Migrating data to Odoo from another system is often the most underestimated part of implementation. It looks simple: export CSV, import in Odoo. In reality, mapping, cleaning, and validation take more time than everything else.
Planning
Document all data sources: customers, products, orders, invoices, inventory history. Decide what actually needs to be migrated. Old order history? Maybe, maybe not. Closed projects? Probably not. Migrating everything "just in case" costs time and creates noise.
Data cleaning
Clean source data before migration. Duplicates in the customer register, products without category, invoices with incorrect amounts. It's easier to clean in a spreadsheet than in Odoo after import. We typically spend 40 percent of migration time on cleaning.
Mapping
Map source fields to Odoo fields. Customer's "Company Name" maps to res.partner.name with is_company=True. Product's "Article Number" maps to product.template.default_code. Document the mapping in a spreadsheet. It becomes your reference throughout the process.
Import
Odoo's built-in import function (CSV) works for simpler imports. For complex data: write a Python script with XML-RPC or ORM. Import in the right order: contacts first (vendors, customers), then products, then orders and invoices (they reference contacts and products).
Validation
Spot checks: manually verify 50 random records. Totals: does total inventory value match? Does accounts receivable match? Relations: do all orders have a linked customer? Do all products have a category?
Do a test import in a test database. Fix errors. Do another test import. Repeat until it's correct. Only then run on the production database.