The World Wildlife Fund (WWF) sends ~1.3 Million symbolic adoption catalogs each holiday season to inspire donations for conservation. These catalogs are mailed to a targeted set of existing donors (an “in-house file”) in three drops between October and December.

This project challenges students to use two years of rich data to optimize the drop timing and frequency for future campaigns. Students will explore how donor engagement varies depending on the timing and number of catalogs received per household and develop a data-driven mailing strategy for FY26.

Seasonality plays a major role in the success of the WWFGifts catalog.  Donor activity spikes after Thanksgiving, especially around Black Friday, Cyber Monday, Giving Tuesday, and the final guaranteed shipping window before the holidays (usually mid December). Students must balance campaign effectiveness, timing, cost, and donor experience.

Students will receive a two-year SAS dataset (model training) and a separate universe scoring dataset, each containing:

• Past catalog audiences (and non-catalog audiences) from FY25 and FY24, with information regarding the number of catalogs received and which drop they received.
• Giving outcomes: response and amount
• Household and demographic data (age, income, household type, hobbies,  home value, census info, etc.)
• 12- and 24-month donation history from other nonprofits and catalogs
• Campaign cost info.
• WWF-specific donation behavior patterns and giving history.

Students are expected to:

1. Analyze donor response by drop timing, region, and donor segment
2. Train predictive models to estimate response likelihood and expected gift amount
3. Recommend a drop strategy that maximizes net income (accounting for catalog costs)
4. Develop a solution for:

• a constrained total mail volume (~1.3M catalogs across ~933K households)
• an unconstrained total mail volume that maximizes net income

Deliverables

1. A presentation summarizing:

• Analytical findings
• Model results
• Key donor behavior trends
• Strategic recommendations

2.  A FY26 drop allocation plan:

• How many drops each segment should receive
• When each drop should occur
• Volume plan that aligns with budget and prior year totals

3. Supporting materials:

• SAS based analysis code
• Visualizations (e.g., projected response curves, revenue by timing, percentile plots, etc.)