Editorial draft for professional review. No unapproved author or reviewer is named. Updated:
What you will check
- Reconcile row and document counts to the source export.
- Check debit and credit totals, then test completeness.
- Preserve the original and document every correction.
Agree what the export represents
Record the entity, reporting period, currency and source system. Distinguish posting date from document date: both can be useful, but a filter based on one must not silently become a filter based on the other. Establish whether opening balances, closing entries and reversals are included.
Keep the original file. Make working corrections in a separate copy and record the reason for each change. A later difference can then be explained without guessing which version was used for analysis.
Define the meaning of each column
Prepare an entry identifier, document number, account, date and debit/credit amounts, or a clearly defined signed amount. ERP column names vary; a column called Debit does not establish how reversals are recorded.
Treat account and document codes as identifiers, not arithmetic values. Excel can remove leading zeroes or turn a code into a date. For CSV files, check encoding, delimiters, decimal separators and quoting before importing the complete file.
Create a pre-import control sheet
Compare row counts and totals with a control export from the source system. Isolate rows without a date, account or amount. Review repeated identifiers; matching amounts alone do not prove duplication.
Inspect a few characteristic entries: an ordinary invoice, a reversal, an opening balance and a document with multiple lines. This checks the interpretation of the data, rather than merely whether the file can be read.
Synthetic example: balanced totals, incomplete scope
Imagine an export with EUR 120,000 in debits and EUR 120,000 in credits. Equality establishes arithmetic balance, not completeness. If both lines of a EUR 5,000 document are omitted, the totals still match.
Alongside balance, check the date range, document count, included accounts and reconciliation to the trial balance. Record each difference with the expected result, actual result, cause and reviewer’s decision.
What to check in a FynTwin demonstration
Use a synthetic sample to explore column recognition, mapping and warnings. Ask how your specific reversal format is handled and how a repeat import differs from the first. Verify available controls in a demonstration without sending real personal data through the contact form.
An export that passes the checks is only a starting point. An accounting professional still confirms the period scope, account classification and applicable reporting rules. Continue with the guide to Croatian AOP positions.
Work through an example
Download six entries covering three documents: a EUR 1,200 sales invoice, its receipt and a EUR 300 purchase invoice. In a copy, remove both INV-001 entries. Which controls change?
Synthetic learning example. This is not a FynTwin export or a real client report. Amounts are in EUR; VAT, opening balances and closing entries are omitted for simplicity. Account codes are illustrative. The CSV is not a universal import template.
| Check | Expected result |
|---|---|
| Original dataset | 6 rows; 3 documents; EUR 2,700 each in debits and credits. |
| Without INV-001 | 4 rows; 2 documents; EUR 1,500 each in debits and credits. |
| Balance check | Debits minus credits remain EUR 0 in both datasets. |
| Conclusion | A EUR 1,200 document is missing. Arithmetic balance does not establish completeness. |
Before proceeding, obtain a control list of documents from the source system. Do not add an artificial entry merely to reach an expected total.
Downloadable resource
Download the synthetic ledger (CSV)6 rows, English headers, UTF-8, comma-separated with decimal points. Source documents: INV-001, PAY-001 and BILL-001.