ERP data migration is the stage where projects quietly lose their history. It is also the stage where timeline expectations drift furthest from reality, because moving rows is easy and making them agree is not.
This post explains what actually drives the timeline, so you can plan realistically — and it mirrors how we run migration projects.
What is actually being moved
A migration moves more than transactions. It moves:
- Master data — customers, suppliers, items, warehouses, chart of accounts
- Open balances — receivables, payables, stock valuation
- Transaction history — what makes comparative reporting survive the cutover
Each of those has a different cost. Master data is the fiddly part. History is the heavy part.
The three real drivers of time
1. Data quality
Messy master data takes longer to clean than structured exports take to load. Duplicate customers, stale item codes and inconsistent supplier names must be resolved before the first trial load. Cleaning after go-live is archaeology.
2. History depth
A year of history is not the same as ten. Transaction volume drives extraction, mapping and reconciliation time. You usually only need two full years plus a read-only archive, but some estates and regulations require more.
3. Source complexity
A single supported accounting system is straightforward. Multiple systems, legacy databases or spreadsheet estates mean more field mapping, de-duplication across sources, and more reconciliation passes.
A realistic schedule
- Included in implementation — most standard migrations run alongside implementation, roughly two to four weeks, because the data team and the configuration team share the field map.
- Extended history — multi-source or very deep estates add a few weeks, priced separately.
- Unusual sources — legacy exports and bespoke databases need a paid discovery phase before a fixed price can be scoped.
What should not change
Regardless of scope, the migration is not complete when the data loads. It is complete when a variance report is signed off stating that the new system agrees with the old one at the cutover point.
Planning for yours
The timeline depends on your data, so the honest answer starts with a look at it. Our migration service covers extraction, cleansing, mapping, opening balances and read-only history with reconciliation at every step — and it is included in most implementations.