Sizing segments: the arithmetic

Length is capacity. A three-character numeric cost centre segment gives you 999 values, which sounds like plenty until you meet a mid-size client who hits 700 in year two. Size for the value count you expect at year five, then add one character. The cost of an extra character is close to zero. The cost of running out is a re-implementation. Numeric versus alphanumeric: alphanumeric gives enormous headroom and readable codes, and it also breaks sorting assumptions and some legacy integrations. That is the trade-off, and it is the client's to make rather than yours. Then combination arithmetic. Multiply the value counts across segments and a perfectly reasonable six-segment design produces a combination space in the hundreds of millions. Mostly that is fine, because dynamic insertion means only used combinations exist — but hierarchies and cubes still feel it. And leave gaps. 1000–1999 assets, 2000–2999 liabilities. Never assign sequentially from…

In this module: Module 4 · Chart of Accounts and Ledger Design

1. 1What we're building
2. 2The decision you can't undo
3. 3Anatomy of a chart of accounts
4. 4Segment labels and qualifiers
5. 5Running the design workshop
6. 6Sizing segments: the arithmetic
7. 7Building value sets
8. 8Structure, instance, deployment
9. 9Loading account values
10. 10Account hierarchies and why reporting needs them
11. 11Building and publishing a hierarchy
12. 12Accounting calendars
13. 13Currency and conversion rate types
14. 14Primary, secondary, reporting: the decision tree
15. 15Building the primary ledger
16. 16Ledger options and completing the definition
17. 17Cross-validation and segment value security
18. 18What breaks
19. 19Lab briefing: Meridian Trading