File requirements, the 500-row limit, encoding, merge logic and every mapping option explained.
Supported files
- CSV - UTF-8 recommended. Hebrew/RTL content is fully supported.
- Excel (.xlsx) - the first sheet is used by default.
- JSON - an array of flat objects.
The 500-row limit
A single import is capped at 500 rows to keep AI mapping and deduplication fast and reliable. Options for larger books:
- Split the file into batches of 500 and import sequentially - merge logic keeps everything consistent.
- Use the REST API for programmatic bulk loading (see API guide).
- Order a done-for-you integration - our team migrates your entire book for a fixed price. Details here.
How mapping works
Every column in your file is matched to a Ricoberi field:
- System fields: name, email, phone, company, status, deal value, last contact date, next follow-up, notes, tags.
- Custom fields: any column you keep but don't map - stored as-is on the contact and visible on the contact page. Engagement columns (e.g. "last open", "SMS status") are also read by the scoring engine.
- Skip: columns you don't want at all.
Your mapping is saved automatically and re-applied the next time you import a file with the same columns (manage it under Mapping).
Duplicate handling
Before anything is written, each row is checked against your existing book by email first, phone second:
- Merge (recommended) - existing contact is updated; new info fills empty fields, and the import source is appended to the contact's association history.
- Create - every row becomes a new contact (use only for guaranteed-clean lists).
Common gotchas
- Emails with typos ("gamil.com") are auto-flagged with suggested fixes before import.
- Dates - use ISO (YYYY-MM-DD) where possible; common local formats are auto-detected.
- Empty rows / header rows in the middle of a file are skipped silently.