Shopify Store Products Scraper — Prices, SKUs & Variants

Scrape every product from any Shopify store through its public product feed: title, vendor, type, tags, images, and all variants with price, compare-at price, SKU and stock. No API key, no account, no browser.

Input

Field	Type	Default	Description
storeUrls	array	required	Store domains, store URLs, or collection URLs
flattenVariants	boolean	false	One row per variant instead of per product
maxProductsPerStore	integer	1000	Stop a store after this many products
maxPages	integer	20	Pages per store (250 products each)
maxConcurrency	integer	3	Stores in parallel

A collection URL scopes the scrape: https://www.allbirds.com/collections/mens reads only that collection (94 products in testing), while https://www.allbirds.com reads the whole catalogue (294).

Output

Product rows carry Shopify's own fields verbatim plus derived ones:

{
  "_input": "https://www.allbirds.com",
  "_source": "S1-products-json",
  "_scrapedAt": "2026-09-07T04:34:59Z",
  "storeUrl": "https://www.allbirds.com",
  "productUrl": "https://www.allbirds.com/products/mens-dasher-nz-anthracite",

  "id": 7369944137808, "title": "Men's Dasher NZ - Anthracite",
  "vendor": "Allbirds", "product_type": "Shoes", "handle": "…",
  "tags": [...], "images": [...], "options": [...], "variants": [...],
  "body_html": "…", "created_at": "…", "published_at": "…", "updated_at": "…",

  "variantCount": 13, "availableVariantCount": 11, "inStock": true,
  "minPrice": 140.0, "maxPrice": 140.0, "imageCount": 6
}

With flattenVariants: true you get one self-contained row per variant instead — the product fields are copied onto each row (and the nested variants array dropped) so you never have to join back to a product table:

{
  "title": "Men's Dasher NZ - Anthracite",
  "variantId": 42436493115472, "variantTitle": "8.5", "sku": "A12417M085",
  "variantPrice": 140.0, "variantCompareAtPrice": 140.0,
  "variantAvailable": false, "onSale": false, "variant": { … }
}

40 products became 472 variant rows in testing, every field populated. That is the shape you want for price and stock monitoring.

How it works

Every Shopify storefront exposes /products.json (and /collections/<handle>/products.json) as a public, unauthenticated JSON feed. There is no HTML to parse and nothing to reverse-engineer.

Pagination. limit is capped at 250 server-side — asking for 500 returns 250 — and paging is ?page=N until a page comes back empty. Verified: page 1 = 250, page 2 = 44, page 3 = 0. The actor stops on the empty page rather than guessing a page count.

Not every Shopify store is reachable. The feed has no anti-bot layer, but the storefront in front of it can: allbirds.com answered a plain request while gymshark.com returned 403 (Cloudflare) and shop.tesla.com returned 406. The actor rotates a small TLS ladder for those, and most stores never need past the first rung.

Three failure modes, three error codes — because "it didn't work" is not actionable:

_error	Meaning
bad_input	Not a hostname at all (caught before any request is spent)
not_shopify	The site answered, but not with a Shopify feed — wrong platform, or a password-protected storefront
blocked	Shopify store, but the storefront refused us (Cloudflare etc.)

Known limits

Limit	Detail
Published products only	The feed shows what the storefront shows; drafts and hidden products are not in it
No inventory counts	Shopify exposes available (boolean) here, not quantity on hand
Password-protected stores	Report not_shopify — the feed is not served behind the password page