A migration is the perfect moment to start with clean data, or to drag old clutter along with you. A good data migration runs in three phases: preparation, migration and quality control.
1. Preparation
Create a formal migration plan with timelines, dependencies and milestones, and appoint a project manager to steer the schedule. Only begin the actual migration once the Salesforce build is complete, the data has been received and the environment is ready for deployment. Schedule the migration for a period in which teams can give it sufficient attention.
Determine which data should come across
Identify which data needs to be migrated and what the authoritative source is. Not all business data comes from backend systems. Map out per department what is needed: sales (communication from calls and emails), service (cases linked to accounts), marketing (multiple sources) and finance (invoices and revenue recognition).
Data mapping
Give all records a unique ID across the various source systems, include related data IDs, and pull samples of metadata to understand the structure. Take into account the differences in complexity between the source and the Salesforce data model.
Cleaning the data
Remove duplicates and outdated records, resolve missing IDs with the help of users, and make clear who is responsible for the clean-up and within what timeframe.
Quality control
Carry out user acceptance testing with the business goals in mind, check whether data is correctly formatted and whether relationships are accurate, and repeat earlier phases if errors stem from the preparation.