From the Nominator
"Nanyang Technological University has over 260,000 alumni. In previous years, the University has consistently reached out to our alumni and students to raise awareness and cultivate philanthropic support. However, we experienced an uphill battle in expanding the alumni donors pool, retaining existing alumni donors, increasing ROI of emails, direct mails and our Phonathon resources were stretched to the maximum with diminishing margin of returns.
To improve the fundraising effectiveness, the Data Analytics team was formed within the University Advancement Office (NTU UAO) in Jan 2021. With support from volunteer data scientists from Dell, initiated by an NTU alumna who is also part the Dell's data science team, we developed predictive models using LightGBM, tapping fundraising and alumni data collected in the past. Instead of the previous "spray and pray" approach, the team adopted various data-driven strategies by leveraging ML technology to support the annual giving fundraising process, from AB testing, data collection and classification, model training/testing, and implementation.
We achieved remarkable results from the collaborative effort. In the last year, our Alumni Giving team cut down 40% of outreach to our alumni (with significant reduction in email bombings), while we achieved a more than 70% improvement in response to our appeals. Our alumni annual participation rate increased 28% in Year 2021. The new data-driven process transformation has successfully improved the University's fundraising initiatives, paving way for us to reach greater heights."
From the Judges
The judges were impressed by this entry and especially with the inclusion of the statistics to support their work. It is a standout for the Grand Gold for its uniqueness in supporting the institution, as well as the demonstrated measurable results and the descriptive use of internal and external resources. The future of fundraising will depend on how advancement services use technology, and this entry demonstrates this opportunity well.