FIELD GUIDE 05 / RETENTION

Before You Call Them Lapsed

Retention · 5 min read · September 2026

Reconcile the donor count before choosing who to contact.

You have 75 donors this year and 100 last year. Twenty are new. Does that mean retention is 55%?

Only if every other current donor also gave last year. Donors returning after a gap can make that shortcut wrong.

All figures below are fictional. The example compares the complete calendar years 2024 and 2025.

The answer you can explain

Group	Donors	Where they belong
Gave in both 2024 and 2025	45	Retained donors.
First-ever gift was in 2025	20	New donors; outside the 2024 cohort.
Gave before 2024, skipped 2024, gave in 2025	10	Returning donors; outside the 2024 cohort.
Gave in 2024, no qualifying gift in 2025	55	Members of the prior cohort to investigate.

2024 cohort: 45 retained + 55 absent in 2025 = 100.

2025 donors: 45 retained + 20 new + 10 returning = 75.

Retention: 45 ÷ 100 = 45%.

Subtracting the 20 new donors from 75 gives 55. That includes ten returning donors who were never in the 2024 cohort. It answers a different question.

A WORKED EXAMPLE · FICTIONAL

75 donors this year. 45 retained.

2024 · 100 donors

45 retained + 55 absent in 2025 = 100

2025 · 75 donors

45 retained + 20 first-ever + 10 returning = 75

PRIOR COHORT RETENTION45 ÷ 100 = 45%

The 10 returning donors skipped 2024.
They do not belong in its retained count.

FIG. 05.01 Returning after a gap adds to this year’s donors, not last year’s retained cohort.

A paragraph for the board report

Board report · fictional example

Of the 100 donors who made a qualifying gift in 2024, 45 gave again in 2025: annual donor retention was 45%. The 75 donors in 2025 comprise 45 retained, 20 first-ever and 10 returning donors. We have not yet established why the other 55 prior-year donors had no qualifying gift in 2025. The next step is to review payment issues, identity changes and giving timing before deciding on outreach.

Use that structure with your own definitions and counts. If you do not have enough history to distinguish first-ever from returning donors, write “30 current-only donors; earlier giving history not yet reconciled.” Do not label them all new.

Get two lists that mean the same thing

Before exporting, complete this note with the person who owns the report:

Decision	Record your rule
Periods	[Start and end dates for each period.]
Donor identity	[Individual, household or organization; stable ID field.]
Qualifying gifts	[Included gift types and how refunds, reversals and intermediaries are treated.]
Credit	[How direct gifts and soft credits are treated without counting the same relationship twice.]
Date used	[The gift date field applied consistently to both exports.]
Report source	[Report names, filters and export date.]

Use received gifts rather than unpaid pledges for this calculation. If you need a different metric, name it separately. For an unfinished year, use a clearly defined year-to-date comparison with the same cutoff in the previous cycle; do not present it as completed annual retention.

Export one stable donor ID per line for each period. Keep names, emails and gift notes out of the comparison lists. Duplicates in transaction exports should not create extra donors.

Check the match with a tiny example

Prior-period sample IDs

EXAMPLE-101 EXAMPLE-102 EXAMPLE-103 EXAMPLE-104

Current-period sample IDs

EXAMPLE-102 EXAMPLE-104 EXAMPLE-105

The intersection is EXAMPLE-102 and EXAMPLE-104: 2 ÷ 4 = 50%. EXAMPLE-101 and EXAMPLE-103 are absent from the current list. EXAMPLE-105 needs older history before you can call it new or returning.

If the prior-period list is empty, retention is undefined. Report “no eligible prior cohort,” not 0%.

The Atlas cohort calculator compares exact, case-sensitive IDs, ignores duplicate lines and trims surrounding spaces. It cannot correct two different IDs that refer to the same donor. If a migration changed IDs, reconcile the identities before relying on the result. Use these CRM migration acceptance checks to check record counts, attribution, and history before comparing periods.

COHORT CHECKCOMPARE → EXPLAIN → REVIEW

Start with the fictional example below, or replace it with your periods and anonymous stable donor IDs, one per line. Exact matches are case sensitive. Repeated IDs and extra spaces are ignored.

Prior period starts

Prior period ends

Current period starts

Current period ends

Who and what count in both lists?

Prior-period donor IDs

Current-period donor IDs

50.0%retention

2 / 4retained / prior donors

2prior IDs absent this period

1current-only IDs; history needed

Your report paragraph and counting note

Of the 4 donors with a qualifying gift from 2024-01-01 to 2024-12-31, 2 gave again from 2025-01-01 to 2025-12-31: donor retention for this comparison was 50.0%. The current period has 3 donors: 2 retained and 1 current-only. Earlier giving history is needed to distinguish first-ever from returning donors among the current-only group.

2 prior-period donors are absent from the current list. Absence alone does not explain why. Review payment issues, identity changes, giving timing and contact preferences before deciding on outreach.

Counting rule: Individual donor IDs with at least one received gift; unpaid pledges excluded.
Prior period: 2024-01-01 to 2024-12-31. Current period: 2025-01-01 to 2025-12-31.
This comparison uses exact, case-sensitive IDs after trimming spaces and removing duplicates. Confirm consistent identities and comparable periods before relying on the result.

Current-only IDs may be new or returning; older history is needed. Inputs stay on this page and disappear on reload.

TAKE IT TO YOUR TEAMThe retention reviewEditable guide · Cohort rules, board paragraph and investigation record

Turn a missing ID into an investigation

What you find in the record	What happens next
A scheduled recurring payment failed	Check the payment status and the appropriate support response with the responsible person.
Giving moved to a household, business or new ID	Resolve the identity rule and recalculate before interpreting the change.
A multi-year giving schedule explains the gap	Record the agreed timing; do not assume a missed annual gift means the relationship ended.
The donor asked for a pause or no contact	Preserve that preference. Absence does not override it.
There is no clear explanation	Have the relationship owner review the history and choose an appropriate next step.

Copy this into the review record:

Blank investigation record · replace every bracketed field

ID: [Internal ID] Observation: [What the records show.] Still unknown: [What would change the decision.] Next action: [Specific check or response.] Responsible person and review date: [Name, date.]

The calculation identifies which records need attention. The relationship history determines what attention is appropriate.

READING & SOURCES

Definition reference: Fundraising Effectiveness Project terminology. The examples and report wording are original Gratona tools.

Explore the reading list ↗