This tutorial shows how to scrape Amazon reviews into CSV with the Amazon Reviews Scraper Lite for US template. You will import the workflow, replace the sample URLs, confirm the export path, run one product first, and validate the rows.
Before setup
Prerequisites and permission checks
You need UScraper installed as a local desktop app, the free template JSON, one or more Amazon.com product detail URLs, and a folder where the CSV can be saved. Start with one ASIN so blocked-page diagnostics do not get mixed into a larger export.
Review Amazon's current robots.txt, customer reviews help, and your internal data policy before automation. This guide is for controlled research from pages you can inspect in a normal browser, not for bypassing CAPTCHA, sign-in walls, account pages, or blocked widgets.
Technical access is not permission. Collect only the fields you need, keep request volume modest, and get legal review before commercial reuse or redistribution.
Workflow anatomy
What the Amazon reviews scraper does
The JSON export is the authoritative workflow definition. The template navigates to each product URL, waits, runs an injected JavaScript collector, waits for hidden export rows, then writes a structured CSV.
The collector uses three best-effort paths: visible review snippets, direct /product-reviews pages, and Amazon's same-origin reviews-render response when exposed. It deduplicates likely repeats, adds product context, and stores results in hidden #uscraper-review-rows elements for Structured Export.
Use the maintained Amazon Reviews Scraper Lite for US template as the download path instead of rebuilding the graph by hand.
| Stage | What happens | Why it matters |
|---|---|---|
| Navigate | Opens the configured Amazon.com product URLs | Keeps the run URL-driven and auditable |
| Inject JavaScript | Detects ASIN, product context, review rows, review pages, and fallback state | Centralizes fragile review logic in one block |
| Wait for Element | Waits for #uscraper-review-rows .uscraper-review-row | Prevents export before collection finishes |
| Structured Export | Appends mapped fields into the local CSV | Produces spreadsheet-ready review rows |
| Loop Continue | Moves to the next product URL | Supports controlled multi-product runs |
Runbook
How to scrape Amazon reviews to CSV
Import the template
Open Amazon Reviews Scraper Lite for US, download the JSON, and import it into UScraper.
Replace the sample URLs
In the Navigate block, replace the bundled Amazon.com examples with product detail URLs you are allowed to process. The workflow detects ASINs from /dp/, /product-reviews/, pd_rd_i, or asin values.
Confirm waits and timeout
Keep the page-load wait, short sleep, JavaScript completion wait, second pause, and review-row wait in place for the first run. Shortening waits can create false blanks.
Set the export folder
In Structured Export, confirm amazon-reviews-scraper-lite-for-us.csv, headers enabled, append mode, and a project-specific local save folder.
Run one ASIN
Compare the CSV against the browser for ASIN, product title, reviewer, rating, review title, body, date, location, helpful count, and image links.
Add more inputs
Only after the first rows pass QA, add more URLs. Stop the run if Amazon shows verification, unavailable content, or repeated fallback rows.
Because the file mode is append, rerunning the same product can duplicate rows. For tests, rename the previous CSV or clear the output file.
Output
What the Amazon reviews to CSV export includes
The bundle does not include a finished CSV sample. Use this export shape as the expected schema, then treat your first validation run as the sample for that product and browser session.
amazon-reviews-scraper-lite-for-us.csvColumn
Product_price
Visible price.
Column
Product_title
Product name.
Column
Product_link
Product URL.
Column
ASIN
Product ID.
Column
Reviewer_name
Reviewer.
Column
Review_title
Headline.
Column
Review_content
Body text.
Column
Product_attributes
Variant text.
Column
Review__date
Review date.
Column
Rating_count
Rating count.
Column
Review_Count
Review count.
Column
Data_source
Static source value, Amazon.
Column
Product_stars
Average stars.
Column
Review_location
Review locale.
Column
Review_rating
Review stars.
Column
Helpful_count
Helpful votes.
Column
Reviewer_avatar_link
Avatar URL.
Column
Review_image_link
Review images.
One review card should create one row. If ASIN, Reviewer_name, Review_rating, and Review_content are all blank, stop and inspect the browser.
Validation
Validate the export before analysis
Open the CSV beside the browser and audit rows from the first product, a later review page, and the final URL. Sort by ASIN, Reviewer_name, Review__date, and Review_title, then compare product context, rating, body, helpful count, and image links against the page.
| Symptom | Likely cause | Fix |
|---|---|---|
| Diagnostic fallback row | Amazon did not expose review rows, showed CAPTCHA, redirected, or served unavailable content | Stop the run, inspect the browser, and do not bypass access controls |
| Product fields export but review text is blank | Review body markup changed or collapsed content was not visible | Confirm text is visible and update the body parsing selectors only after a small test |
| Rows duplicate after testing | The export block uses append mode | Rename or clear the CSV before rerunning the same ASIN |
| Review count differs from Amazon's page | Amazon may filter, localize, sort, or hide reviews by session | Treat the CSV as a snapshot of the visible session, not a canonical review database |
The collector finished but found no usable review cards. Check for CAPTCHA, robot checks, unavailable reviews, redirects, sign-in prompts, or pages where Amazon did not render the review module.
Alternatives
UScraper vs Python and review APIs
A no-code Amazon reviews scraper is useful when the goal is a supervised CSV for research, QA, or voice-of-customer analysis. If your team searches for scrape Amazon reviews Python, the trade-off is control versus maintenance: Python gives code ownership, but you own rendering, pagination, retries, selector drift, compliance checks, and exports. An amazon product reviews API or data provider is usually a better fit for contracted production ingestion.
For wider scraping workflows, browse the UScraper template library. For more tutorials and comparisons, start from the UScraper blog.
FAQ
FAQs
Is it legal to scrape Amazon reviews?
Amazon reviews may be publicly visible, but automation can still be limited by terms, robots directives, marketplace policies, privacy law, and local regulations. Use modest runs and do not bypass access controls.
What does the Amazon Reviews Scraper Lite for US export?
It exports product context, ASIN, reviewer name, review title, review content, product attributes, review date, counts, stars, helpful votes, avatar links, and review image links to CSV.
Do I need an Amazon product reviews API?
No API key is built in. The workflow opens editable Amazon.com product URLs, detects the ASIN, checks visible review snippets and review pages, then exports exposed fields.
Can I scrape Amazon reviews with Python instead?
Yes. Python with Playwright, requests, BeautifulSoup, or a scraping API can work, but your team owns rendering, pagination, retries, compliance review, and formatting.
Why did the Amazon reviews scraper return a fallback row?
A fallback row means Amazon did not expose review rows in that browser session. CAPTCHA, robot checks, unavailable review pages, blocked widgets, or layout changes may be present.