SPRIG STUDIO

Fix duplicate SKUs and missing photos in a product CSV

A valid CSV can still contain the wrong product identifiers. Use this three-row exercise to separate table errors from photo-matching errors.

Written by the Sprig Studio team · Checked against the importer on October 10, 2026 · Fictional products and sample prices.

1. Start with a deliberately broken sheet

Download the practice CSV with errors and open it in the free CSV checker. The checker reads your file inside your browser, without an account or file upload. This example should flag the repeated CAN-001 SKU and a missing title. Do not use this deliberately broken file as your production product list.

2. Fix each product identity

The first row represents Amber Jar Candle; the second represents Sage Jar Candle. Give them distinct identifiers, CAN-001 and CAN-002. Restore the third title to Linen Jar Candle. Keep these identifiers unchanged in future campaigns. Treat numeric-looking SKUs as text in your spreadsheet so leading zeros are not lost.

3. Check the filename against the actual photo ZIP

The third row deliberately references CAN-999.png. The matching photo ZIP contains CAN-003.png instead. The free checker cannot detect this mismatch because it does not inspect your photos. Correct the filename, then import the sheet and photos into the studio to verify the association.

In filename mode, front.png searches by basename and can be ambiguous if several folders contain that name. A full relative path such as candles/front.png narrows the match. In SKU mode, the filename without its extension must match the SKU; CAN-001-front.png does not match CAN-001. Matching ignores letter case, but it does not guess which variant you meant.

4. Compare with the repaired file

Download the corrected CSV. It has three distinct SKUs, three titles and filenames found in the sample photo ZIP. A clean table does not prove the prices or photos are commercially correct: open each preview and check the jar label, size and displayed price.

When the import still fails

Use unique, non-empty column headers and map them in the studio. Save CSV as UTF-8 or use XLSX. Import 1–200 product rows per batch. Product photos must be JPG or PNG; a filename ending in PNG is not enough if the file is another format. Password-protected ZIPs are not supported. Do not share account recovery codes when asking for help.

Follow the full import workflow → · See the finished example →