Happy N Yummy · Practical data troubleshooting
If each CSV row appears in a single Excel cell, import an unchanged copy and choose the file’s actual delimiter in the preview. Check the columns before loading. A one-column result is a parsing problem; changing the font or column width will not split it.

The workflow below targets desktop Excel builds with Data → From Text/CSV. Menu availability varies by edition and platform. The examples were checked with Python’s CSV parser; the Excel UI steps were checked against Microsoft documentation, not executed in Excel for this article.
First identify which problem you have
| What you see | First setting to inspect | Success check |
|---|---|---|
| The entire row is in column A | Delimiter | The preview has the expected column count. |
| A description splits at its internal punctuation | Delimiter and quotation structure in the original file | The description remains one field. |
| Columns are correct but letters are garbled | Encoding / File Origin | Known names or words match the source. |
| Columns are correct but an ID loses zeros | Data type conversion | The exact identifier text survives. |
These can occur together. Fixing the column boundaries does not prove that the characters, identifiers or decimal values were preserved.
A three-column practice file with two traps
Copy the block into a plain-text file called delimiter-practice.csv. It contains invented items and uses semicolons between fields. The comma in “Cup, blue” belongs to a description; the semicolon in “Tray; small” is inside quotation marks.
item;description;amount A17;"Cup, blue";12,50 B08;"Tray; small";7,25
Expected result, keeping every column as text at first:
| item | description | amount |
|---|---|---|
| A17 | Cup, blue | 12,50 |
| B08 | Tray; small | 7,25 |
There should be three columns and two data rows. The quote characters mark a field boundary convention; they are not part of either displayed description. A simple “split at every semicolon” operation would mishandle the second description. This fixture and its expected values were verified using Python’s standard CSV reader with a semicolon delimiter.
Import with a visible preview
- Keep the original CSV unchanged. Open a blank Excel workbook.
- Choose Data → From Text/CSV and select the copy. Microsoft documents this route in its text-file import guide.
- Set the delimiter to the character that separates fields. For this fixture, choose Semicolon.
- Use the preview to check the three headers, both descriptions and both amounts. If the preview is wrong, correct the import settings before loading.
- Where available, turn off automatic type detection for this practice import. The Power Query Text/CSV documentation describes that option. Load or transform only after the preview agrees with the expected table.
For real files, inspect several raw lines in a plain-text editor. A comma somewhere in a row is not enough to prove that commas are the separators. Compare the header and several records, paying attention to quoted fields. If the file contains inconsistent field counts or broken quotes, ask the exporter for a corrected file instead of forcing a delimiter that makes only the first row look right.
Why the amount column stays as text for this check
12,50 deliberately uses a comma as a decimal mark. Our first checkpoint verifies the string, not a numeric calculation. After the import boundaries are correct, decide which numeric locale matches the source and convert that column deliberately. Do not replace all commas in the file: that would also alter Cup, blue.
Likewise, adding a made-up header line or changing global regional settings may affect another tool that reads the same file. This workflow keeps the original bytes intact and makes the choice in the import step. If your Excel build lacks this interface, use its documented text-import workflow and the same expected-result table rather than assuming the menu names will match.
A final check before saving
- Count columns and data rows.
- Compare one punctuation-containing field against the source.
- Check a known decimal and a known identifier separately.
- Save your working workbook as a new file; retain the original CSV so a mistaken import can be repeated.
For identifier conversion, follow our leading-zero preservation guide. A delimiter fix and a text-type fix solve different parts of the import.
Verification: the synthetic semicolon sample, expected field values and wrong-tab-delimiter case were executed locally with Python. No Excel edition or operating-system UI was tested for this article. Documentation checked November 16, 2026.