Costs
Part of Donor retention: the parts worth your attention
Building a donor retention benchmark: numbers from your own program
Build a donor retention benchmark from your own program data with defined cohorts, real formulas, a review cadence, and a defensible baseline.
What to take away
- Write four fields before you calculatepopulation, period, measure, decision.
- The core formula is R / (S - N)prior-period donors who gave again, divided by prior-period donors eligible to repeat.
- Report repeat-donor, value and first-year retention side by side, on one population and one window.
- Set the threshold from your own history and record who set it.
- Attach the working, the gaps and a review date, or the number becomes folklore.
Start with the four fields
A benchmark is a figure with a population, a period, a measure and a decision attached. Drop one field and you have an anecdote nobody can restate.
Write the four fields down before you open the database. The order matters: the decision field tells you which population to pull.
Four fields of a benchmark
- Populationwho is included, who is excluded
- Periodstart, end, maturity window
- Measuredefinition, numerator, denominator
- Decisionthreshold and a named owner
| Field | What to record | Acceptance test |
|---|---|---|
| Population | Who is counted, and who is excluded | A colleague rebuilds the list from the description alone |
| Period | Both windows, with start and end dates | Both windows run the same number of days |
| Measure | The retention figure, named as a formula | Substituting other figures still gives a computable number |
| Decision | The action and its threshold, such as a stewardship review below 55% | The figure changes what someone does next quarter |
The same discipline sits behind setting a donor acquisition benchmark, where arrivals mid-year make the population harder to pin down.
The retention formula, written out
Donor retention is repeat donors divided by donors eligible to repeat, over a fixed window.
S = donors who gave in the prior period R = donors from S who gave again in the current period N = first-time donors in the prior period, not yet eligible to repeat Donor retention = R / (S - N)
Write the exclusions next to the result.
Two figures get confused here. Overall retention counts every donor who gave again. Repeat-donor retention counts only donors with a prior gift. Mixing them is the most common reason two teams disagree about one program.
Value retention is separate. Let V1 be prior-period revenue from S and V2 the current revenue from the same group. Value retention is V2 / V1. A program can hold donors and still lose revenue.
Retention formulas, written out
- R / (S - N)donor retention
- V2 / V1value retention
- Sdonors who gave in prior period
- Rprior donors who gave again
Seven figures worth calculating
| Figure | Formula | What it answers |
|---|---|---|
| Repeat-donor retention | R / (S - N) | Do donors with a giving history keep giving? |
| Overall retention | repeat donors / all prior-period donors | The headline figure sector studies publish |
| Value retention | V2 / V1 | Does revenue from the same donors hold? |
| First-year retention | year-two repeat donors / that first-gift cohort | Your long-run file size |
| Recapture rate | lapsed donors who give again / donors lapsed 12 to 24 months | Does win-back mail earn its cost? |
| Cost per retained donor | retention program spend / donors retained | Gives stewardship spending a denominator |
| Donor lifetime value | average gift x gifts per year x years retained | A budget planning figure, not a report card |
Run every measure on one population and one window, or the figures will not reconcile. Overall retention can look flat while first-year retention falls and average gift rises. The questions behind these numbers are set out in 4 plain facts about donor retention questions.
Example: 4,200 donors at a food bank
A food bank counted 4,200 donors in the prior fiscal year. Nine hundred were first-time donors not yet eligible to repeat. Of the remaining 3,300, 1,980 gave again, so repeat-donor retention is 60%.
The same group gave $1.24M last year and $1.19M this year. Value retention is 96%. Donor count held while revenue per prior-year donor slipped about 4%. The two results disagree because donor count and gift size move independently.
Sixty percent sits above the program's 55% floor, so the stewardship calendar runs unchanged. A reading below 55% triggers a mid-year review of the lapsed-donor sequence and a rebuild of the second-gift ask.
Recalculate before you compare
A published rate is comparable to yours only when the population and the window match. Pull the raw records, recompute the figure, then compare.
The Fundraising Effectiveness Project, run by AFP and the Giving Tuesday Data Commons, reports overall donor retention in the low 40s percent. Repeat-donor retention typically runs about 20 points higher. Check the edition year on the page before you quote it.
If your number and the published one differ, find which field differs before you change anything. Where records are incomplete, calculate with unknown-date gifts in, then out, and report both. The recurring gaps are collected in the parts worth your attention.
The W3C Privacy Principles give shared vocabulary for donor data. Confirm consent and receipting rules with your charity lawyer or the CRA, since retention periods vary by jurisdiction.
What to record beside the number
Attach six lines to every published figure.
- Source of the records and the pull date
- Population definition and exclusions
- Formula with the variables filled in
- Known gaps and the range they produce
- Threshold, and who set it
- Owner and next review date
The FTC advertising substantiation policy requires a reasonable basis for objective claims before they go out. That covers a retention figure in a board deck or a grant report. If you cannot show the working, do not publish it.
Choosing the tool that holds the benchmark
What you need is cohort retention reporting and a stored first-gift date. Bloomerang and Neon CRM report retention by cohort. DonorPerfect, Little Green Light and Blackbaud Raiser's Edge NXT keep the first-gift date a cohort depends on.
The GOV.UK guidance on choosing technology asks for control of your own data and attention to total ownership cost. Export the raw records once a year so the benchmark survives a platform change. Run the donor retention checklist against that export before you trust the cohort labels.
