How One Small Church Used AI to Recover 40 Lapsed Donors in 90 Days

How One Small Church Used AI to Recover 40 Lapsed Donors in 90 Days
A 320-member church in the Midwest recovered 40 lapsed donors in 90 days by letting AI flag quiet givers early, drafting personal outreach for the pastor to send, and routing follow-up to the right staff member before the relationship went cold. No new hires. No new campaigns. Same congregation, same budget, same team of four.
Here is exactly what happened, in the order it happened, so you can judge whether the approach fits your church.
The situation: strong Sundays, soft giving
By late winter, the lead pastor at Grace Fellowship (name changed at their request) noticed the pattern most small churches see and cannot quite explain. The room looked full. Volunteers showed up. But the monthly giving report kept sliding, roughly 11 percent under the prior year, and nobody could point to a single reason.
The bookkeeper had a spreadsheet of givers, but no time to cross-reference it against attendance, small group check-ins, or the last time someone opened a church email. The pastor knew a handful of families had gone through hard seasons. He suspected others had drifted. He could not tell you which was which without spending a week on it, and he did not have a week.
That is the honest problem behind most lapsed giving in a small church: the information exists, scattered across Planning Center, a giving platform, a Mailchimp list, and the pastor's memory. Nobody has the hours to stitch it together.
What they tried
The church connected ChurchAI to their existing Planning Center account. No migration, no ripping out systems. Within a day, the platform had pulled two years of giving history, attendance, and communication engagement into one view.
Three things happened in the first week:
- A ranked list of at-risk givers appeared. Ninety-seven households had given consistently in the prior year and had either stopped or dropped their gift by more than 40 percent. The list was sorted by giving history, tenure, and recent engagement, so the pastor could see the highest-signal names first.
- Each name came with context. Last gift date, last service attended, last email opened, small group status. Instead of guessing, the pastor could see that the Hendersons had not attended since a family move in February, while another family was still showing up weekly but had quietly stopped giving in March.
- Drafted outreach, ready to edit. For each household, ChurchAI generated a personal message the pastor could adjust and send from his own phone. Not a mass email. A short, specific note referencing the last time they had connected.
What actually moved the number
The pastor blocked 45 minutes on Tuesday and Thursday mornings for three months. He worked the list top down, sending 8 to 12 messages per session, editing each draft in about a minute. His associate handled the ones flagged as pastoral care rather than giving conversations.
By day 90:
- 40 households had resumed giving at or above their prior level
- 12 more had responded with a life update but had not yet given
- 9 had asked to be removed from communication, which the church treated as useful information, not failure
- The remaining 36 got a second, different touchpoint queued for the next quarter
Recovered monthly giving from those 40 households came to roughly $8,400 a month, which annualized covers the platform many times over. But the pastor was clearer about the softer result: he stopped feeling like he was flying blind, and two of the recovered families ended up leading a new serving team by summer.
The reusable takeaway
Small churches do not lose givers because pastors do not care. They lose them because the signals are buried in five systems and nobody has the hours to look. Fixing that does not require a bigger team. It requires one place that watches the patterns and tells you who to call, in what order, with what to say.
If that sounds like the gap in your week, ChurchAI was built by pastors who lost people they should not have, specifically to close it. You can book a demo to see what your own giving data would surface, or read more stories on the ChurchAI blog. Questions about how the integration works with Planning Center, Breeze, or CCB are answered on the support page.
The people you have already reached are worth reaching again. The hard part is knowing which ones, and when. That part, at least, is solvable now.