Skip to content

Site guides

Getting Amazon product data: the official API first, and the limits of public pages

Use the Product Advertising API where it fits, understand Amazon's terms, and assess the legal and technical limits of public product pages. No login or paywall bypass.

Last updated 3 min readBy CrawlerBros Editorial

Amazon product data is available through the Product Advertising API (PA-API) for eligible associates, and through public product pages for limited, low-volume use. The API is the stable path; public pages are a terms-of-service and anti-bot minefield. This guide starts with the API and states the limits of the alternative.

This page is research material, not legal advice. Amazon’s terms and your jurisdiction govern your use.

Start with the Product Advertising API

PA-API 5.0 provides product data to Amazon Associates who meet the programme’s eligibility requirements. It is the documented way to get:

  • product titles, ASINs, and URLs;
  • prices and availability (subject to the programme’s rules);
  • images and feature bullets;
  • offers and variations, where the API supports them.

Access requires an Amazon Associates account, qualifying sales, and API credentials. The API uses AWS Signature Version 4 signing. Read the current documentation for the exact resources, parameters, and rate limits; they change.

A request is a signed HTTPS call, not a browser scrape. The response is JSON. Keep your access key and secret key in a secret manager, never in a repository.

What PA-API does not give you

  • Arbitrary search results at unlimited volume. The API has rate limits and eligible resources.
  • All customer reviews or personal data. Review access is restricted.
  • A licence to republish content outside the programme’s terms. The Associates programme has its own display and attribution rules.
  • Access without eligibility. If you do not qualify, the API is not an option.

Public product pages

Amazon’s Conditions of Use prohibit certain automated access, and its robots.txt disallows many paths. Public product pages also use anti-bot measures. A crawler that ignores those signals is taking on contractual and technical risk.

If you have a permitted, low-volume use case, the defensible approach is:

  1. Read the current Conditions of Use and robots.txt.
  2. Check whether the data is available through PA-API or another official feed.
  3. Use a descriptive user agent and a very low rate.
  4. Do not bypass a login, CAPTCHA, or block.
  5. Do not collect personal data (reviewer names, addresses, profiles).
  6. Do not republish copyrighted descriptions or images without a licence.
  7. Stop on a block or a cease-and-desist.

This site does not provide a step-by-step Amazon page scraper. The official API is the path that respects the terms; a page-scraping tutorial would be a recipe for a dispute.

Data fields and stability

If you use PA-API, design around stable identifiers:

  • ASIN is the product key. Marketplace is separate.
  • Marketplace changes currency, availability, and legal terms.
  • Prices can change by the minute and may include marketplace-specific tax/shipping rules.
  • Variations (size, colour) are separate child ASINs; do not collapse them.
  • Images have usage rules; store the URL and attribution, not a rehosted copy unless the terms allow it.

For price monitoring, store the observation time and the marketplace. A price without a timestamp and a marketplace is not a fact you can compare.

A safe architecture

PA-API request (signed)
  -> validate JSON schema
  -> store raw response (short retention)
  -> normalise ASIN, marketplace, price, currency, availability
  -> deduplicate by (ASIN, marketplace)
  -> change detection on price/availability

Do not put the API credentials in the crawler’s frontend. Do not log the signature or the secret key. Rotate credentials if they leak.

  • Contract: the Associates programme and PA-API terms govern the data and its display.
  • Copyright: product descriptions and images are protected; a licence or the programme’s rules are needed to republish.
  • Database rights: if you build a product database in the EU/UK, the database right may apply.
  • Personal data: reviews and seller profiles can contain personal data; minimise and do not build profiles you do not need.
  • Competition/consumer law: price comparison claims must be accurate and dated.

The terms of service guide and the GDPR guide cover the broader analysis.

Sources

Sources checked