Content Planning
Tutorial

Import a CSV Without Changing Its Values

Editorial responsibility: SearchCub

Created: Oct 9, 2026, 4:42 PM UTCEdited: Oct 9, 2026, 4:42 PM UTC

Import identifiers and text as text before loading a CSV, then compare the resulting values with an untouched source copy. A column that looks numeric may contain a code whose exact characters matter more than its numeric value.

That matters when a sheet supports your blog topic plan: changing a resource identifier can break the connection between a row and the file you meant to use. The same precaution applies to article lists, research notes and other files you work with before writing.

Decide What the Columns Mean

A CSV stores rows of fields. Your spreadsheet still has to decide how to interpret those fields. Microsoft's Excel guidance, checked on October 8, 2026, describes conversions that can remove leading zeros or change large numbers.

Start by naming which columns are identifiers, which are literal text and which genuinely need arithmetic or date operations. Keep identifiers as text even when every character is a digit. A monthly-looking code is also text if its job is to identify a file rather than represent a reporting date.

Keep the original file separately from your working sheet. If you have already saved changed values back over your only copy, a later formatting change cannot establish what the original contained.

Use the Filled Teaching File

Download the teaching CSV and its completed expected-value map. They contain three invented article records; the names, codes and identifiers are teaching inputs.

Every column in this sample is intended as Text. The table highlights values worth checking; the companion file supplies the complete values for every row.

Column	Exact teaching value	What to preserve
`article_id`	`00037`	The leading zeros are part of the identifier.
`resource_id`	`1234567890123456789`	Every digit belongs to a literal identifier.
`date_like_code`	`03-04`	This is a code, not a date to interpret.
`article_title`	Café shelf labels	The accented character belongs to the title.
`notes`	Keep "spare" on the label.	The quotation marks are part of the field's text.

The CSV also includes a title containing a comma. RFC 4180 describes quoting fields containing commas or quotation marks and doubling a quotation mark within a quoted field. Those file characters keep a comma inside one field; they are not instructions to delete punctuation from the title.

The download is UTF-8 without a byte-order mark, or BOM: an initial encoding marker. Its encoding is known because we created the file that way. A different file needs its own encoding information.

Import Through the Text/CSV Preview

The following path follows Microsoft's Power Query Text/CSV documentation, checked on October 8, 2026. It documents Excel among the supported products and notes that capabilities can vary by host. These instructions were checked against the documentation; an Excel UI import was not performed.

Start a Text/CSV import through Get Data, select the downloaded file and open its preview.
Set File Origin to the file's known character set. For this teaching file, choose UTF-8. Check that the accented title appears correctly.
Choose comma as the delimiter. Confirm that the header and each record occupy the intended columns.
Turn off automatic data type detection so the columns default to Text. Do this before continuing; inferring types from more rows is still inference.
Choose Transform Data. Confirm the column types and exact values before loading the result into your sheet.

Microsoft's connector documentation says UTF-8 is inferred when a file starts with a UTF-8 BOM. This teaching file has none, so select its known encoding explicitly rather than relying on detection.

For a real file, retain the text version first. Convert a separate, clearly named working column only when its meaning requires a number or date, with the appropriate units or date convention.

Compare Values Before Using Them

Use the completed map alongside the imported sheet. Compare the full identifier and code strings, not just whether a column looks tidy. Check the titles and notes as well: correct digits do not establish correct character encoding or column boundaries.

For this download, a local CSV-parser check recovered all three records and every supplied string, including the leading zeros, long identifiers, accent, comma and quotation marks. Writing those parsed fields back with the declared CSV format reproduced the original file bytes. That check establishes the file's contents and roundtrip under the tested parser; it does not establish an Excel import result.

If a value has changed in your sheet, stop before using that row. Return to the untouched CSV and redo the import with explicit types. Microsoft also notes that formatting cannot restore leading zeros already removed, and Excel's numeric precision is limited to fifteen significant digits; a long identifier should remain text from the start.

Once the values agree, keep the original and the import choices with your notes. If the sheet contains topic ideas, continue with keyword grouping and page mapping. When you choose an article, transfer the relevant checked inputs into its writing brief.