Do not migrate everything. Migrate the gap.

Why big-bang rollouts stall

The usual plan is to configure every module, import five years of history and go live on the first of the month. Three weeks in, the masters are half clean, the team is running both systems, and everyone quietly returns to the sheet.

The plants that succeed start with the one place money is leaking, and only then extend.

Week one, the commercial line only

Four things. Nothing else.

• Products and product families, only the ones you sold last quarter.
• One price book, with your current rates and slabs.
• Users, roles and approval thresholds.
• Open enquiries and live quotations, not closed history.

Week two to four, the plant side

Once quotations and orders are being created in one place, bring the Sales Admin review, production orders and dispatch on. This is the point where committed dates start being real, because sales and the plant are reading the same record.

Quality, inventory holds and the customer portal come after that, in whatever order your complaints suggest.

What to leave behind

Closed orders older than a year, dead leads, and any sheet whose owner cannot explain what it is for. History belongs in an archive file, not in your live masters.

Keep accounting where it is. Your ERP or Tally continues to invoice and keep ledgers, and pulls order and dispatch data from the new system.

The three mistakes

• Running both systems in parallel with no cut-off date, which guarantees the old one wins.
• Importing dirty master data and then blaming the software for wrong rates.
• Rolling out without an approval policy, so the price book is bypassed from day one.

Takeaway

Start where the money leaks, set a cut-off date, and let history stay history.