Why do commas and quotes in my product copy break the CSV import?

Locate the broken boundary

A fictional tray description says: Oak finish, three compartments. Another sentence includes the nickname “Desk Friend.” A manually assembled comma-separated row can split that description into extra fields. Removing every comma would change the prose without addressing how the file represents its fields.

RFC 4180 documents a common CSV quoting convention: fields with commas, quotes or line breaks are enclosed in double quotes, and an internal quote is doubled. Platform import rules remain the destination authority. Export from a spreadsheet or a known CSV writer rather than joining arbitrary copy fragments with commas.

Inspect a small punctuation specimen

Keep HTML encoding, CSV quoting and visible punctuation separate. A product description can contain quotation marks even when its serialized CSV needs extra escaping. Never approve the file merely because its raw text looks like the final sentence.

Here is a small syntax specimen, not a Shopify template. The intended description is Oak finish, called "Desk Friend". Its comma-separated representation with a fictional identifier is:

example_id,description
tray-01,"Oak finish, called ""Desk Friend""."

After parsing, there must be two fields in the data record. The doubled straight quotes are serialization, not extra visible quotation marks in the approved sentence. Curly quotation marks can be legitimate copy inside a field; they cannot replace the straight double-quote characters that delimit a CSV field. Do not globally remove or replace approved punctuation to repair a boundary error.

Run an export and reopen comparison

1. Place the approved specimen in one authorized pilot record.
2. Export using the actual required delimiter and encoding.
3. Reopen the saved file in an appropriate CSV-aware editor.
4. Compare the header count, parsed fields, record count and exact description.

This fictional punctuation exercise is not an import-ready Shopify template. Use your store’s current exported column schema for an actual update. If the destination rejects a multiline field, adapt the publishing format deliberately and retain the same facts; do not claim that every importer follows the same interpretation.

For a second specimen, put an actual paragraph break inside the quoted description field and reopen with a CSV-aware parser. Expect one product data record containing that break, even though the file occupies more physical lines. Record parsed field counts and the exact resulting description, not just a line count. A correct local round trip demonstrates this serializer/parser pair; it does not certify a destination importer.

If the description contains HTML such as an anchor, its attribute quotes still need CSV serialization around the complete field. CSV escaping does not validate HTML, sanitize a link or prove that the destination preserves that markup. Check the saved visible description separately.

Separate a syntax repair from copy editing

Record the offending field and the import error before changing anything. Shopify’s troubleshooting page distinguishes quote-related errors; read the current message instead of assuming every failure is a model-generated wording problem. Test the corrected pilot and inspect the saved product description.

A complete result preserves the approved text and the intended record boundaries. If a paragraph break is intentionally removed for a specific destination, have an editor approve the resulting sentence flow. CartVisual does not validate external CSV imports. Keep the original copy, exported file and inspected result together so the next handoff does not reintroduce a manual quoting mistake.

Preserve the exact importer message and file version. Encoding damage and an unmatched delimiter may produce similar symptoms but require different repairs. Follow the UTF-8 round trip if characters themselves change, and the record-identity check before an authorized retry. No CSV was imported into a store for this article.