Amazon Best Sellers Scraper

This Actor extracts public Amazon Best Sellers ranking cards from category pages and can optionally fetch the selected product detail pages. It supports Amazon India, US, UK, Germany, Japan, Canada, Australia, and Brazil.

Code structure

• src/main.js owns the thin process entrypoint.
• src/actor.js owns Apify lifecycle, abort handling, and run failure reporting.
• src/runner.js owns bounded public Amazon requests, parsing, diagnostics, and dataset output.

Source-specific output

Listing records can include rank, productTitle, ASIN, canonical product URL, category context, page number, price/currency, rating, review count, Prime signal, and public image. With deepSearch:true, records can additionally include brand, productDescription, availability, price/list price, breadcrumbs, feature bullets, attributes, product overview rows, media, and up to ten JSON-LD reviews when extractReviews is enabled.

Every row includes recordId, actor/source URLs, targetKind, direct-transport marker, access/fetch status, found, dataAvailable, fieldCoverage, and dataQuality. Empty arrays, empty objects, blank strings, null values, raw page text, and diagnostic records are omitted.

Access boundary

Requests use serialized, bounded public HTTP. Apify Residential proxying is enabled by default and can be explicitly disabled or replaced through proxyConfiguration. Each request uses bounded pacing, timeouts, and retries; a retry obtains a fresh proxy connection.

The Actor does not solve CAPTCHAs, log in, access paywalls, or use alternate data sources. Amazon challenges, HTTP access controls, redirects outside the requested marketplace, timeouts, and pages without public cards are recorded as bounded failure evidence in OUTPUT. They are never mixed into the product dataset. A run with no verified products fails unless valid source records were all excluded by user filters.

Input

• mode: category, categoryUrl, or productUrl.
• category / categoryUrl: public Best Sellers listing target.
• productUrls: direct Amazon product pages for productUrl mode.
• country, locale, maxItems, maxPages, deepSearch, extractReviews, includeMedia, includeTechnicalDetails, filters, deduplicate, navigationTimeoutMs, minDelayMs, maxDelayMs, maxRetries, and proxyConfiguration.

Mode-specific values are validated before network work: categoryUrl must point to the selected marketplace's Best Sellers path, and productUrl requires supported Amazon product URLs with valid ASINs. extractReviews requires deepSearch=true.

Local development

apify validate-schema
apify run --purge --input-file qa-inputs/local-category.json

Local output is written beneath the configured Apify storage directory. Inspect the product dataset and the default key-value store's OUTPUT record after a run.

Use only public pages and follow Amazon's terms, robots guidance, rate limits, and applicable law.