The Pre-Cutover Audit: Catching TBO Tek's Migration Failure Points
Objective: Given a synthetic migration plan document, apply the lesson's three failure points (lost historical data, broken automations, IP warmup) to catch what the plan is missing before cutover, not after.
You're reviewing a draft migration plan at TBO Tek, the Gurugram-based B2B travel distribution platform (Nasdaq-adjacent NSE/BSE listing, ~$1.79B IPO market cap) ahead of a CRM and ESP switch serving its 147,000+ travel-agent network. The plan looks complete on paper, your job is to find what it's missing before the migration team signs off.
Read a synthetic 1-page migration plan and score it against the lesson's three common failure points, flagging any gap that would cause silent data loss, broken automations, or a deliverability collapse.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, and both artifacts are simple tables well suited to a spreadsheet
The process
2 steps
Step 01 of 02
The lesson's second failure point: every workflow trigger references specific field names and values, so a renamed or reformatted field breaks the automation silently, or fires it incorrectly, unless every single automation is tested against sandbox data before cutover.
TBO Tek's draft plan says 'test top 3 automations in sandbox.' Their marketing stack actually runs 14 automations, including agent-tier upgrade emails tied to a field that's being renamed during migration. What's wrong with 'test the top 3'?
Procedure
- List all 14 automations from the plan's appendix, not just the top 3 the plan calls out
- For each, identify the trigger field it depends on
- Cross-reference against the field-mapping doc to flag any automation whose trigger field is being renamed or reformatted
- Mark which of the flagged automations the plan actually tests in sandbox
- Flag the gap: automations tied to a changed field but not sandbox-tested
TBO Tek automation audit (excerpt) Tested in plan: 3 of 14 automations Flagged as trigger-field-affected: 6 of 14 UNTESTED + AFFECTED (the real risk): 4 1. Agent-tier upgrade email -> trigger field 'agent_tier' being reformatted, NOT in test-3 list 2. Renewal reminder sequence -> trigger field 'contract_end_date' format changing, NOT tested 3. Win-back campaign -> untouched field, low risk, not urgent to test 4. High-volume booking alert -> trigger field renamed, NOT tested
Healthy
All automations whose trigger field changes are tested in sandbox before cutover, regardless of how prominent they are.
Unhealthy
Only the 3 most visible automations get tested, while 4 lower-profile ones tied to renamed fields go untested and break silently on cutover day.
What this means
'Test the top 3' is a false sense of coverage, the automations that actually break are the ones tied to a changed field, not the ones anyone remembers to check by default.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| 4 automations depend on a renamed field but were never sandbox-tested | Add all 4 to the sandbox test list before the plan gets sign-off | half day |
Step 02 of 02
The lesson's third failure point: domain reputation usually carries over to a new ESP, but IP reputation does not, so sending full volume on day one from a brand-new IP reads as a spam signal and craters deliverability.
The draft plan schedules the ESP cutover for a Monday with 'resume full sends immediately.' TBO Tek sends to roughly 60,000 travel-agent contacts. What's missing from that plan?
Procedure
- Confirm the plan has zero mention of a warmup ramp before full-volume resumption
- Calculate a realistic warmup schedule: start with the most-engaged 10% of the 60,000-contact list, scale over 6-8 weeks
- Add the schedule as a required addition to the plan before it goes to sign-off
- Flag the specific risk: sending 60,000 emails from a brand-new IP on day one
Recommended warmup schedule (excerpt) Week 1: 6,000 contacts (most-engaged 10%) Week 2: 12,000 contacts Week 3: 24,000 contacts Week 4-6: scale to 42,000 Week 7-8: full 60,000 list, only after inbox placement holds steady PLAN GAP: original draft schedules full 60,000-send for cutover Monday, no ramp
Healthy
A documented multi-week ramp starting with the most-engaged subscribers before resuming full volume.
Unhealthy
Full-volume resumption on cutover day, treated as a formality rather than a deliverability risk.
What this means
IP warmup is the one failure point unique to ESP migrations specifically, and it's the easiest to miss because domain reputation surviving the move creates false confidence that everything else did too.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Migration plan has no warmup schedule before full-volume resumption | Insert a 6-8 week warmup ramp starting with the most-engaged 10% of the list | half day |
Final deliverable
An automation-risk checklist flagging untested-but-affected automations, plus a documented 6-8 week IP warmup schedule to add to the migration plan.
See a reference example
Razorpay marketing ops, pre-cutover audit (excerpt) AUTOMATION RISK: 3 of 11 automations tied to a renamed field are untested WARMUP GAP: plan has no ramp, resumes 40,000-contact full volume on cutover day RECOMMENDATION: add all 3 automations to sandbox test list, insert 6-week warmup ramp before sign-off
Success criteria
You're done when you can:
- Correctly identifies every automation tied to a changing field, not just the plan's stated top 3
- Flags the missing IP warmup ramp as a required addition, not an optional nice-to-have
- Produces a realistic week-by-week warmup schedule scaled to the actual list size