Your donor database knows everything that happened. It does not know what is about to happen.
Every gift recorded. Every campaign response logged. Every event attendance tracked. The CRM is a perfect archive of donor history. It is completely blind to donor future.
That distinction is where the $847,000 Leak lives.
The signal that a donor is about to leave does not appear in a gift record. It appears in the 60 days before the gift does not arrive. In the email open rate that dropped from 58% to 12%. In the event invite ignored for the second time in a row. In the survey response that was technically completed but tonally distant. None of those signals exist in your CRM as a risk flag. By the time the lapse shows up in a report, the window to prevent it has been closed for months.
What Manual Monitoring Actually Costs
A mid size nonprofit with 1,200 donors cannot meaningfully monitor all 1,200 relationships manually. The math does not work.
A development director managing 200 relationships has bandwidth for roughly 15 to 20 meaningful touchpoints per week. That is one touchpoint per relationship every two to three months, on a good week, with no administrative interruptions. For the remaining 1,000 donors in the file, monitoring is effectively nonexistent. Those donors are either retained by inertia or lost to silence.
The cost of that silence is measurable. A 10% improvement in retention is worth $847,000 over five years for the average organization. The donors generating that improvement are already in your database. They are signaling their departure right now. Nobody is watching.
That is not a staffing failure. It is a structural one. And no amount of additional headcount solves a monitoring problem that requires continuous, simultaneous analysis of 1,200 behavioral profiles.
What the Donor Risk Alert Agent Monitors
Donor Risk Alert
Tracks 20 plus engagement signals per donor continuously. Not weekly. Not monthly. Continuously. Each signal feeds a real time engagement score between 0 and 100 that updates the moment new behavior arrives.
The Donor Risk Alert agent tracks 20 plus engagement signals per donor continuously. Not weekly. Not monthly. Continuously.
Email behavior: open rates, click through rates, response rates, and patterns in subject line engagement over time. Event behavior: registration history, attendance history, and RSVP response times relative to historical patterns. Giving behavior: gift frequency, gift amount trends, and timing patterns compared to the donor's historical cadence. Communication behavior: survey response sentiment, reply rates, and language signals where available.
Each signal feeds a real time engagement score between 0 and 100. The score updates continuously. When a donor's trajectory drops below a threshold correlated with lapse risk, an alert fires.
The alert does not ask a development associate to review a spreadsheet. It surfaces a specific donor, a specific risk level, a specific context, and a specific recommended action.
The Jennifer Martinez Example
Jennifer Martinez, three year donor. Email open rate dropped from 58% to 12%. Two declined event invitations. 89 days of silence. Lapse probability: 78%.
Jennifer Martinez was a three year donor with strong prior engagement. Over 90 days, her engagement score dropped from 72 to 38. Her email open rate fell from 58% to 12%. She declined two event invitations. She accumulated 89 days of silence. Her lapse probability reached 78%.
Under a manual system, Jennifer would have required a development associate to pull her record, notice the pattern, and flag it. That sequence requires someone to be looking for Jennifer specifically. Nobody was.
Under the Donor Risk Alert system, her trajectory triggered an alert automatically. The director received full context and a recommended intervention. A personal call was made. Jennifer attended the spring showcase and increased her gift the following month. The difference was not the quality of the outreach. It was the fact that the outreach happened at all, at the right moment, because the system surfaced it.
What the Activity Tracker Adds
Donor Risk Alert surfaces who needs attention. Activity Tracker provides the context for what to do.
Every interaction logged in connected systems appears automatically in the donor's Activity Tracker timeline: emails opened, calls completed, events attended, gifts received, personal notes added. The director opens the alert, pulls the timeline, and sees the full relationship history in seconds.
Last meaningful interaction: four months ago at the annual luncheon. Last gift: smaller than the previous two years. Program area of deepest engagement: STEM workforce development. Recommended next action: personal call referencing program impact, with an invitation to the spring impact event. That is a relationship strategy synthesized from three years of data, available in under 30 seconds, requiring no manual research.
The Operational Shift
Before StewardWise AI, donor monitoring was reactive. After, it is predictive.
Reactive Monitoring
- Quarterly reports identify donors who already lapsed
- Subjective judgment about who might be at risk
- Outreach built on incomplete signals
- 1,000 of 1,200 donors effectively unmonitored
- Save rate below 15% because intervention arrives too late
Predictive Monitoring
- Continuous tracking of 20 plus signals per donor
- Specific risks surfaced with specific context
- Recommended interventions ready to personalize
- All 1,200 donors monitored simultaneously
- Save rate of 70 to 85% because the team acts inside the window
The development team stops generating reports to identify who might be at risk. The 1,200 donors who were impossible to monitor manually are now monitored continuously. The ones who need attention get it at the right moment. The leak slows.
What the Early Warning System Delivers
- 20 plus engagement signals tracked per donor continuously
- Real time engagement scores from 0 to 100, updated as new behavior arrives
- Automatic threshold alerts with churn probability and recommended action
- 90 day early warning window surfaced for every at risk donor
- 70 to 85% save rate within the window versus less than 15% post lapse
- Activity Tracker auto logs every interaction across connected systems
- 1,200 donors monitored simultaneously with no additional staff capacity required
Your CRM was never designed to predict the future. StewardWise AI was.
Build the Early Warning System
See exactly which agents stop the $847,000 leak and how they fit into your current operations.
