Global Re-Engagement Under a Hard Send Cap
A B2B SaaS firm has 1.1M marketable people. Its instance is capped at 900,000 email sends per month by a contractual deliverability agreement. Global email opt-out sits at 14% and has climbed 3 points in two quarters; the growth team blames a 2025 re-engagement blast that mailed the entire database twice. The lifecycle model currently has a single 'Dormant' bucket defined as no activity in 180 days, holding 610,000 people. Sales complains that 'Dormant' includes accounts with open opportunities. Marketing wants to re-engage all 610,000 within one quarter. The instance already runs 40 active engagement programs, and the ops team is at capacity.
1. The CMO wants the full 610,000 re-engaged this quarter. What do you recommend, and how do you reconcile it with the send cap?
Read the worked answer
The request is infeasible and the cap is the forcing function, not the obstacle. 610,000 × a multi-touch cadence (say 3 sends) = 1.83M, double the monthly ceiling and well over a quarter's headroom once lifecycle and transactional mail are counted.
Recommendation:
- Refine the definition first. 'Dormant = no activity in 180 days' is too blunt. Split by (a) open opportunity / active account ownership, (b) engagement recency bands (180–365, 365–730, 730+), and (c) source and consent vintage. Open-opportunity contacts should be suppressed from any re-engagement stream entirely - sales owns them.
- Tier the treatment. Only the middle bands get a multi-touch email cadence. The oldest, lowest-signal band gets a single low-cost touch plus paid retargeting and a preference-center prompt, not a 3-email sequence.
- Budget against the cap. Reserve ~15% of monthly capacity for lifecycle/operational mail, then allocate the remainder between re-engagement and business-as-usual. Re-engagement gets a defined slice, e.g. 150–200K/month.
- Protect deliverability. Use a dedicated subdomain/IP pool for re-engagement, throttle daily volume, and hard-stop a cohort on non-open after N touches.
The judgement being tested: you do not accept a volume target that conflicts with a contractual cap and a rising opt-out trend. You re-scope the population and sequence the work, and you say so explicitly with the math.
2. Sales says 'Dormant' is polluted. How do you change the model without breaking 40 running programs?
Read the worked answer
Treat this as a data-model change with a migration plan, not a rename.
- Don't redefine the existing field in place. Programs reference the current value; flipping its meaning silently changes who qualifies everywhere. Instead, introduce new fields (e.g.
Engagement Recency Band,Lifecycle Stage v2) and backfill, leaving the legacy field intact during transition. - Audit dependencies. Inventory which of the 40 programs use the Dormant value in smart-list filters and triggers. Each one needs an owner decision: migrate to the new field or retire.
- Sequence the cutover. Backfill → run new and old in parallel and compare counts → migrate programs one by one → only then deprecate the old field.
- Account-level suppression. Add a smart-list constraint that excludes contacts on accounts with open opportunities, sourced from the CRM sync, so the fix holds even as opportunities change.
This tests whether you understand that lifecycle fields are load-bearing infrastructure and that a 'small definition tweak' can ripple across every program in the instance.