Your ERP has been running for years. Transactions have been posted, vendors have been added, customers have been created, invoices have been processed. The system works — in the sense that it keeps running and people keep using it.
But "the system works" is not the same as "the data is correct."
What typically goes wrong
The most common problems we see fall into three categories.
1. Duplicate master data
Vendors and customers get created multiple times. Sometimes it's a typo: "Acme Corp" and "ACME Corp" and "Acme Corporation" are the same company. Sometimes it's a system migration artifact. Sometimes it's a user who couldn't find the existing record and created a new one.
The downstream effect: your AP liability is potentially overstated. Your AR aging is potentially understated. Your cash flow projections are wrong because they're based on the wrong vendor and customer counts.
We have seen companies with 40% duplicate vendor rates. A company spending $5M annually with 200 vendors might actually have 280 vendors in their system — 80 of which are duplicates of existing records.
2. GL entries that don't tie to subledgers
The general ledger is supposed to be the source of truth. Every GL entry is supposed to tie to a corresponding subledger entry — an AR invoice, an AP bill, a payroll record.
In practice, many GL entries exist without matching subledger records. They were posted directly to the GL during a period close. They came in through a journal entry that bypassed normal posting controls. They're the result of an import that mapped incorrectly.
The result: you cannot fully reconcile your GL to your subledger. Your AR aging report does not match your AR GL balance. Your AP aging report does not match your AP GL balance. Month-end close requires manual adjustments to paper over the gap.
3. Reconciliation gaps between periods
Every month, the closing trial balance for month N should tie to the opening trial balance for month N+1. Retained earnings should roll forward correctly. Beginning balances should match ending balances from the prior period.
They often don't — for small amounts that someone dismissed as rounding, or for amounts that were manually adjusted without documentation, or because a period was accidentally left open and a transaction was posted after the close.
How to find out how bad it is
The fastest way is a structured data health score. Specifically:
- Export your chart of accounts, vendor master, customer master, and a trial balance with all periods.
- Export your GL detail for the same periods.
- Export your AP aging and AR aging as of the most recent period end.
From these five exports, you can run:
- A duplicate scan on vendor master and customer master (fuzzy matching on name, address, tax ID)
- A GL-to-subledger reconciliation for AP and AR
- A period-over-period continuity check on the trial balance
- A posting completeness check (entries with no corresponding subledger record)
The first two are the most common sources of real financial risk. The last two are the most common sources of close pain.
What the numbers tend to look like
Based on the data we have processed:
- Companies with more than 5 years of ERP history typically have 15–40% duplicate rates in vendor master
- Companies that have migrated systems once have a higher-than-average rate of GL-subledger mismatches in the migration period
- Companies where close is done manually (not through a structured close checklist) tend to have the most period continuity gaps
None of these are permanent problems. They are data problems, and data problems are fixable.
The free data health score
If you want a specific answer — not estimates, but actual findings from your actual data — the free data health score is the right first step. Submit your exports, and we run the analysis and send you a report with specific findings and dollar amounts.
It takes us about two hours of automated pipeline time, plus 30 minutes of human review. You receive a PDF. No credit card, no commitment.
If the findings are minor, you'll know that too. Not everyone needs a full rehab. Some companies are in better shape than they think. But most aren't — and it's better to know.