A recurring pattern in finance transformation work: an organisation invests in reporting technology, connects it to the ledger, and discovers the reports it wanted were never possible. The dimension it needs to slice by does not exist in the data.
The chart of accounts is not administrative housekeeping. It is the schema of the business. Every management question you will want to answer for the next several years has to be expressible in it, or the answer requires manual reconstruction each period.
Design from the report backwards
Start with the reports the board, the lender and the regulator will ask for, then work back to the structure that produces them without intervention. Segment, cost centre, project and product should be separate dimensions rather than encoded into a single account code — a habit that seems efficient at implementation and becomes unworkable at the second reorganisation.
Common structural mistakes
- Encoding department and expense type into one flat account, so neither can be reported independently
- Creating accounts for individual vendors or customers that belong in the sub-ledger
- Leaving no numbering headroom, forcing unrelated accounts into whatever gaps remain
- Allowing postings directly to control accounts that should only ever receive sub-ledger entries
- Designing for the current organisation chart rather than for the reporting obligation
Changing it later
Restructuring a chart of accounts mid-life is possible but rarely painless, because comparatives must be remapped for every period a reader will compare against. The practical window is at implementation or at a deliberate reset — which is why the design deserves more time than it usually receives.
This article is general commentary and not advice on any specific set of facts. For guidance on your own circumstances, speak to our team.

