Children’s Hospital of Philadelphia


In fall 2021, Children’s Hospital of Philadelphia (CHOP) received a seven-figure gift from a generous donor.

Not so unusual for an experienced major gifts officer. But what was unique about this experience for CHOP was that the prospect had not shown up in any previous donor research and screening.

But this is only one of the remarkable outcomes.

In 2020 the newly hired Director of Philanthropic Technology at CHOP was asked a question: “Do we have the pipeline to hire more gift officers?”

Hospitals seeking philanthropic leads consistently wealth screen families of patients. Due to resources and marketing and patient databases that don’t connect, the very best development departments tend to take wealth screen approved leads and pass them off to major giving teams. The rest are either ignored or are put in regular direct response cultivation, with no real understanding of their connection to the institution, resulting in lack of performance, and the curtailment of the best source of donors possible.

Beyond traditional wealth screening, additional data flags including donor behavior activity to other nonprofits within our SimioAudience co-op, along with number of medical appointments, length of stay, distance traveled, severity of treatment, and behavioral activity sourced from the 250+ million U.S. adults in our constituent data platform, such as social media activity, were combined with key technology to produce a better-informed set of Grateful Family marketing data scores.

The importance of building models based on such a robust set of data is absolutely necessary because you don’t want to simply know if a prospective donor is likely to make a gift. You want to know if they are likely to give to you!


Our approach was to determine if donors who received the mail were motivated to give as a result. To make this determination, we identified the online only cohort and followed their response in the days following when they received a direct mail piece until they received their next direct mail piece. The timing varied from two weeks to a month between mail pieces. Then we identified online giving that was unattributed to any other campaign – including giving directly to the digital efforts for the campaign.


A specific version of the CHOP model was created for direct mail acquisition selection through family patient files. Initial testing has been very positive for the first time out this past year.

  • 85% increase in response rate (improved from 0.7 to 1.3% on the first mailing alone)
  • First average gift increased to $20 – $30



Based on the success of the two initial models, CHOP has commissioned the following propensity models which were just recently delivered:

  • Planned Gifts
  • Event Captain Acquisition
  • Likely to Donate via Digital
  • Likely to Donate via Direct Mail

This is the type of modeling expertise we can bring to Lutheran Social Services to identify donors of higher value and from new sources.