How to scrape Amazon in 2026 (complete guide)

Amazon is the largest e-commerce storefront in the world, and its product pages are a reference for price monitoring, availability tracking and review analysis. The trade-off: Amazon actively detects automated access and serves verification pages the moment behavior steps out of the ordinary. Here's how to collect cleanly.

What data you can collect on Amazon

• Prices: current price, strikethrough price, changes, third-party offers.
• Availability: in stock, lead times, "buy box" seller.
• Reviews: average rating, number of reviews, public verbatims.
• Listing: title, features, images, category and ranking.

What blocks a naive scraper

Amazon spots automated access and switches to a control page (a message inviting you "to discuss automated access", or a CAPTCHA). As elsewhere, detection stacks several signals: IP reputation, TLS and browser fingerprint, header consistency and behavior. Amazon also adapts its content by country and by account, which makes geolocation important.

The detection mechanism is detailed in our article on anti-bot systems.

How to get the data cleanly

curl -G "https://api.wyndpath.com/v1/" \
  --data-urlencode "api_key=YOUR_KEY" \
  --data-urlencode "url=https://www.amazon.fr/s?k=casque+bluetooth" \
  --data-urlencode "render_js=1" \
  --data-urlencode "country=fr"

WyndPath picks the right route, handles IP rotation and retries, and returns the page. Exiting through country=fr gives you the prices and assortment of the French market. Pay-per-success: a verification page you don't clear costs nothing.

The legal framework

Listings and prices are public, but Amazon's terms of use govern their reuse, and reviews contain personal data (pseudonyms) subject to the GDPR. Minimize what you collect, respect robots.txt, and keep volumes reasonable. Details in web scraping and the GDPR in France.