The merged-consent fix behind the 4-versus-250 audience
Evidence for: A green checkmark said 4 people were reachable. The real number was 250.
The audience builder re-checked consent against the ecommerce platform alone on top of a customer-data flag that already merged both platforms. The double check discarded about 250 genuinely opted-in contacts and left a sendable cohort of 4. The fix replaced the redundant check with a merged-consent rule and added a conflict check that holds back contacts whose two sources disagree.
- Sendable before the fix
- 4
- Reachable pool after the fix
- about 250
- Cross-source consent conflicts excluded
- 134 (the conflict file has exactly 134 entries)
- Final sendable after conflicts, deliverability and frequency filters
- 228 (the segment file has exactly 228 lines)
- Consent rule as applied
- sendable if either platform shows opt-in and neither shows an explicit unsubscribe; any explicit negative wins
What it proves
A real, dated, hash-verifiable fix in which a consent-handling bug was silently suppressing about 98 percent of a legitimately opted-in audience, and the fix that restored it also tightened compliance.
Excerpt
fix(email): merged-consent model in attach-remainder cohort build. BUG: cohort-build stage 3 re-checked live platform-only consent on top of the customer-data opted-in flag; double-gating discarded about 250 genuinely consented contacts, crushing the sendable cohort to 4. FIX: merged-consent model (either platform opt-in, negative wins) plus cross-source conflict exclusion from a read-only three-way reconciliation dated 2026-09-07.
Where it lives: commit · freya-brain (the ecommerce brand's own send pipeline) · 7c5d66d504eb85e2c4f54f757f9357f2ab3f5fac ·
Redaction: Counts, the commit hash, the date and the rule as written. No customer records, no file paths beyond the repo name.