Why does the close get slower every time you add an entity?
Because what scales isn't the data — it's the disagreement. Each entity arrives with its own definition of revenue, margin and cut-off, and the close absorbs the cost of making them agree, every month. New tooling makes the mechanical work faster without deciding what the numbers mean, which is why the close date rarely moves.
Most finance leaders describe the same curve. The first entity closes in five days. The third takes eight. By the seventh, the close has its own project plan, and the team has grown twice without the calendar moving.
The instinct is to blame volume — more transactions, more reconciliations, more people to chase. But volume scales linearly, and the close does not. Something else is compounding.
What compounds is disagreement
Each entity arrives with its own history: its own chart of accounts, its own cut-off convention, its own idea of when revenue is earned and what belongs inside "other operating costs". None of these are wrong locally. Each was a reasonable decision made by someone solving a local problem.
The close is where those decisions meet. And the cost of the close is not the cost of collecting numbers — it is the cost of negotiating which definition wins, every month, in the same arguments.
That is why the curve bends. Two entities produce one relationship to reconcile. Seven produce twenty-one.
Why new tooling rarely moves the date
A consolidation system, a new BI layer, a bot on the manual steps — each makes the mechanical work faster. None of them decide what revenue means. The definitional work lands back on the same three people who were already carrying it, now behind a better interface.
Teams that break the curve do the unglamorous thing first: they write the definitions down, name an owner for each, and make the system enforce them at the point of entry rather than at the point of consolidation.
What that looks like in practice
- One definition per metric, in plain language, with a named owner. Not a data dictionary nobody opens — a short list covering the fifteen numbers the board actually sees.
- One cut-off convention, applied everywhere, even where local practice differs. Local reporting can diverge; group reporting cannot.
- Lineage a non-technical reader can follow. When a director asks where a number came from, the answer should take one click, not one analyst-day.
- Enforcement at entry. A definition that lives in a document is a suggestion. A definition the system applies is a rule.
None of this requires a platform migration. It requires deciding once, and then holding the line.
The test
Ask two people in two entities for the same number, on the same day, without warning.
If the figures differ, the gap is not a reporting problem. It is a definitional one — and no amount of tooling will close it.