What is data conversion?

Suppose your accounts have been in the same package for nine years. Your account numbers follow a layout that once suited the company, items have codes with a logic of their own and some relations appear twice in the system. If you load that data unchanged into a new package, you take every quirk along. That is why there is an intermediate step.

In that step the data is translated. The chart of accounts is mapped to a standard such as RGS, so the layout becomes recognisable. Relations are merged or left out if they are no longer active. Items get the fields the new system expects. Then you check the result before anything becomes final.

A good conversion happens in a test environment. You put the balances, the open items and the stock next to the figures from the old package. If something is wrong, the conversion is adjusted and repeated, without your real administration being touched. Only when the trial is right does the final transfer follow.

Conversion is not the same as data migration, although the words get mixed up. Data migration is the whole process of moving data. Conversion is the translating inside that process. In a switch you see both, and usually one of them is used for the whole.

Be aware of what a conversion does not do. It does not turn bad data into good data. If your customer file is full of duplicates, someone has to decide which one stays. That is why, in our switching service, we agree beforehand what is converted and what is not. Cleaning up and data from several sources are not included automatically; we discuss them beforehand.

When the conversion is done, your team starts with an administration that matches the opening balance. That is the moment the new system starts counting.