Snapchat catalog feed for Shopify

In Ads Manager, a catalog's data feed is a hosted file Snapchat fetches from a URL (HTTP, HTTPS, FTP or SFTP) on a refresh schedule of hourly or daily, at the time of day you pick. Accepted formats are CSV, TSV, TXT and XML in Google's RSS/ATOM shape; gzipped files are fine, zipped ones are not, and a file can hold up to 10 million products. If the file needs a login, the credentials go into their own form fields; Snapchat asks you never to put them in the URL.

Feedflock renders the TSV (a header row of Snapchat's column names, one tab-separated row per variant) and hosts it on a stable URL, regenerated every hour. You paste the URL once and never upload again. A manual CSV upload exists for small catalogs, up to 500 MB and recommended under 50 products, but a hosted feed is the way to keep a real catalog current.

Snapchat says you can use a catalog file you already maintain for another platform, as long as it adheres to its specification, and the vocabulary is Google's. The differences are in the details. The required set is id, title, description, link, image_link, availability and price, plus any one of brand, gtin or mpn. Where Google wants a brand and a barcode, one identifier is enough here. Availability takes five values (in stock, out of stock, preorder, discontinued, available for order), and there is no backorder. Sold-out products stay in the feed: Snapchat excludes them from serving itself.

Price is lenient (“199 USD” or “USD 199”, zero allowed, the currency optional only when it's USD), and the template always appends the ISO code, the safe side of the rule. google_product_category is optional but recommended for every product, at least three levels deep, because it informs Snapchat's ranking models; every example in the spec is the English path, and that is what the template sends.

There is no product database to pair with on Snapchat; the pixel is the join. Dynamic Ads retargeting matches the item_id your Snap Pixel sends from product pages to the id column in the feed, or, if you match on groups, item_category to item_group_id. That makes the id choice a decision to make before the first import: the template's default, the numeric Shopify variant ID, is unique and stable, and a duplicate id can fail the whole catalog upload.

Two more things live outside the feed's columns. Attaching a pixel to the catalog is required for Dynamic Ads for Web. And URL parameters (your UTM tracking) must be included on the links in the catalog before upload; Feedflock appends them when the file is served, so the links the pixel sees and the links the ads use carry the same parameters.

Catalog Diagnostics splits issues into errors (problems on required fields, which block the specific product) and warnings on optional fields, which don't. The downloadable report lists every feed error by id, so you can fix row by row. Media failures are the exception: a failed image never appears in the report, only on the diagnostics screen, so a product showing in ads without its photo is the telltale.

The image rules: at least 200 × 200 px, no larger than 5 MB, no GIFs. The field list allows WebP, JPEG and PNG, but the diagnostics page lists JPEG and PNG as the supported formats. Keep the gallery to those two and you are safe on both pages. And let Snapchat's fetcher through: it downloads images with its own user agent, and a blocked agent shows up as AUTH_ERROR or NOT_FOUND.