Dimensions encoded in accounts
Separate accounts per office, per product line, or per project — the classic sign that the chart is doing a dimension’s job. It multiplies accounts and makes consolidation manual.
Services
Companies replace accounting systems to get reporting their chart of accounts was structurally incapable of producing. The new system arrives, the chart comes with it, and the reporting is the same — at which point the conclusion is usually that the new system was also wrong.
Send your chart of accounts and one report you cannot produce. We will tell you if the chart is why.
The situation
Four structural problems account for nearly every chart we are asked to look at.
Separate accounts per office, per product line, or per project — the classic sign that the chart is doing a dimension’s job. It multiplies accounts and makes consolidation manual.
A single account holding three economically different things because it was convenient once. Every report above it inherits the ambiguity permanently.
A flat list where roll-ups have to be maintained in a spreadsheet, which is how two versions of the P&L come to exist.
Each subsidiary with its own chart, mapped monthly by hand. The mapping lives in one person’s workbook and nowhere else.
A redesign that abandons comparability is not worth doing. Prior periods are remapped so year-over-year still works from day one.
A written policy on when a new account is justified, which is what stops the chart re-accumulating over the following three years.
The single principle behind every redesign is that an account should answer what kind of economic event this was, and dimensions should answer where, who, and what for. Chart bloat is almost always a dimension trying to live in the account number because the system had no better place for it.
A chart with 1,400 accounts usually has about 180 real ones and the rest are combinations. It is not a discipline failure; it is a structural response to a genuine constraint. But it constrains everything above it and no reporting layer can undo it.
Anyone can design a better chart. What makes a redesign usable is remapping several years of history onto it so that comparison still works and last year’s numbers do not become unreachable.
That mapping is rarely one-to-one. An account that conflated three things has to be split retrospectively using transaction-level evidence — vendor, description, department, amount pattern — and some of it needs a person. This is most of the engagement and it is what separates a redesign from a document.
A chart of accounts redesign is possible on QuickBooks, Xero, NetSuite, or anything else. The constraint is what dimensions your system supports — QuickBooks classes and Xero tracking categories are limited, which is why the graph is useful.
But the restructuring itself, and the discipline it introduces, delivers most of its value wherever your ledger happens to live.
Charts re-bloat within three years unless somebody owns them. The deliverable includes a written policy — who may create an account, what justifies one, and the standing answer that a request for a new account is usually a request for a dimension.
That last reframing is the durable part. Most new-account requests are somebody needing to see something separately, which a dimension does better and without permanent structural cost.
Where to start
Every account by usage, volume, and value. Dormant accounts, conflated accounts, and dimension-shaped accounts identified from transactions rather than names.
Accounts for economics, dimensions for context, hierarchy that rolls up natively. Agreed with your controller before anything moves.
Two to five years mapped onto the new structure, with splits derived from transaction evidence and the ambiguous cases reviewed by a person.
New structure live, comparatives working, and a written policy on account creation so it does not re-accumulate.
Questions
With one report you cannot produce. We will tell you whether the chart is why.