How Save The Children is using AI donor predictions to build a better world for every child
In the challenging fundraising environment of 2021, Save The Children Australia were looking for ways to improve their appeals targeting to increase net revenue.
So they ran a real-world experiment with their Christmas 2021 appeal to see if Dataro’s AI-powered donor predictions would generate more gifts with less waste when compared to their traditional segmentation approach.
The results showed that the Dataro selections reliably generated more gifts than the RFV selections, with an 18% increase in response rate and more revenue received from less mail sent. In their first 6 months using Dataro, Save The Children saved $16k across three appeals through better donor targeting.
- 18% increase in response rate with Dataro
- >$12k in additional net revenue with Dataro
- 3.6% increase in Dataro campaign net revenue
- 14.5% decrease in mailing volume with Dataro
What we tested
Using their Christmas appeal as an experiment, Save The Children wanted to test which appeal targeting method would generate better returns: Dataro’s AI-powered donor predictions versus their traditional RFV (recency, frequency, value) segmentation approach.
Setting up the test
Save the Children used their traditional RFV segmentation strategy to select donors for the appeal.
Dataro’s direct mail appeal propensity scores and ranks were then integrated with their CRM and used to select a list of donors with the highest likelihood to give.
The two lists were combined so we could identify:
- donors selected by both Save the Children and Dataro = 19,854 donors
- donors selected by Save the Children only = 4,813 donors (that Dataro predicted were extremely unlikely to give)
- donors selected by Dataro only = 1,244 donors (who were likely to give but had been missed by RFV)
By sending the campaign to all donors selected using either method, we were able to compare the results from the Dataro version of the campaign versus the Save The Children’s version.
Analysing the Results
Dataro’s campaign raised more than the charity’s version, generating 35 more gifts in total from 3,569 fewer direct mail letters sent. The net return of Dataro’s campaign was +$12,000 better than the charity’s campaign.
Had Save The Children solely used Dataro’s predictive scores to make their appeal selections they could have reduced the mail file by 14.5% and saved +$8,000.
The additional 1,244 donors added by Dataro gave 59 gifts at a reasonable 4.75% response rate, netting roughly $,3000 in extra revenue. In comparison, the extra donors that Dataro recommended removing from the campaign (selected by the charity using RFV), performed very poorly, with a response rate of just 0.5%. This cohort actually cost the charity more to mail than they gave in gifts, costing >$8,000.
Confident in the Christmas appeal results, Save the Children ran subsequent experiments for their next two appeals using the Dataro donor scores and ranks as the default appeal targeting method to test reliability of the new selection method.
The results of these campaigns built on the strong initial appeal experiments and confirmed that using Dataro’s AI scores and ranks as the sole selection criteria was the best approach.
If you are using a segmentation approach for your direct mail appeals, you are missing out on gifts from donors in your database and wasting costs targeting donors who are highly unlikely to give. The time spent building your data lists, usually with specialist support, can also be lengthy and often starts months in advance of the appeal going live.
Nonprofits using Dataro’s machine learning to identify the donors most likely to give to their direct mail appeal are able to target better and as a result:
- Find more gifts from your existing donors
- Raise more money
- Reduce mail volume and costs
- Increase appeal ROI
- Save time and costs making their data selections inhouse and with a click of a button
With prediction-based targeting Dataro will help you raise more from your existing donors, with less waste and saving you time and effort.
What they said
“Over the last 6 months we have been working closely with the team at Dataro to support our appeal data selection with some really positive results in a challenging fundraising environment. Starting cautiously, wanting to minimise any potential risk to income, we have quickly gained confidence and tactically increased our reliance on the Dataro ranking of our database. This has helped us to identify a smaller cohort of supporters with a higher propensity to give and a better ROI for our appeals. ”
Supporter Growth Manager
More Case Studies
Testing Dataro across Australia for UNHCR direct mail appeals
Australia for UNHCR is the UN Refugee Agency’s national partner in Australia, raising funds and awareness to support displaced people across the globe. Direct mail appeals form a core part of Australia for UNHCR’s fundraising calendar. So in September 2021 the team embarked on a partnership with DataroRead more
How UNICEF Australia use AI to raise more for children in need
UNICEF is a leading global non-profit working to improve the lives of children in need. As an organization that is 100% donor-funded, UNICEF Australia relies on the generosity of individual donors to fund their work. In 2020 UNICEF Australia partnered with Dataro to improve their direct mail appealRead more
Erskine donors leave a lasting legacy for Scottish veterans
Charitable income from legacies has quadrupled over the past 30 years and was forecast to top £4 billion per annum by the end of last year, experts say. It is little wonder that legacies (Gifts in Will) have become a key focus for charities that have one eyeRead more
Red Cross uses AI to find mid-value donors that help support communities in times of need
Australian Red Cross have a bold objective to increase their mid-value program income from $7 million to $9 million by FY26/27 to help them improve lives and build resilient communities in times of need. In 2021 the charity launched a new mid-value strategy and communications plan, including aRead more