Giving Pattern Analyzer

Pastoral signal from giving data — never surveillance. Aggregates go public; a two-person check-in list stays private, and nothing else.

Category	Operations
Church-health domain	Close the Back Door
Data-privacy tier	Maximum — donor giving is among the most sensitive data a church holds; no LLM on donor data
Mastery-ladder target	Level 5 · Systems
Build stack	MortarStone / Planning Center Giving export → statistical analysis (no AI); aggregate dashboard + 2-person list

---

The problem

Giving trends carry early signals of congregational and personal change, but raw individual giving is among the most sensitive data a church holds — so most XPs avoid it entirely. That's an overcorrection. Pastors often want to know about "any major change from the norm... an indicator of a change in spiritual or financial well-being which might merit pastoral attention." The challenge is getting that signal without building a surveillance machine.

What good looks like

Two distinct layers:

1. A public aggregate layer — cohort retention, recurring-vs-one-time mix, and lapsed-giver anomaly counts — that exposes no individual.
2. A private list of "households worth a pastoral check" that goes to exactly two named people (senior pastor + one elder) and appears nowhere else.

Giving stays "private, but not secret": the church gets pastoral signal, and no one builds a donor-targeting tool out of it.

Market scan

Tool	Fit	Pricing (verify)
MortarStone	Giving analytics; new/lapsed givers, retention	Subscription + setup; free tier for PC churches
Pushpay / ChurchStaq Insights	Giving trends, "at-risk donation" recovery	Enterprise; ~$1,475/mo cited
Planning Center Giving	Giving + added analytics	From 2.15%+$0.30/txn; $14–199/mo
Tithely	Giving + ChMS	Plus $149/mo
Vanco	ACH/text giving; limited analytics	From under $10/mo

The gap: incumbents push toward individual donor targeting and "at-risk donation" recovery — more granular surveillance, framed for revenue. None defaults to aggregate-public plus a 2-person named list and nothing else. The gap is an explicitly anti-surveillance posture.

Data privacy & security — the core of this build

The governing norm: giving should be "private, but not secret." Financial commitment shouldn't be public record, yet the financial secretary may flag families needing pastoral attention.

• Aggregate-only public layer. Cohort/retention/mix charts expose no individual.
• The named list goes to two people only. Senior pastor + one elder; names appear nowhere else, are never emailed in plaintext to a list, and access is logged.
• No LLM on donor data, ever. Anomaly detection here is plain statistics (e.g., a z-score on giving cadence) — not an AI prompt. Donor data never leaves church infrastructure.
• Anti-surveillance discipline. A lapsed-giver list is pastoral only if it never becomes a solicitation or judgment tool. ECFA's 2025 guidance emphasizes donor privacy; GDPR/CCPA can apply. Encrypted storage, locked records.
• Build and demo on synthetic donor data.

How to build it

1. Weekly authenticated pull from MortarStone / Planning Center Giving.
2. Compute aggregate metrics — cohort retention, recurring-vs-one-time mix, lapsed-giver z-score anomalies — with plain statistics. No AI.
3. Publish only the aggregate dashboard to a broad leadership role.
4. Generate a separate, access-logged "households worth a pastoral check" list visible to exactly two named accounts; store names nowhere else.
5. Build and demo on synthetic data.

Rollout plan

• Q1: Agree the governance model first — who sees what, the 2-person rule, the anti-surveillance commitment. Then build the aggregate layer.
• Q2: Add lapsed-giver anomaly detection and the tightly gated check-in list.
• Q3: Review quarterly that the list stays pastoral and the access log is clean.

Effort & cost

• Build: ~1–2 weeks (most of it in the aggregation/anomaly logic and the strict 2-person access gate).
• Run: existing giving platform / MortarStone (free tier possible for Planning Center churches) + modest hosting. The real cost is governance discipline.

Sources

• UMC Discipleship, privacy of giving — https://www.umcdiscipleship.org/resources/privacy-of-financial-giving
• MortarStone analytics — https://mortarstone.com/analytics/
• Who should see giving records — https://churchanswers.com/blog/who-should-be-able-to-see-individual-church-giving-records/
• Ethical donor analytics — https://grassrootsdigital.org/using-donor-analytics-to-grow-your-impact/
• ECFA 2025 developments — https://ministrywatch.com/ecfa-report-lists-tax-and-legal-developments-for-2025/

Honesty flag: Pushpay/MortarStone don't publish standard pricing; figures are from third-party roundups. Treat as estimates.