The Universal Importer brings in data from any support platform or spreadsheet. Find it at Settings > Team > Import.
What You Can Import
- Tickets -- Full conversation history, statuses, priorities, tags, custom fields
- Customers -- Profiles with email, name, phone, sentiment, tags, metadata
- Articles -- Knowledge base content with categories, tags, publish status
Supported Formats
- CSV -- From any spreadsheet or platform export
- JSON -- Structured exports from most platforms (nested objects handled)
- JSONL / NDJSON -- One JSON object per line
- XLSX -- Excel and Google Sheets exports
- XML -- Structured markup exports
- ZIP -- Compressed archives containing any of the above
Limits
- File size: 50MB max
- Records per import: 10,000 max
- Larger datasets: Split your file into chunks and import each one. Save your configuration after the first run so field mappings, merge strategies, and match keys carry over automatically.
The 7-Step Process
1. Upload
Drag and drop your file or click to browse.
2. Detect
cStar reads the file structure and identifies the source platform. Built-in detection for Zendesk, Freshdesk, Intercom, Help Scout, HubSpot, Front, Crisp, LiveAgent, and Zoho Desk. Falls back to generic mode if the format isn't recognized.
3. Map Fields
Two-column layout with auto-suggested mappings based on field names. Accept, change, or skip each field.
4. Match and Merge
Choose how to handle existing records: match by email, short ID, external ID, slug, or custom field. Set per-field merge strategies.
5. Preview
See creates, updates, and skips with sample data before anything changes.
6. Execute
Real-time progress tracking with batch processing.
7. Summary
Results with successful creates, updates, and any errors.
Large Migrations (10,000+ Records)
- Split your file into chunks of 10,000 or fewer
- Import the first chunk through the full wizard
- Save the configuration when prompted
- For each remaining chunk: load the saved config, upload the file, preview, execute
With saved configs, each subsequent import skips detection and mapping entirely. A 50,000-record migration becomes 5 imports, each taking a couple of minutes.
Tips
- Export from the old platform first. Most platforms have data export somewhere in their admin settings.
- Start with a small test file (10-50 records) to verify mappings before committing to the full dataset.
- Original timestamps are preserved, keeping your analytics and SLA history accurate.
- Use a match key (email, ID) to prevent duplicates when importing across multiple batches.