Estée Lauder Australia Product, Price & Variant Scraper

Extract live Estée Lauder Australia catalog data for product discovery, price monitoring, beauty-market research, catalog enrichment, and AI/agent workflows. The Actor supports keyword search, category discovery, category browsing, and deterministic single-product lookup. Product results can include prices, discount prices, stock state, ratings, review counts, images, SKUs, barcodes, descriptions, sizes, shades, and other variant details when the source exposes them.

Quick start

The default run is buyer-useful and searches for Advanced Night Repair:

{
  "operation": "search",
  "query": "Advanced Night Repair",
  "page": 1,
  "limit": 5
}

Search normally tries the retailer's Endeca search endpoint with a short bounded deadline. If that endpoint stalls, the Actor falls back to the public sitemap for discovery and resolves matching product IDs through the existing exact product GraphQL path. Fallback results are semantically checked against the meaningful query tokens; impossible queries return an empty Dataset instead of unrelated products.

Operations

Operation	Purpose	Main inputs
search	Find products by keyword	query, page, limit
categories	List live category IDs/slugs	none
category	Browse one category	categoryId or slug, page, limit, details
item	Fetch one exact product	productId or url

A reliable automation pattern is search → exact item: take the returned product id and send it back as productId. IDs emitted in a PROD12345 form are normalized automatically before exact GraphQL lookup.

Output contract

The default Dataset is the canonical machine-readable output. Product rows keep the existing business fields and add a small self-describing envelope: record_type=product, operation, retailer=Estee Lauder, country=AU, currency=AUD, retrieved_at, and source. Category-discovery rows use record_type=category. The Actor also writes an OUTPUT summary with status, operation, count, no_match, and as_of, which is convenient for MCP clients, AI agents, schedulers, and downstream workflows that need to distinguish a valid no-match from a failed run.

Typical product fields include:

• id, name, brand, slug, category, breadcrumbs
• current_price, price, discount_price, currency
• availability, in_stock
• rating, review_count
• sku, barcode, images
• description, short_description, variants
• source_url, source, retrieved_at

price is the regular/list price when a separate lower discounted price is exposed; discount_price is that lower sale price. current_price is the effective current value (discount_price when present, otherwise price). Missing retailer data is left absent/null rather than guessed.

Category example

First run categories, then reuse a returned ID:

{
  "operation": "category",
  "categoryId": "26389",
  "page": 1,
  "limit": 5,
  "details": true
}

Exact item example

{
  "operation": "item",
  "productId": "101853"
}

Exact item results are detailed and may include all available variants. A product URL can be supplied instead of the ID.

Reliability and limitations

Retailer search and catalog backends can change without notice. The Actor deliberately bounds the historically unstable Endeca search path and does not convert transport failures or impossible queries into fake matches. Category and item workflows use the retailer's current catalog surfaces and can expose richer data than the fast search route. Large rich rows, especially products with many shades or images, can be substantial.

The optional top-level customProxyUrl is treated as a secret input. Normal runs do not require buyers to provide credentials. This release does not change Actor pricing; review the current Store pricing shown on the Actor page before running large jobs.

This Actor is unofficial and is not affiliated with or endorsed by Estée Lauder. Users are responsible for complying with applicable website terms, policies, and laws.