Changing management software without re-entering data by hand

What the company had to do

A company was moving from an old management system to a new one. Alongside it stayed the company's ERP, where orders and items are kept.

Why the migration risked never ending

Once it was live, the new system had to be kept in line with the company's ERP by hand: every closed job had to be copied across to the other side. On top of that, years of master data, orders and history had to be moved.

How I moved the data and connected the two systems

I extracted and imported the data with scripts that can be re-run, and connected the new system to the company's ERP in both directions.

• Extraction with scripts. Master data, orders and history come out of the old system through scripts.
• Repeatable imports. They can be re-run as many times as needed, without creating duplicates.
• The client's corrections, without deleting. I set up the import of the master data corrected by the client so that it updates the data without deleting anything.
• The automatic return. When each job is closed, the data goes back to the company's ERP on its own, through a fixed-IP connection.
• Reading orders and items from the company's ERP. Orders and items arrive in the new system.
• Quick access. Anyone working from a shared workstation gets in instantly, without typing a password.

What has changed

• The new system went live with the historical data.
• Synchronisation starts on its own, without manual work.
• 100% of the client's requests are closed.
• Multi-line orders, a case that came up during the work, were fixed straight away.

What you can take away

• A migration is only finished when the data flows on its own.
• Write imports so they can be re-run: the first time is never the last.
• When you import data corrected by the client, update the existing records instead of deleting them.
• Test the less common cases straight away, such as multi-line orders.

Details have been changed to protect the client's confidentiality.