T-0 is the rehearsed cutover, for real: legacy frozen and set read-only at the cut-off, the final cockpit load repeating the timed dry run, the trial balance reconciled and signed by the finance controller, integrations switched to production, users logging in — and the first invoice posted in production, usually before lunch.
Keep the legacy archive queryable after it goes read-only. Rule 4 depends on history staying available without being migrated.
Hypercare, shrinking on purpose
T+1 to T+10, a daily 15-minute standup on yesterday’s exceptions, fixes and same-day configuration changes; T+11 to T+20, twice weekly, on trends and preparing the first month-end close; T+21 to T+30, absorbed into the normal operating rhythm, with T-Plus wave planning.
| Exception | Example | Handled by |
|---|---|---|
| Posting or reconciliation | Unmatched bank line, suspense balance | Finance key user + Ledger Architect |
| User cannot complete a task | Missing permission, unclear step | Adoption Navigator — ASK, role fix |
| Integration message failed | Bank file rejected | Ops Pilot + partner |
| Configuration change request | A new approval level | Ledger Architect, the same week |
Hypercare ends on evidence, not a calendar date: exceptions below threshold for five working days, the first month-end closed in TRANXACT, and no open critical or high defects.
How this meets the service contract: T-Minus hypercare, T+1 to T+30, is part of the method. Extended post-go-live support with daily check-ins — typically 4 to 8 weeks — is the separately contracted H1 Hypercare package in the Roles & Responsibilities workbook. After hypercare, the steady-state service takes over: service requests, incidents, releases and drills.
T-Plus: growing by configuration
The T-Plus backlog — every T-Plus Fit Ledger line from wave 1, plus more entities, regions, modules and any period data a new report needs — is prioritised with the Countdown Owner, then run as a mini-countdown per wave: rehearse the change, toggle it on. Plan the first wave before hypercare ends, while the countdown team is still together; each wave reuses the same Fit Ledger, plans and artifacts. The measure is modules live per quarter without re-implementation.
The artifacts, and where they live
The Project Plans workbook holds one countdown plan per source system and a dashboard across them: enter the go-live date and holidays once, keep the sheets for the customer’s sources, assign owners, and read the dashboard daily. Everything else — the Fit Ledger, migration scope and object register, configuration workbook, test plan, reconciliations, cutover runbook, rehearsal script and gate, the adoption, change, stakeholder and comms sheets, and the RAID log, countdown board, hypercare log, T-Plus backlog and KPIs — is one sheet each in the Artifact Toolkit.
Calderwood posts its first invoice at 10:40 on 1 October. Exceptions fall below threshold by T+14, but hypercare does not close then: it waits until October’s month-end has been closed in TRANXACT, in the first days of November, with no critical or high defects open. The HRMS wave and the deferred tax integration are planned in the T+25 standup.
Try it Exceptions at Calderwood have been below threshold for eight working days by T+18. Can hypercare exit?
Not yet, on one criterion: the first month-end has not been closed in TRANXACT. Exit needs all three — exceptions below threshold for five working days, the first month-end closed, and no open critical or high defects. The cadence can still drop to twice weekly as planned.