Why your CSV opens in one column

Check the separator, preview the records, and export the delimiter you need.

A separator mismatch can look like missing columns

A file ending in .csv is not necessarily comma-separated. Some export systems use semicolons, tabs or pipes. If the importer expects commas and the file uses semicolons, a complete record may appear in a single column.

id;description;quantity
00127;"Notebook; blue";12
00904;Café mug;3

In this teaching example, each record has three fields. The semicolon inside the quoted description belongs to the description. Replacing every semicolon with a comma would change that value as well as the real separators.

Confirm the structure before changing it

1. Open CSV check & repair and load the original file.
2. Preview the detected separator. If the tool says the separator is uncertain, choose it explicitly rather than treating the suggestion as proof.
3. Confirm the number of columns and compare several values, especially fields containing punctuation.
4. Choose the output separator required by the system you are importing into, then download a new file.

Check the delimiter →

Use a preview as an acceptance check

For the sample above, the description should be Notebook; blue in a single cell and the quantity should be 12 in a separate cell. When exporting comma-delimited CSV, the original description’s semicolon remains part of its text. The parser, not a global find-and-replace, identifies field boundaries.

Excel’s text import tools allow you to choose delimiters and view a preview. You can use that workflow instead of converting the file. What matters is that the import settings agree with the file’s actual structure.

If different records have different widths

One record with three fields and another with four can indicate a missing quote, an extra separator or a genuinely irregular export. TableMender lists example record numbers and blocks the download. It does not pad records, delete extra values or guess which field belongs where.

Cases that need a different approach

A one-column source may genuinely contain one field per record. Delimiter detection has no basis to invent more columns. A proprietary export might also contain a preamble before the real header. Inspect the source and remove a preamble only when you know it is metadata rather than a data record. The current tool supports comma, semicolon, tab and pipe separators.

← All guides