Giving is Caring:
Understanding Donation Behavior through Email
Yelena Mejova, Yahoo Labs
Qatar Computing Research
Institute
Presenter: Mobile HCI Lab Andy
Abstract
In this paper, we describe a comprehensive large-scale data-driven study of donation behavior.
-
Intro
- RELATED WORK
- Observing donations in email
- Tables
- Contributions
Intro
$211.77 billion online donation in 2010
The 2008 US presidential campaign
4 respects: (1) demographics, (2) topi- cal interest, (3) social influence, and (4) external influence
RELATED WORK
Individual capacity and inclination.
personal reputation, “empathetic joy”,Name-Letter Effect
Interest
To orgs, beneficiaries, performance and accountability,
RELATED WORK
Social influence
socially reward donors and punish non-donors,
positive in-group identity -> individual self-esteem
email: increase the salience of social identity
selective incentives, tax incentivesAfter the 2010 earthquake, SMS donation over $7M in micro donations.
1/2 page about crowdfunding.
Observing donations in email
two-month period of July 19 – September 19, 2012
all user identifiers (= email addresses) were
replaced by random numbers (“hashes”).
480 charities
100,000 users.
data anonymized
Spam Filtering
gathering user action statistics
Tables
Tables
average accuracy of these regular ex- pressions to be 86.2%
Tables
July 31: “Photo going around on Facebook” (Obama),
Aug 11: Paul Ryan is Republican VP candidate
Aug 18: T-shirt promotion “I built this” (Romney),
Aug 27-30: Republican National Convention
Sep 4-6: Democratic National Convention
Findings
-
Major donations coincide with major events, especially in political.
-
Explicit solicitations for donations spur more donation activity than newsletters or updates.
-
Solicitations connected to real-world events are more ef- fective than general-purpose ones.
The effect of (number of friends who are donors to the charity) is by far the greatest: for each friend who donates to a charity, the likelihood for the user to also donate to that charity goes up by 24%.
Contribution
Our paper provides a unique exploration of the donation behavior of hundreds of thousands of user.
We take a data-driven approach in which we examine the incoming and outgoing emails for these users over the span of two months.
*Cons
illustrated by some case of public broadcasting orgs
offline donations
Future work
What messages and solicitations are the most convincing to donors? What strategies can charities employ to find new potential donors?