A customer ID loses its first two zeros. A delivery date changes meaning. A name with an accent turns into broken characters. Your spreadsheet looked fine before export, but the CSV now tells a different story. These problems often happen because a workbook, a text file, and an importing app each handle data differently. A careful export protects the values, and a careful import keeps them usable.
A reliable CSV export preserves both the text and its intended meaning.
- Agree on the delimiter, encoding, and date format before exporting.
- Treat identifiers as text, especially when they contain leading zeros or many digits.
- Keep the original workbook and validate the exported file in its destination.
Understand What CSV Can Actually Preserve
A CSV file stores rows of text divided into fields. It does not carry the full structure of an Excel workbook. Colors, charts, cell comments, merged cells, validation rules, and multiple worksheet tabs do not survive as workbook features.
Formulas usually become their current results during a standard CSV export. The receiving app gets a value, not a calculation it can update. Refresh external data and check formula results before exporting. Otherwise, a perfectly readable CSV may contain stale figures.
CSV also lacks built-in column types. A field containing 00127 has no attached instruction saying “keep this as text.” An importing program may decide it is a number and display 127. That change can happen even when the exported file contains every original character.
Keep your XLSX file as the master copy. Treat CSV as a delivery format for one rectangular dataset. Export separate worksheets into separate files when needed, and use stable identifiers to preserve relationships between them.
Choose a Delimiter That the Destination Accepts
The delimiter marks where one field ends and the next begins. A comma is the usual choice for CSV, but some workflows expect semicolons. Tab-separated files use tabs and are normally described as TSV.
Regional settings can affect spreadsheet export behavior. A system that uses commas for decimal values may expect semicolons between fields. Do not assume the filename proves which separator is inside. Check the destination’s requirements and inspect the exported text.
If everything appears in one column after import, the receiving app may be using the wrong delimiter. Select the actual separator in its import settings before changing the file.
Let Quoting Protect Commas Inside Values
A company name containing a comma should stay in one field. CSV handles this with double quotes. Quotes within a quoted field are doubled. These CSV quoting rules also cover fields containing line breaks.
customer_id,company,note
00127,"River, Stone Ltd","Requested ""priority"" shipping"The second record contains three fields. The comma inside the company name belongs to the company name. The doubled quotes around priority represent literal quotation marks in the note.
Avoid replacing every comma with a semicolon in a text editor. That can change addresses, notes, and quoted content. Use a converter that reads fields correctly, then writes the requested delimiter.
Use an Encoding That Preserves Names and Symbols
Encoding defines how characters are stored as bytes. If an exporter writes one encoding and the importer assumes another, names, currency symbols, and non-Latin text can become unreadable.
UTF-8 is a sensible default for modern data exchange because it supports a wide range of writing systems. In Excel versions that offer it, select CSV UTF-8. If your export menu only says CSV, verify the resulting encoding instead of assuming it is UTF-8.
Check representative text after export. Include accented names, an address in another script, and any symbols your dataset actually uses. Plain English letters alone cannot reveal many encoding problems.
Some programs use a UTF-8 byte order mark, often called a BOM, to recognize the encoding. Other importers expect UTF-8 without it. Follow the receiving system’s requirements. If the first header acquires unexpected characters or fails to match, investigate encoding and BOM handling.
Protect Leading Zeros Before They Disappear
Postal codes, account references, product codes, and telephone numbers are identifiers. Arithmetic is rarely useful for them. Store these columns as text from the start.
Formatting an existing numeric cell as Text does not restore zeros that have already disappeared. Check the underlying values. If an identifier has a known fixed width, you can create a helper column with =TEXT(A2,"00000") to produce five characters. Use this only when five characters is a confirmed rule.
Custom number formatting can make 127 appear as 00127 inside Excel. Appearance alone is not a dependable preservation strategy across export tools. Create explicit text values in your export copy, then inspect the CSV.
Long identifiers need extra care. Excel retains only 15 significant digits of numeric precision. If a longer code was already stored as a number, some digits may have changed before export. Converting that damaged value to text cannot recover the original. Retrieve it from the source and import the column as text.
Wrapping an identifier in CSV quotes does not guarantee that a spreadsheet will keep it as text. The importing app still controls type detection. Avoid adding formula wrappers to force a display, since that changes the field’s content and can cause compatibility problems.
Make Dates Clear Across Regions
A date written as 04/05/2026 could mean April 5 or May 4. Both interpretations are plausible. Use YYYY-MM-DD for date-only fields when the receiving system accepts it. For example, 2026-05-04 leaves little room for confusion.
In an export copy, a helper formula such as =TEXT(A2,"yyyy-mm-dd") can create consistent text from a valid Excel date. Inspect existing text dates separately. A cell that looks like a date may already contain an ambiguous string.
For timestamps, agree on a format and time zone. A value such as 2026-05-04T14:30:00Z identifies UTC. Use the Z suffix only when the time truly represents UTC. An ordinary Excel date-time value does not carry a time zone automatically.
Check decimal values too. Agree on the decimal separator and whether thousands separators are allowed. A destination expecting 1234.50 may reject or misread 1.234,50.
Build an Export Copy With Clear Boundaries
Before saving, create a copy containing exactly the records and columns intended for delivery. Remove report titles above the header row, unnecessary totals, and decorative blank rows. Give every column a clear, unique name.
Do not assume hidden columns or filtered-out rows will be excluded by every export method. Create a dedicated export range or worksheet and verify its contents. This also helps prevent internal notes or personal information from entering a shared file.
Watch for distinctions that flat text cannot explain on its own. An empty field might mean missing, unknown, or intentionally blank. If those meanings matter, agree on separate fields or a documented missing-value convention. Avoid casually replacing blanks with zero.
Follow a Repeatable Export Sequence
- Save the source workbook. Keep its formulas, worksheets, and formatting available for future work.
- Confirm the destination’s requirements. Check delimiter, encoding, headers, date format, decimal separator, and missing-value rules.
- Prepare the export copy. Make identifiers explicit text, standardize dates, and remove content outside the intended dataset.
- Export the correct worksheet. In Excel, use Save As or Save a Copy, depending on your version, and choose the required CSV format.
- Inspect the raw text. Check headers, separators, quotation marks, zeros, dates, and representative characters.
- Test the actual import. Select column types where available and compare the imported records with the source.
If you prefer a browser workflow, an XLSX to CSV converter offers another route. Check its worksheet selection and output settings before relying on the result. For private or regulated records, confirm that its data-handling practices meet your requirements before uploading. Use a local export when external processing is unsuitable.
Check the Output Without Triggering New Changes
Double-clicking a CSV in a spreadsheet can trigger automatic interpretation. Dates may be reformatted, identifiers may become numbers, and long values may appear in scientific notation. That makes visual checking less reliable.
Start with a plain text editor. Then import the file through the receiving application’s import controls. In Excel, this commonly means using Data > From Text/CSV and setting identifier columns to Text before loading. Available controls vary by version.
Use this final check before sharing:
- Compare parsed record counts and field counts with the intended export.
- Check identifiers containing leading zeros and long digit sequences.
- Inspect dates, timestamps, decimals, and negative amounts.
- Verify names, symbols, quoted text, and multiline notes.
- Compare a useful total and review the first and last records.
Count parsed records rather than physical lines when fields contain line breaks. One customer note can span several text lines while remaining part of a single record.
A matching total helps catch missing records, but it cannot prove every field is correct. Pair totals with targeted checks of risky columns. If a file looks wrong only after import, compare its raw text before exporting again.
Send a File the Next Person Can Trust
A successful export gives the next person enough information to read the data correctly. Send a brief format note alongside the CSV. State its encoding, delimiter, date convention, identifier columns, and meaning of blank fields.
Keep the workbook, export copy, and delivered CSV clearly named. Your source preserves the working dataset. Your checked CSV provides a portable snapshot. With both available, a routine format conversion stays routine, and the values arrive with their meaning intact.