Before You Call Them Lapsed

Build the prior-period cohort

Export distinct donor IDs with qualifying gifts in the prior period. Keep one row per donor. Confirm the count against the relevant CRM report and investigate duplicates or missing attribution.

Find the intersection

Match that list to distinct qualifying donors in the current period. Retained donors are the IDs in both lists. Divide that count by the prior-period donor count. If the prior cohort is empty, the rate is undefined.

Reconcile this year’s donors

Partition current-period donors into retained, first-ever donors, and returning donors who gave before the prior period but skipped it. Use sufficient history to distinguish new from returning. If history is incomplete, retain an unknown category and investigate it; do not force it into ‘new.’

Segment without changing the cohort

Break the prior cohort down by prior-period giving band, donor type, or acquisition source where reliable. Freeze giving bands using prior-period amounts so current-year upgrades do not move donors between comparison groups. Show numerator and denominator next to each percentage.

Investigate one pattern

Review a small set of records behind an unexpected result. Look for missed commitments, changed timing, identity changes, program concerns, or a donor-requested pause. Choose one action and an owner, then set a date to learn whether the process changed.

Put it to work

Suggested session: 60 minutes once exports are ready. Bring the relevant records and allow additional time for preparation and delivery.

• Choose two complete, comparable periods. If the current year is unfinished, label the result year to date and compare it with the same cutoff in the prior cycle, not a completed annual rate.
• Write down whether a donor means an individual, household, or organization. Use stable IDs and apply the same rule in both periods.
• Define included gift types, how refunds and reversed gifts are handled, and how intermediary or soft-credit attribution is treated. Apply the same policy in both periods. Do not count unpaid pledges as received gifts.

When the plan changes

• A changed household rule or CRM migration can create false gains and losses; reconcile identities before interpreting the trend.
• A recurring payment failure is different from a decision to stop giving. Investigate the payment and the relationship separately.
• A donor who has not yet reached their usual renewal month should not automatically enter a lapsed-donor outreach sequence.
• Small cohorts: report counts and inspect cases. A dramatic percentage may represent one donor.

What finished looks like

• Periods, identity rules, and gift rules are documented.
• Retained donors are members of the prior-period cohort.
• The current donor count reconciles or the missing history is explicit.
• One observed pattern has an investigation, owner, and review date.

See whether it helps

Separate annual retention, first-year donor retention, gift frequency, and revenue. Use external benchmarks only when their period, sample, and definitions are sufficiently comparable. A process improvement and a revenue change occurring together do not establish causation.