Amazon Reviews Scraper

Amazon India public reviews

Reads actual customer review cards from Amazon.in product pages, then tries the dedicated reviews route when the per-product target remains unmet. Dedicated reviews can require sign-in; this Actor does not provide login access or bypass that requirement. It never pads reviews with product descriptions or search snippets.

Input

productUrls is required and accepts HTTPS Amazon.in product/review URLs (strings or url objects). ASINs are validated, product URLs normalized and duplicate products removed. Other marketplaces must use their own supported adapter. maxReviewsPerProduct is 1–50 (default 10); optional maxItems caps all products combined. If maxItems is absent, the total cap is per-product maximum times the number of supplied products.

proxyConfiguration accepts Apify or custom proxies; the sample uses datacenter. One browser page at a time and bounded navigation/retries reduce duplicate work. Anonymous product pages often expose fewer than ten reviews per product. The ten-record sample requests two real products; it does not claim ten available public reviews for every individual product.

Output

Records retain source-native reviewId, available title/body/rating/reviewer/date/images/helpful votes and product context. reviewUrl points to the original review. sourceUrl records the actually observed page and scrapedAt is retrieval time, not publication time. Only an observed verified-purchase badge emits verifiedPurchase=true; absence is omitted. Unknown names, counts, rating and flags are omitted rather than invented. Zero counts are preserved only if displayed by the source.

Reviews require a native ID and actual review text/title. IDs deduplicate across pages. Product rating counts (totalRatings) remain separate from explicit review counts (totalReviews). Source card HTML is saved in run key-value storage for inspection. RUN_SUMMARY gives per-product counts and access errors. Partial datasets are reported as partial by the audit; zero-review runs fail.

Run and export

Use apify run --input-file input.json locally. Inspect the default Dataset and RUN_SUMMARY before requesting larger limits. Export JSON/CSV from Apify after a production run. Data availability depends on the source; compute and configured proxy usage are billed by the platform.

The source can pool reviews across product variants. productAsin is the requested product, while reviewedAsin preserves the native card ASIN when exposed. Review URLs use actual source links when available. Both current reviewTitle/reviewRichContentContainer markup and older review-title/review-body markup are supported.