Turn a shared, drifted GP account framework into a governed D365 chart of accounts — classifying segment meaning, designing account structures, and moving from GL batches to journal names.
This is the module where a GP programme succeeds or acquires its first irreversible problem. The GP account framework is shared across every company, hard to change, and semantically drifted. The D365 chart of accounts is governed, dimension-led, and validated by account structures.
GP's account framework is established centrally: segment count, individual segment lengths, and total account length are configured once and shared by every company database. It is difficult to change after the fact, which is precisely why estates adapt around it rather than to it.
The main account is the natural account. It carries type, category and posting behaviour, and it is a single value rather than a position in a string. Financial dimensions carry every other analytical meaning explicitly — a dimension called Department is called Department, and its values are governed as a set.
The migration pattern in the source material has four steps. They are sequential, and skipping the first is the classic failure.
Fiscal periods and years map structurally well: GP fiscal periods become fiscal calendar periods, and the calendar itself binds to the ledger. The mapping is rated direct in the source material, with medium confidence — the structure carries, the behaviour needs review.
GP's GL batch is a container for entries pending posting. D365's nearest construct is the journal name, which governs the header and lines — but a journal name carries considerably more than a container does: number sequence, approval behaviour, posting restrictions, default offset behaviour and voucher rules.
Currency behaviour must be confirmed per company rather than assumed from the group. A D365 ledger has an accounting currency and a reporting currency; revaluation, translation and exchange-difference account design sit alongside the segment redesign as separate decisions with their own approval.
The framework model, the segment-drift pattern and the target constructs are well supported. The following are not, and each is a validation item for every engagement:
The segmented-account redesign is the first gate and the largest single decision in a GP programme.