Walmart Reviews Scraper

Collect every written review for any Walmart product — text, rating, author, date and helpful votes — plus an optional product rating summary with the average score and a full star breakdown. Paste product links or bare item IDs and get back a clean row for each review, ready for sentiment analysis, competitor monitoring or voice-of-customer research. Reviews are collected in the order you choose: most relevant, newest, highest rated or lowest rated. Built for anyone who needs Walmart's customer feedback as data: brands tracking what shoppers say, retailers watching rivals, analysts mining real shopper language and teams feeding review text into topic or sentiment models.

What you can do with it

• Mine customer language — pull the review text, titles and star rating for any product in one dataset, ready for analysis.
• Isolate complaints or praise — sort by highest rated or lowest rated to surface what delights or frustrates buyers first.
• Monitor a product over time — sort by newest and schedule the actor to catch fresh feedback as it appears.
• Read Walmart the way shoppers do — every row carries the review date, helpful votes, whether it was a verified purchase and any photos the reviewer attached.
• Get the full rating context — add one summary row per product with the average rating, total reviews and the count at each star level (5 to 1).
• Reuse a link you already have — paste full Walmart product addresses or just the numeric item ID; mixed together in the same list is fine.

What you get

One row per review. Abridged sample of a single review:

{
  "rowType": "review",
  "productId": "12345678",
  "productUrl": "https://www.walmart.com/ip/12345678",
  "productTitle": "Sony WH-1000XM5 Wireless Headphones",
  "reviewId": "r-9f3a2c1b-8a7d-4e1f-9a2b-1c3d4e5f6a7b",
  "author": "James M.",
  "authorName": "James M.",
  "rating": 5,
  "title": "Best headphones I have owned",
  "text": "Incredible noise cancellation and battery life. Comfortable for long sessions and the sound is crystal clear.",
  "date": "2024-03-15T00:00:00.000Z",
  "displayDate": "March 15, 2024",
  "helpfulVotes": 12,
  "verifiedPurchase": true,
  "images": [
    "https://i5.walmartimages.com/seo/review-image-abc123.jpeg"
  ],
  "requestedInput": "https://www.walmart.com/ip/Sony-WH-1000XM5-Wireless-Headphones/12345678",
  "scrapedAt": "2026-08-28T12:00:00.000Z"
}

When the product summary is switched on, one extra row per product is added:

{
  "rowType": "summary",
  "productId": "12345678",
  "productUrl": "https://www.walmart.com/ip/12345678",
  "productTitle": "Sony WH-1000XM5 Wireless Headphones",
  "avgRating": 4.6,
  "totalReviews": 3421,
  "totalRatings": 3421,
  "histogram": { "5": 2310, "4": 780, "3": 180, "2": 80, "1": 71 },
  "reviewsCollected": 100,
  "requestedInput": "https://www.walmart.com/ip/Sony-WH-1000XM5-Wireless-Headphones/12345678",
  "scrapedAt": "2026-08-28T12:00:00.000Z"
}

date is an ISO timestamp when the original wording can be parsed (for example "March 15, 2024" or an ISO string the site already publishes). displayDate always preserves what Walmart showed. Relative dates such as "2 weeks ago" are resolved to the time of collection. When a date cannot be parsed, date is empty and displayDate still carries the original text.

Input reference

Setting	Type	Default	What it does
Walmart products	list of text	—	Products whose reviews you want. Paste full Walmart product links (walmart.com/ip/…) or bare item IDs (for example 12345678) — mixed together is fine. Duplicates that point to the same item are collected once. Required.
Review order	choice	Most relevant	The order Walmart serves the reviews in. Most relevant is Walmart's default ranking; newest puts fresh feedback first; highest and lowest rated surface praise or complaints.
Maximum reviews per product	whole number	50	Stop after this many written reviews per product. Products with fewer reviews simply return what exists. Set 0 to skip review text entirely and collect only the product rating summary. Maximum 2,000.
Include the product rating summary	true/false	true	Add one summary row per product: average rating, total reviews and the count at each star level (5 to 1). The summary is collected once per product.
Parallel products	whole number	5	How many products to work on at the same time. Higher is faster. Leave at 5 unless you have a reason to change it. Maximum 10.

Output fields

Review rows (rowType: "review")

Field	Type	Description
rowType	text	Always review for review rows.
productId	text	Walmart's item number for the product the review belongs to.
productUrl	text	Clean link to the product page.
productTitle	text	Product title as published, when available on the page.
reviewId	text	The review's stable identifier.
author / authorName	text	Who wrote it, as Walmart publishes it.
rating	number	Stars given, 1 to 5.
title	text	Review headline / title.
text	text	Full review body.
date	text	ISO timestamp parsed from Walmart's date; empty when the wording cannot be parsed.
displayDate	text	The date exactly as Walmart showed it.
helpfulVotes	number	How many shoppers found the review helpful.
verifiedPurchase	true/false	Whether Walmart marks it as a verified purchase.
images	list of text	Photo URLs the reviewer attached, when any.
requestedInput	text	Exactly what you supplied for the product, so results are easy to join back.
scrapedAt	text	When the row was collected (ISO timestamp).

Summary rows (rowType: "summary", one per product when switched on)

Field	Type	Description
rowType	text	Always summary.
productId / productUrl / productTitle	text	The product the summary belongs to.
avgRating	number	Average customer rating (0–5).
totalReviews	number	Number of written reviews the product reports.
totalRatings	number	Number of ratings; on Walmart this matches written reviews.
histogram	object	Count at each star level: { "5": 2310, "4": 780, "3": 180, "2": 80, "1": 71 }. Missing levels are empty rather than guessed.
reviewsCollected	number	How many review rows this run collected for the product.
requestedInput	text	What you supplied for the product.
scrapedAt	text	When the row was collected.

Every row is priced in USD where a price appears — Walmart is a US-only marketplace. Ratings and reviews are a snapshot at the moment of collection.

Pricing

You pay per result, with no monthly minimum.

What you are charged for	Price
Review — one review row added to your dataset	$0.001 per review — $1.00 per 1,000
Product summary — one summary row added to your dataset (when switched on)	$0.002 per summary — $2.00 per 1,000

You are only charged for rows that are actually added to your dataset. A product that turns out to have no written reviews returns no review rows and therefore costs nothing for reviews; a summary is still added when that option is on and reports the totals Walmart publishes. Products that cannot be read at all cost nothing and are listed in the run summary.

Worked example. A run over 10 products collecting 100 reviews each is 1,000 review rows: 1,000 × $0.001 = $1.00. With the product summary switched on it adds 10 summary rows: 10 × $0.002 = $0.02, for a total of about $1.02. A daily run collecting 50 reviews across 5 products is about $0.05 per run, roughly $1.50 a month.

Limits & what this actor cannot do

• Prices, ratings, stock and review ordering are a snapshot at the moment of collection and keep changing afterwards; Walmart's ranking and the set of reviews it surfaces are not identical from one moment to the next.
• Only written reviews Walmart publishes openly are returned. Shoppers who left a star rating without writing anything are counted in the summary totals but do not produce a review row.
• The summary's totals and the number of review rows can differ. Both are returned as published or collected; neither substitutes for the other.
• There is a practical ceiling per product imposed by what Walmart will actually serve; the actor stops when the reviews run out rather than inventing more. To cover many products, run a longer list rather than expecting thousands from a single listing.
• Images are links to what the reviewer uploaded. Not every review carries photos; where there are none the list comes back empty rather than guessed.
• Speed depends on the size of the job and on Walmart's own response times; no fixed throughput is promised.
• Walmart's terms govern automated access. You are responsible for using the data lawfully and in line with the source site's terms, and for handling any personal data in line with applicable privacy law.

FAQ

Do I need a Walmart account? No. The actor reads only what Walmart shows the public on its product and reviews pages.

Does it need my login or password? No. There is nothing to connect and nothing to configure — enter a product link or ID and press start.

Can I paste a product link I already built on Walmart? Yes. Paste any link that contains /ip/ and the actor extracts the item ID. A bare numeric ID works too.

How many reviews can I get per product? Up to 2,000 per product in one run. Walmart stops serving sooner when the product has fewer, so the actor stops when the reviews run out.

How does sorting work? Most relevant is Walmart's default relevance ranking. Newest surfaces fresh feedback first. Highest rated and lowest rated surface praise or complaints. The option sets the sort that is sent to each reviews page.

What if a product has no reviews? It is reported as an empty product (not a failure) in the run summary and contributes no review charge. When the summary is switched on, the summary row is still produced and reports the product's zero.

Can I get only the summary without reading every review? Yes. Set the maximum reviews per product to 0 and leave the summary switched on. The actor fetches one product page per product and skips all review pages.

Can I schedule it? Yes. Schedule it hourly, daily or weekly from the Apify console, and each run writes a fresh dataset you can compare against the last — ideal for tracking new feedback over time.

Is the data complete? It is complete for what Walmart publishes openly on its reviews pages. Where a page omits a field the actor leaves it empty rather than guessing, and it reports exactly the review rows Walmart actually serves rather than an estimated total.