Data migration
Data migration moves master records, documents and history from an old system into a new one. The hard part is not the copying but the decisions: what to move, how to convert it, and how to tell that the result adds up.
Migration runs in stages. Master data first: customers, item codes, price lists, trading terms. Then open items: unfulfilled orders, receivable and payable balances, stock levels as of the switchover date. History last, or not at all. Every stage is first run as a trial migration on a copy, usually several times, until the numbers on both sides meet.
For a while the old and the new system run in parallel, and this is where rollouts fall apart. Decide up front what is created in the new system from which date and what is finished off in the old one, and name the person who enforces it. The longer the two-track period lasts, the more people drift back to the tool they know.
Historical data does not have to move in full. A frozen read-only copy of the old system, or an export to files, is often enough, while only recent periods go into the new system. The boundaries are set by the statutory retention period for accounting records on one side, and by GDPR on the other, which requires deleting personal data you no longer have grounds to keep.
After the migration you need a reconciliation, defined before the switch: record counts, balances, stock value and quantities, turnover totals as of the same date in the old and the new system. Here is the risk signal: nobody in the company can say how many customer or item records should be there once the move is done.