may identify people who would have given anyway
   • To save postage only
   • Because regulators require it
   • To train the language model
8. According to the guide, what do well-used fundraising models often do?
   • Always increase the number of asks
   • Replace gift officers entirely
   • Identify donors to contact less often or through a different channel
   • Remove the need for clean data