Fix an import without guessing which rows arrived
Diagnose concrete CSV header, quoting, column, date, amount and mapping errors, then safely retry a complete sales import.
Import errors fall into three stages: the file cannot be parsed, its columns cannot be mapped or one of its sales values fails validation. Read the stage and row number before editing a working copy. The importer rejects a file containing invalid rows as a whole, rather than silently saving the valid-looking remainder.
The useful part
Correct the reported cause, preview the complete file again and reconcile its totals before importing. A file error and a later commission-review question are different problems.
When the file cannot be read
“Include a header row and at least one sale” means the parser did not find both parts. “Every column needs a unique, non-empty header” points to blank or repeated heading cells. “Invalid CSV quoting” and “A quoted field is unfinished” usually need a fresh CSV export or a repaired quoted value. Preserve the source file and work on a copy. A spreadsheet workbook is not CSV simply because its filename extension has been changed.
When one row has the wrong number of columns
An error such as “Row 7 has 9 columns; expected 8” identifies a structural mismatch. Inspect that row and its original headers, including commas inside item descriptions or amounts. A comma-containing value must be quoted correctly in a comma-separated file. The parser also supports semicolon and tab separators, but every data row must match the detected header width. Remove report notes or subtotals only from the prepared working file, keeping the untouched report for reference.
When a mapping cannot be applied
“Map the Earning staff member column” asks for the actual earning-person source field. “Each source column can only be mapped once” means two app fields point to one heading. A saved mapping that reports missing mapped columns cannot be applied until those selections are corrected. If several profiles match the headers exactly, select the intended one and choose Apply mapping. Use Not in this file for unavailable optional fields. When mapping-name duplication is reported, apply that existing mapping before updating it, or choose a different name.
When dates or amounts fail validation
Date errors distinguish unsupported text from impossible calendar dates. Choose the correct DD/MM/YYYY or MM/DD/YYYY option, or use YYYY-MM-DD. Amounts must be numeric values with at most two decimals; formulas, dashes standing for blanks and extra decimal places are not accepted as money. A blank required amount is an error. A non-zero tax value must share the sale’s sign and cannot exceed its magnitude. Row numbers include the header row, making the first data row row 2.
| Input | Problem | Checked replacement |
|---|---|---|
| 31/09/2026 | September has no day 31 | Correct date from source |
| 110.005 | More than two decimal places | Source currency value |
| Sale −110; tax +10 | Opposite signs | Verify refund tax −10 |
| Amount =100+10 | Formula, not a value | Export calculated value 110.00 |
When the file is too large
The supported import limit is 8 MB and 15,000 sales. Split a larger source into clearly defined periods or other non-overlapping groups, preserving headers in each file. Record the scope and total of every part so the split remains checkable. Adding a later file can overlap existing draft rows; replacing the current draft removes its editable rows. Choose deliberately, compare coverage and do not split a file merely to hide a validation error in the remaining section.
Know whether to correct, review or reload
“No rows were imported” means correct the file and rerun its preview; the displayed errors may be only the first few. Missing flat rules, dated staff identities or agreement dates, refunds and duplicate warnings after a successful import belong to review. If saving fails or another tab changed the workspace, preserve attempted changes when offered and reload the saved state before retrying. Inspect Imports and Sales to check what arrived. To undo a confirmed draft import, use Remove draft rows and its counted confirmation. Closed rows are kept; older filename-grouped imports remove all matching remaining draft rows.
Put it to work.
Inspect a working CSV before adding its rows to a business draft.
Check the file structure