The wave plan, the rehearsals and the go/no-go gates

Sequence the migration into nine dependency-ordered waves, prove the operating chain through golden scenarios, rehearse the cutover until it is boring, and hold the go-live decision against criteria agreed in advance.

What you will be able to do

Introduction

The wave plan exists because migration objects depend on each other, and loading them in the wrong order produces failures whose cause is several steps upstream of where they appear. The rehearsals exist because a cutover is a timed operational procedure, and timed procedures are proven by executing them rather than by planning them.

The nine waves

Several of the dependencies are worth naming explicitly, because they are the ones teams try to work around.

The three centrepieces, restated as cutover objects

Three objects carry disproportionate cutover risk in a SYSPRO migration, and it is worth restating what each needs at the moment of transition.

The golden scenarios

The playbook specifies nine end-to-end scenarios, and their common property is that each crosses module boundaries. That is what makes them useful in a migration from a tightly coupled source: the boundaries are where the target's greater separation shows.

Rehearsing the cutover

The distinction that matters here is between a load rehearsal and a cutover rehearsal.

Go/no-go criteria

The playbook's criteria are specific enough to be checked, which is what makes them usable:

Cutover choreography

The window is a sequence, and every step needs an owner, a duration and a completion check.

Hypercare, designed rather than staffed

Hypercare is where the SYSPRO-specific risks concentrate, because the coupling that the target has separated is now separated in production.

Knowledge check

Summary

Nine waves, dependency-ordered, each with a gate signed by a named business owner. On-hand, open work in progress and the open financial book are the three cutover objects that carry disproportionate risk. Golden scenarios cross module boundaries, run on real data, executed by the people who will do the job, and end in a reconciliation.