Shein Featured Products Scraper

Scrape featured products from Shein regional homepages across 7 countries. Returns product ID, name, URL, prices (sale + retail), image, discount percentage, and category.

What it does

• Fetches the Shein homepage for the selected country
• Parses ~50 server-rendered featured products from the embedded productList JSON
• Per-product bracket-counted JSON extraction tolerates non-JSON fragments in the surrounding page
• Returns prices in local currency (USD, GBP, EUR, AUD) with sale/retail/discount
• Hardcoded residential proxy (Shein hard-blocks datacenter IPs)

Scope note

This scraper returns currently-featured Shein homepage products, not a keyword/URL-based catalog lookup. Shein's category (/dress-c-1727.html) and product detail (/...-p-39521541.html) pages redirect every anonymous session to /risk/challenge?captcha_type=903. Only the root homepage (https://us.shein.com/) is reachable without cookies, so the scraper extracts featured products from the homepage server-side render.

The featured set rotates per session ID, so each run can return different products. Run the scraper multiple times to accumulate variety.

Input

Field	Type	Default	Description
country	enum	US	Shein regional homepage: US, UK, DE, FR, ES, IT, AU
maxItems	integer	30	Maximum featured products to return (1-200)

Example input

{
    "country": "DE",
    "maxItems": 20
}

Output

Each product is a JSON record with the following fields (missing fields are omitted, never null):

Field	Type	Always present	Description
productId	string	✓	Shein goods_id
title	string	✓	Product name
url	string	✓	Product page URL
countryCode	string	✓	2-letter country code
scrapedAt	string	✓	UTC ISO 8601 scrape timestamp
mainImage	string	—	Main product image URL
currency	string	—	Local currency (USD, GBP, EUR, AUD)
salePrice	number	—	Current sale price
salePriceText	string	—	Formatted price with currency symbol
retailPrice	number	—	Original retail price (before discount)
discountPercentage	number	—	Discount % off retail
catId	string	—	Shein category ID
brand	string	—	Brand name if present
images	array	—	Additional product images

Sample output

{
    "productId": "323288283",
    "title": "Amplova Women's Slogan Print Yellow Bandeau Top And Low Waist Mini Shorts Set",
    "url": "https://us.shein.com/-p-323288283.html",
    "countryCode": "US",
    "mainImage": "https://img.ltwebstatic.com/v4/j/pi/2025/12/11/af/1765421201aae6799000f50a389fe2f87ee624a82d_thumbnail_405x552.jpg",
    "currency": "USD",
    "salePrice": 18.23,
    "salePriceText": "$18.23",
    "retailPrice": 26.39,
    "discountPercentage": 31,
    "catId": "1780",
    "scrapedAt": "2026-04-14T08:42:22Z"
}

FAQ

Do I need a proxy? No configuration needed — a residential proxy is hardcoded and applied automatically. Shein blocks direct datacenter access.

Why featured products only, not keyword search? Shein blocks category and product detail pages with a /risk/challenge?captcha_type=903 redirect. Only the root homepage is reachable without pre-seeded cookies or a captcha-solver, so this scraper extracts the featured products embedded in the homepage server-side render.

How many products per run? The homepage ships 6-60 products depending on session / IP. The scraper returns up to maxItems (default 30).

Do different countries return different products? Yes. Each regional site (us.shein.com, uk.shein.com, etc.) has its own feature set with local pricing and currency.

Is the data fresh? Every run fetches the live homepage and the featured set rotates per session, so repeated runs build up a rolling catalog.

Use cases

• Daily trending monitor — track what Shein is featuring on each regional homepage
• Price sampling — accumulate a rolling catalog of Shein prices across countries
• Market research — monitor product themes / categories across 7 markets
• Cross-market comparison — compare US vs EU featured picks for the same brand

Notes

• Pricing is configured in the Apify UI (pay-per-result).
• The daily Apify test run uses country=US, maxItems=5.
• Parser: _iter_product_blocks walks every {"goods_id":"..."} object with a bracket counter + json.loads, tolerant of nested JSON fragments in the enclosing array.