What to do when an upload, mapping, or validation step fails, and how ambiguous dates are handled.
Most import problems are one of four kinds: the file won't upload, a column
won't map, values fail validation, or the import itself misbehaves. Find the
symptom below; each fix is something you can do yourself.
Upload problems
| Problem |
Fix |
| File too large |
Split into files under 10,000 rows each |
| Invalid format |
Re-save as "CSV UTF-8" from Excel |
| Upload fails |
Check your connection, then try a smaller file |
Mapping problems
| Problem |
Fix |
| A column didn't auto-map |
Pick the field from the Map To dropdown yourself |
| A custom field is missing from the dropdown |
Check the field's code in Settings → Custom Fields |
| A required field shows red |
Map one of your CSV columns to it |
Validation problems
| Problem |
Fix |
| Invalid email |
Correct the format: user@domain.com |
| Invalid date |
Use ISO, European, or American format consistently |
| Invalid Record ID |
Re-export to get fresh IDs |
| Unknown select option |
Use the exact option label from the field's settings |
In the review step you can also Fix a value once to correct it across
every row that has it, or Skip a value. The rows still import, with that
one field left empty.
Ambiguous dates
The wizard reads three date formats and detects which one a column uses:
| Format |
Pattern |
Example |
| ISO |
YYYY-MM-DD |
2024-05-15 |
| European |
DD/MM/YYYY |
15/05/2024 |
| American |
MM/DD/YYYY |
05/15/2024 |
When a day value is greater than 12 the format is unambiguous, so
31/01/2024 can only be European. When both positions are 12 or less
(01/02/2024, January 2 or February 1?), the wizard shows a warning and a
format dropdown on that column: pick European or American to say
which reading is correct. ISO dates are never ambiguous, which is why they're
the safest format to export from other tools.
Import problems
| Problem |
Fix |
| Stuck at "Processing" |
Large imports run in the background; check back in a few minutes |
| Some rows failed |
Download the failed rows, fix them, re-import just those |
| Unexpected duplicates |
Map a unique column (email, domain, or ID); see Update existing records |
Good to know
- There is no undo. Test with a 5-10 row sample before the full file.
- A failed import keeps its successful rows. Fix and re-import only the
downloaded failures.
- One record type per file. Importing people can still auto-create their
companies if you match the Company relationship by Name.
- Every run is logged. The Import History page under your workspace
name shows created, updated, skipped, and failed counts for each import.