Sequence the load waves so no design depends on an unratified decision, build a reconciliation an auditor accepts, and apply the source's own readiness test — a programme is not cutover-ready while subaccount decomposition or interface replacement is unresolved.
Everything in this path has been building towards a sequence and a set of proofs.
The playbook gives a nine-stage plan. The stage labels are the playbook's own and the ordering logic is what matters.
The validation set covers more than balances, and it is useful to group it into four lenses that answer different questions.
A mock cutover is a rehearsal, and the value of a rehearsal is proportional to how closely it resembles the performance.
A recurring conclusion across the Inventory, Manufacturing and Trade modules deserves to be stated once as a programme decision: arrange for fewer partial states rather than migrating the ones you have.
Finance asks for history because inquiry matters: an auditor's question about a transaction from two years ago, a customer dispute, a warranty claim. The reflex answer is to load history into the target.
The go-live decision needs criteria set in advance, because a decision made in the room on the night is made under pressure by people who want to go.
The load sequence follows the target's dependency graph: currencies and units, then the chart and dimensions, then grouping and posting segmentation, then products and warehouses, then open transactions, then projects and production, then custom data and interfaces, then the finance cutover, then the controlled delta.