An ERP data migration checklist helps leaders prove that the new system contains the right records and supports real business work. A successful file upload is only the beginning. The important question is whether staff can invoice customers, receive stock, pay suppliers, and close the accounts using trustworthy information.
Imagine launching on Monday with the right number of customer records but the wrong payment terms. The import report looks clean. Cash collections, credit control, and customer service soon tell a different story. Migration therefore needs business ownership as well as technical execution.
Use this guide to set scope, assign responsibility, test reconciliation, and make a go-live decision based on evidence. AB Consulting helps companies connect system implementation with operating processes through its ERP consulting services.
1. Define the scope of your ERP data migration checklist
Start with the transactions the business must complete on its first working day. Work backward to the data those transactions need. A supplier payment, for example, may depend on a supplier record, payment terms, an approved bank account, an open invoice, and the correct entity and currency.
Separate three categories: configuration, master records, and open transactions. Configuration includes items such as currencies, tax codes, and posting rules. Master records include customers, suppliers, and products. Open transactions include unpaid invoices, unfulfilled orders, and stock balances.
Microsoft distinguishes configuration from migration data in its official data-management guidance. Its instructions are product-specific; the business distinction is also a useful way to structure your planning for other ERP platforms.
Decide what historical data belongs in the operational system and what can remain in a controlled archive. Document how staff will retrieve older invoices and evidence. Retention and access obligations depend on your jurisdiction and records; confirm those requirements rather than deleting history simply to simplify migration.
Give each data set a named business owner
Finance should approve financial balances and posting rules. Operations should approve product units, stock locations, and production data. Sales should validate customer terms and addresses. The technical team moves and transforms records, but it should not silently decide which business value is correct.
For every data set, record the source, target, owner, cleansing rule, transformation, validation test, and sign-off evidence. Also identify dependencies. Loading an order before its customer and product records are ready can create errors that later appear to be application defects.
2. Validate migration in five practical steps
Step 1: Profile the source before cleaning it
Count records and examine missing fields, duplicate identifiers, invalid dates, inactive records, and unexpected currencies. Distinguish an empty value from a genuine zero. Confirm which system owns each field when multiple files contain conflicting versions.
Prioritize defects by business impact. An incorrect supplier bank detail is more serious than an optional marketing field. Use approved records to confirm sensitive payment information and preserve the verification trail. Never use a plausible-looking value to fill a gap.
Step 2: Approve mapping and transformation rules
Write down how source fields become target fields. Specify date formats, rounding, currency treatment, product units, and codes. For example, confirm whether a carton becomes one inventory unit or twelve individual units. A technically valid import can still misstate stock by a factor of twelve.
Test difficult cases early: credit notes, part-paid invoices, discontinued products, cross-border customers, and intercompany records. These cases often reveal assumptions that a simple customer sample will miss.
Step 3: Reconcile counts and financial control totals
Compare the extracted population with records loaded, rejected, excluded, and still unresolved. Then reconcile relevant monetary totals by entity and currency. Keep the cut-off consistent. A source balance from Friday cannot fairly be compared with a target balance that includes Monday’s activity.
Expected target population = source population − approved exclusions − approved duplicate removals. Document each subtraction. Do not treat an unexplained mismatch as an acceptable technical variance.
Step 4: Run complete business scenarios
Ask users to follow real workflows in a test environment. Create an order, deliver it, invoice it, apply a receipt, and review the ledger. For purchasing, receive stock, match an invoice, and check the proposed payment. Include returns and corrections as well as clean transactions.
Record the expected result and the observed result. A passing screen test does not prove that downstream reports are correct. Check customer aging, inventory quantities, and financial postings after each scenario.
Step 5: Rehearse the final cutover
Measure extraction, transformation, loading, reconciliation, and approval time using a realistic rehearsal. Include the final changes that occur after your first data extract. Decide how the business will capture transactions during any system freeze.
Microsoft’s cutover guidance emphasizes rehearsal and readiness before switching systems. Your own plan should identify checkpoints, owners, communications, and a practical recovery decision before the deadline arrives.
3. Worked example: a clean record count can hide a balance error
Illustrative example, not a client case. A company extracts 2,000 customer records. It approves removal of 120 duplicate records and 80 inactive records with no relevant open activity. The expected target is 2,000 − 120 − 80 = 1,800 records.
The target contains exactly 1,800 records. That check passes. However, open receivables total €450,000 in the source and €438,500 in the target. The difference is €11,500. Correct master-record counts do not explain missing or altered financial transactions.
| Check | Expected | Observed | Decision |
|---|---|---|---|
| Customer records | 1,800 | 1,800 | Pass count check |
| Open receivables | €450,000 | €438,500 | Investigate €11,500 |
| Payment-term sample | Agreed source terms | Three unexplained changes | Resolve mapping |
| Order-to-cash test | Correct invoice and ledger | Posting account incorrect | Fix configuration and retest |
Investigate the receivables difference at invoice level. Possible causes include an omitted credit note, a currency conversion error, or an incomplete extract. Do not post a balancing adjustment solely to make totals agree. Identify the cause, correct the rule, reload as needed, and preserve the evidence.
Agree acceptance criteria before testing. Some rounding differences may be explainable; a missing invoice or altered bank record needs a different response. One global percentage threshold can hide serious defects inside a large total.
4. Make the go-live decision visible and accountable
Your ERP data migration checklist should finish with a signed decision record. Separate critical blockers from accepted limitations. Give each remaining issue an owner, workaround, deadline, and business impact. An issue without an owner is still a risk, even when the technical team labels it minor.
- Migration scope and exclusions are approved.
- Record counts and control totals reconcile at a common cut-off.
- Critical fields and exceptions have been checked.
- Users can complete essential end-to-end transactions.
- The final migration fits the tested cutover window.
- Support and recovery responsibilities are clear.
Continue checking after launch. Compare the first invoices, receipts, supplier payments, and inventory movements with the agreed expectations. Monitor rejected integrations and duplicate transactions. A successful weekend does not remove the need for operating controls.
For a broader implementation roadmap, read AB Consulting’s ERP development and implementation guide. If systems must exchange property data, explore the ERP and property-system integration article.
Prepare a migration review with AB Consulting
Need an independent view of data readiness, reconciliation, or your cutover plan? Contact AB Consulting to discuss an operational deep-dive. Share your current systems, planned launch date, entities, and the business transactions that must work on day one.
Prepared by AB Consulting. Worked figures are original illustrations. Official references checked in October 2026.
