Import the spreadsheets vendors actually send.

Import the messy reality, not the ideal

Vendor catalogs come in every shape. Some have one row per variant, some have one row per product. Some use semicolons to separate image URLs, some use pipes. Some have option columns named "Size" and "Color," some name them "S" and "C," some don’t name them at all. Vendee Pro’s import is built for the spreadsheets you actually receive, not the spreadsheets you wish vendors sent.

Header auto-detect (no mapping UI required)

Vendee Pro reads your CSV header row and resolves each product field via a list of common column names. Examples: the product handle comes from handle, falling back to name, vendor_style_name, title, sku, vendor_style_number, or faves_item_id. Wholesale price reads wholesale_price, then buy_price, then cost. That fallback list is what lets a Shopify export, a FAVES export, and a vendor’s home-baked line sheet all import without a manual remapping step. Header names are case-insensitive and spaces are normalized to underscores.

Variant grouping by handle

For files with one row per variant, Vendee Pro groups rows by handle (or the first available fallback — name, style name, title, SKU). Shared product fields (name, description, vendor, category, currency, tags, notes) are taken from the first row of each group; per-row variant fields (SKU, barcode, wholesale price, retail price, options) are taken row-by-row. Three variant axes are supported via option1_name/option1_value, option2_name/option2_value, option3_name/option3_value (the Shopify-style format), with a variant_options JSON column as an alternative for tools that emit a single blob.

Image URL fetching

If the catalog has image URLs (column image_urls, images, image_src, or photo_links), Vendee Pro splits the value on semicolons, pipes, or commas that precede an http(s) URL, then fetches each image and attaches it to the product. Images are scaled down to a 2048 × 2048 max and converted to WebP at quality 85 (originals are not retained — the converted asset is what gets stored). Failed fetches are logged per row so a single broken URL never sinks the whole import.

Per-row error reporting

Imports succeed at the product-group level, not all-or-nothing. If 8 product groups out of 200 fail (missing handle, malformed price, image URL 404), the other 192 land successfully and the failed groups are reported with the row number and the specific reason. Fix those rows, re-upload only the failed slice, and you’re done. The job page shows the most recent 200 errors plus a final summary line.

Downloadable template

The import page ships a Default template you can download as a starting point. It demonstrates the column conventions described above (handle, option columns, image URLs) with sample rows for both simple and variant products. Use it as a reference when wrangling a vendor file into shape.

Vendee Pro’s export uses the same column set as the import, so a catalog you build in Vendee Pro can be downloaded, archived, or used as a starting point for the next vendor’s file.

Background processing

Imports run as background jobs with a 1-hour timeout. Upload your file (CSV, up to 50 MB), get a status row in the import list, and continue working. The page polls for progress (total rows, processed rows, success count, fail count) so you can see the import advance in real time.

Bulk import via the AI assistant

For vendor catalogs that aren’t in CSV form — a PDF line sheet, an XLSX with non-standard headers, a photo of a printed catalog page — the AI assistant accepts file attachments directly. Drop the file into a chat, and the assistant extracts a structured product list, shows you a confirm card with every line, and creates the records only after you approve. That path handles the messy formats the direct CSV importer can’t.

Frequently asked questions