Human Services

Food banks, homeless shelters, community services, family and social services, rescue missions.

1. The Retention Thesis

Human-services fundraising is built to lie to you. A food bank runs a matching-gift year-end campaign, raises 40% of its annual individual revenue in six weeks, and the board celebrates. Nobody asks the only question that predicts next year: how many of those donors will still be here in March?

Most won't. The sector's business model quietly depends on renting attention during a crisis or a holiday, then losing most of the people who showed up. You raise a fortune in December and spend January through October re-acquiring the same kind of stranger you already had. Gross-raised hides this because a good year-end can paper over a collapsing donor base for a long time.

"Keep, don't raise" says: the shelter that retains 65% of its donors and grows lifetime value beats the one that raises more this year and bleeds it back out. That is doubly true in human services, where demand is counter-cyclical to easy giving. When a recession hits or a plant closes, your meal counts spike exactly when acquisition gets harder. The only defense against that whipsaw is a donor base you keep across the cycle, anchored by sustainers whose gifts don't wait for a crisis to remember you exist. Retention isn't a nicer way to fundraise. In this sector it's the difference between a stable meal budget and an annual gamble.

2. The Metrics That Matter

Beyond universal sustainer churn (split voluntary vs. involuntary), four numbers govern a human-services shop.

Seasonal-donor conversion rate. Of donors whose first-ever gift arrived in November or December, what share make a second gift in the following 12 months? Formula: (Nov–Dec first-time donors who gave again within 12 mo) / (all Nov–Dec first-time donors). This is the single most diagnostic number in the sector and it is bulletproof: every input is a dated transaction. Expect it to land around 20–25% (illustrative) with no intervention, which tells you exactly how much of your December haul is evaporating.

Second-gift rate (overall). Share of all first-time donors who give a second gift within 12 months, regardless of season. The first-to-second gift is the steepest cliff in the file; a donor who gives twice is several times likelier to give a fifth time. Fully computable from transaction dates.

Sustainer share of individual revenue. Recurring-gift revenue (trailing 12 mo) / total individual revenue (trailing 12 mo). This measures how much of your base is insulated from seasonal and crisis volatility. Under 15% (illustrative) means you're exposed; the plays below exist to move this number.

Mid-level donor retention. Retention rate for donors giving, say, $1,000–$10,000/year (set the band to your file). Report it separately from the mass file. Mid-level is the sector's most under-cultivated segment: too big for the newsletter, too small for the major-gifts officer, so they get treated like $25 donors and leave like $25 donors. Track them apart or you'll never see the leak.

Mark clearly which figures are bulletproof (all four above, since they rest on dated gift records) versus estimated. In-kind and food-drive "donors" often lack clean transaction records, so any retention rate you cite for them is an estimate. Say so.

3. Where Donors Leak

The seasonal-donor leak (voluntary, the big one). A donor gives once in December, prompted by a match, a mailer, or guilt, and never hears from you in a way that earns a second gift. This is a voluntary loss driven by an onboarding failure, not a donor who chose to leave. It is the largest single leak in the sector and the most fixable.

Crisis one-and-done (voluntary). A local disaster or a viral need spikes giving from people with no relationship to your mission. They gave to the moment, not to you. Absent a deliberate bridge from "the crisis" to "the ongoing work," they're gone by the next quarter.

Mid-level neglect (voluntary). A $2,500 donor gets the same three-touch appeal cycle as a $25 donor, feels like an ATM, and drifts to an org that noticed them. Silent, expensive, and invisible unless you segment.

Card failure on sustainers (involuntary, recoverable). Your recurring donors are your best donors, and 20–40% of their annual attrition is failed or expired cards, not a decision to stop giving (illustrative range; measure yours). This is pure recoverable revenue. A shop without dunning (retry logic, pre-expiry updates, account-updater services) is discarding loyal donors over an Amex reissue.

4. Signature Plays

Play 1: The December-donor 90-day welcome journey. Build a fixed sequence that triggers the moment a first-time year-end donor's gift posts. Day 2: a plain, specific thank-you naming what the gift did in local terms ("your $50 is 150 meals at the Elm Street pantry") — no ask. Day 21: one impact story about a neighbor served, told with dignity and consent (see Guardrails). Day 45: an invitation to a warehouse tour or volunteer shift, converting a check-writer into a witness. Day 75: a soft, specific second-ask tied to the same program they first funded. Target: lift seasonal-donor conversion from ~22% to ~35% (illustrative). Ship it as a templated, dated email/mail sequence your CRM fires automatically.

Play 2: The sustainer save-stack. Stand up involuntary-churn defense before you spend a dollar on acquisition. Three layers: a card-account-updater service on your processor; a pre-expiry email at 30 days ("your card ending 4412 expires next month — 20 seconds to keep the meals coming"); and a 4-attempt smart-retry schedule on failed charges (day 1, 3, 7, 14) with a personal call for gifts over $50/month. Every recovered card is a donor you already earned. This typically recovers a large share of "lost" sustainer revenue for near-zero cost.

Play 3: The mid-level "20-minute" portfolio. Pull every donor in your mid band and assign them, by name, to a staff member — 100–150 per person. The commitment: one non-ask personal touch per quarter (a call, a handwritten card, a photo from the floor) plus one tailored ask a year referencing their giving history. No event, no software purchase, just a named human. This is the highest-ROI retention move most shelters have never run, because mid-level lives in the gap between mass marketing and major gifts.

5. Guardrails

Dignity of the people you serve. No "poverty porn." Do not photograph a client's face at their lowest moment to move a donor. The person in your food line is not a fundraising asset. Use imagery and stories that show agency and hope, not degradation. A shelter that exploits its clients to raise money has already failed its mission, and donors increasingly see through it.

Client privacy and consent. Never use a name, photo, or identifying story without explicit, documented, revocable written consent — obtained free of any suggestion that services depend on saying yes. Change identifying details when asked. Domestic-violence and family-services clients carry real safety risk from exposure; a location tag or a recognizable photo can endanger someone. Treat consent as a safety issue, not a paperwork step.

Honesty about outcomes and overhead. Don't inflate "meals served" by counting the same pound of food twice, and don't hide behind a program-ratio number that buries real costs. Human-services donors are local and they talk; a claim that doesn't survive a coffee with a board member will cost you trust you can't re-acquire. If government grants fund most of a program, say so rather than implying donor dollars did it all.

6. The Benchmark Cut

Publish seasonal-donor conversion rate first: of donors whose first gift landed in November or December, the share who gave again within 12 months.

It's bulletproof because every input is a dated, unarguable transaction — no attribution guesswork, no survey, no estimate. It's the number that exposes the sector's core failure (renting December donors and losing them) and it's directly movable by Play 1, so an org can watch the metric respond to the work. Raise this number and everything downstream — sustainer share, lifetime value, revenue stability across the crisis cycle — improves with it. Measure what you keep. Start by measuring how many of December's strangers you turned into March's donors.

Data tier: B (roadmap, not v1-published). This cut is processor-computable but requires cross-gift donor linking (one-time → recurring) and/or reliable campaign tags, so it does not yet clear the "never wrong" bar. In v1 it is private per-org analytics, not a published benchmark. The sector's published number is sector-median sustainer churn, split voluntary/involuntary. See [README](README.md).