If you have data outside the integrations (legacy CRM, partner exports, spreadsheets), backfill it via CSV. The importer maps fields, runs identity resolution against existing records, and integrates into Golden DB.
When to use CSV backfill
- Migration from another platform.
- One-time partner data dump.
- Legacy CRM data that won't be live-synced.
- Bulk corrections after a data audit.
Preparing the CSV
- Use UTF-8 encoding — avoid Excel's default Windows-1252 if special characters matter.
- One row per entity — don't combine multiple contacts into one row.
- Include a unique identifier if available — Salesforce ID, LinkedIn URL, internal ID.
- Use ISO formats for dates (YYYY-MM-DD) and country codes.
Step 1 — Upload
Data → Import → CSV. Drag the file. The importer detects the column headers and shows a preview.
Step 2 — Map fields
Map source columns to canonical fields. Required: at least one identity field (email for people, domain for companies). Optional: as many other fields as you have.
Step 3 — Set source priority
Set the quality score of this import relative to other sources. CSV imports usually rank lower than live API sources because they're snapshot-in-time.
Step 4 — Test and commit
Click Test → shows the first 50 records with proposed merge decisions. Review for surprises. If clean, click Commit. Large imports (>100k rows) run in the background; smaller ones complete in minutes.
