The Sunset Call: Auditing FirstCry's Dormant Subscriber Export
Objective: Given a 30-contact dormant subscriber export with last-open, last-click, and last-purchase dates, apply the lesson's multi-signal dormancy definition to segment by depth of inactivity and flag who is suppression-ready.
You're the CRM analyst at FirstCry (Brainbees Solutions) validating which dormant contacts deserve the win-back sequence versus immediate suppression, ahead of a Diwali sale send.
Apply 'no click AND no purchase AND no site visit in X days' rather than opens alone, then split the export into recently-lapsed, deep-sleepers, and suppression-ready buckets.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, handles formula-based multi-signal filtering without a platform login
Paid upgrades (optional, faster/deeper)
Segments recompute automatically as contacts' activity changes, instead of a static one-time export
No access? Re-export and re-run the Sheets formulas weekly
The process
2 steps
Step 01 of 02
The lesson's Step 1 warns that Apple Mail Privacy Protection inflates opens for roughly half of iOS users, so a clean dormancy trigger needs 'no click AND no purchase AND no site visit', not opens alone.
12 of the 30 rows show 0 opens in 120 days but 3 site visits and 1 click. Are they dormant?
Procedure
- Import dormant-export.csv and freeze the header row
- Add a formula column: TRUE dormant only if clicks=0 AND purchases=0 AND site visits=0 in the window
- Flag the 12 opens-only-zero rows as active, not dormant, since they clicked or visited
FALSE POSITIVE (opens=0, but active) row 14: 0 opens / 3 site visits / 1 click, 118 days -> NOT dormant TRUE DORMANT row 22: 0 opens / 0 clicks / 0 visits, 140 days -> dormant
Healthy
The dormant segment excludes anyone with a click or site visit, even at zero opens.
Unhealthy
Filtering on 'opens = 0' alone and sending sunset emails to people who are actually still browsing.
What this means
Opens are the least reliable signal in the export; clicks and visits are ground truth.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| The dormant filter uses only the opens column | Add clicks and site-visit columns to the dormancy formula before segmenting | 5 min |
Step 02 of 02
The lesson's Step 2 says a subscriber silent for 95 days gets a lighter touch than one silent for 14 months, and cites a Litmus study where targeting only the top decile of lapsed engagers outperformed blasting the full dormant cohort.
Of the true-dormant rows, 9 lapsed 90-120 days ago and 5 lapsed 300+ days ago. Do both get the same sequence, at the same time?
Procedure
- Sort true-dormant rows by days-since-last-activity, ascending
- Bucket into 'recently-lapsed' (90-150 days) and 'deep-sleepers' (150+ days)
- Flag any row past 365 days with a prior failed win-back attempt as suppression-ready, not resend-ready
RECENTLY-LAPSED (9 rows): send Week 1 DEEP-SLEEPERS (5 rows, 300+ days): send Week 2, lighter frequency SUPPRESSION-READY (3 rows): already completed one sequence with zero engagement, skip resend
Healthy
Recently-lapsed contacts get the sequence first; deep-sleepers wait and get fewer touches; repeat non-responders are suppressed, not resent.
Unhealthy
Blasting the full 14-row dormant list in one send regardless of how long each has been silent.
What this means
Depth of dormancy changes both timing and volume, not just messaging.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| The send plan treats 95-day and 300-day silence identically | Split the send into two waves by dormancy depth, deep-sleepers get Week 2 | 30 min |
Final deliverable
A segmented version of the export with three tagged buckets, recently-lapsed, deep-sleepers, and suppression-ready, plus a one-line send-order recommendation.
See a reference example
Instacart, dormant list audit (excerpt) RECENTLY-LAPSED (Week 1 send): 14 contacts, 90-140 days silent, no click/visit/purchase DEEP-SLEEPERS (Week 2, lighter touch): 6 contacts, 150-280 days silent SUPPRESSION-READY (skip): 4 contacts, already completed one sequence with zero engagement
Success criteria
You're done when you can:
- Correctly excludes opens-only-zero rows that had a click or site visit
- Splits true-dormant rows into recently-lapsed and deep-sleeper buckets
- Flags repeat non-responders as suppression-ready rather than resend targets