The close is a lineage problem before it is an accounting problem
The spreadsheet is usually a symptom of missing lineage
Finance teams build a spreadsheet when the official systems cannot answer a simple question: what changed, where did it come from, and who approved it? The spreadsheet becomes a private lineage layer. It works until the author is unavailable, the period changes, or an auditor asks for the source row behind a total.
Replacing the spreadsheet without replacing the question does not help. A new dashboard can make the same opaque number prettier. The useful system stores the relationship between a consolidated balance, its entities, its journal entries, its source transactions, and the evidence attached to the transaction.
Exception queues outperform blanket reconciliation
Most transactions are routine. The close slows down around the small group that is late, unmatched, outside tolerance, or missing evidence. A controlled reconciliation process clears the routine population automatically and puts only exceptions in a queue with an owner, reason, due date, and next action.
- Routine matches
- 94% cleared automatically
- Exceptions open
- 38 ranked by materiality
- Lineage depth
- 4 layers report to source
- Close cycle
- 2 days after controlled rollout
Controls have to travel with the entry
A control that lives in a month-end checklist is easy to skip and difficult to prove. A control attached to the entry can enforce approval thresholds, segregate duties, require supporting documents, and retain the decision next to the transaction. The result is less reconstruction during audit and less dependence on institutional memory.
Frequently asked
How does lineage shorten the financial close?
It removes the search and reconstruction work around balances. A reviewer can move from a report total to entity contributions, journal entries, source transactions, evidence, and approvals without asking someone to rebuild the path in a spreadsheet. The team spends time resolving exceptions rather than proving routine numbers.
What should be automated during reconciliation?
Routine matches within configured tolerances, duplicate detection, source completeness checks, and recurring evidence collection are good candidates. Items outside tolerance, missing evidence, unusual entries, and policy exceptions should route to a named reviewer with materiality and context attached.
Can an ERP replace spreadsheets immediately?
The safe approach is a controlled transition. Map the current spreadsheet logic, identify which rules belong in the ledger or reconciliation layer, validate the output against closed periods, and run both processes for a defined period. The goal is not to hide the spreadsheet. It is to replace its private lineage and control logic with an auditable system.
More from Field Notes
All essays
ERP Closing the books in two days: continuous reconciliation as a default
How continuous reconciliation collapses month-end close from 14 days to 2 — the controls, controls, and audit posture that come with it.
ERP Multi-entity, multi-currency consolidation that survives audit
How to design ERP consolidation for multi-entity, multi-currency operations — intercompany elimination, FX policy, and the audit trail your auditor will accept.
ERP Procure-to-pay automation that survives audit: three-way match, the right way
How three-way match automation actually closes loops between PO, receipt, and invoice — controls, exceptions, and the ERP architecture auditors recognize.