Chrome Web Store Reviews Scraper — Extension Reviews

Scrape every review of any Chrome extension into clean, analysis-ready JSON: star rating, full review text, author, ISO timestamps, helpful / not-helpful votes, the extension version the user was on, the review's language, and the developer's public reply — plus the listing's own metadata (name, average rating, total ratings, weekly users, category, current version, developer).

This is not HTML scraping. The modern Chrome Web Store is an Angular app that serves reviews from Google's internal JSON-RPC transport; this Actor talks to that endpoint directly, the same one the page itself calls. No browser, no login, no API key, no proxies — so it is fast, cheap and immune to visual redesigns.

What you get from a single run:

• Complete review histories, not the first page. Reviews come back 200 per request with unlimited token pagination — verified deep into extensions with tens of thousands of written reviews.
• Developer replies, matched to the review they answer, with their own author and timestamp — the field most review datasets drop.
• All 39 review languages, merged. The Store partitions reviews by language: its English page hides the Spanish, Russian and Portuguese reviews entirely. Set language: "all" and this Actor sweeps every language and deduplicates into one dataset. On a mid-size extension that is often 40% more reviews than the English listing shows.
• Version-tagged feedback. Every review carries the extension version it was written against, so you can see exactly which release caused a rating drop.

Example output item (real review, unedited):

{
    "reviewId": "fae2c3ca-68f6-4467-937d-4c642918096d",
    "extensionId": "gighmmpiobklfepjocnamgkkbiglidom",
    "extensionName": "AdBlock — block ads across the web",
    "extensionUrl": "https://chromewebstore.google.com/detail/gighmmpiobklfepjocnamgkkbiglidom",
    "extensionIcon": "https://lh3.googleusercontent.com/mgNKV-3VMXD556WVUiWSbcukQQN-il4Zlqq03efTjG2B5j9YP7Fxr3idTQ_G0JFD7E6o4TMwvTQTleDn_8UdFLf5VQ",
    "extensionSummary": "Block ads on YouTube and your favorite sites for free",
    "extensionCategory": "productivity/workflow",
    "extensionVersion": "6.45.2",
    "extensionUpdatedAt": "2026-08-18T14:10:09.000Z",
    "extensionRating": 4.473629640346254,
    "extensionRatingCount": 290307,
    "extensionReviewCount": 37963,
    "extensionUserCount": 60000000,
    "extensionDeveloper": "AdBlock",
    "extensionWebsite": "http://getadblock.com/",
    "rating": 1,
    "text": "Adblock (getadblock.com) changed its revenue scheme, which means poorer ad-blocking performance to drive users to a subscription.",
    "author": "Giovanni",
    "authorAvatar": "https://lh3.googleusercontent.com/a-/ALV-UjWK1nx3jX4QdSeoABCET5WvamXrfD-_BiWcwb4KeqPfgHubZ4eJ=s32",
    "publishedAt": "2026-08-04T06:33:42.000Z",
    "updatedAt": "2026-08-04T06:33:42.000Z",
    "edited": false,
    "helpfulCount": 7,
    "notHelpfulCount": 0,
    "totalVotes": 7,
    "version": "6.44.0.1",
    "language": "en",
    "requestedLanguage": "en",
    "developerReply": "Thanks for sharing your thoughts. While we have added new functionality for our Premium users, we continue to provide the same free in-browser ad blocking that you rely on. Our filter list authors are always monitoring for issues and make frequent updates to our free ad-blocking filters. If you have questions, please reach out to us at help@getadblock.com. We'd love to connect with you.",
    "developerReplyAuthor": "AdBlock Support",
    "developerReplyAt": "2026-08-04T18:48:25.000Z",
    "scrapedAt": "2026-08-21T03:13:50.174Z"
}

How to scrape Chrome Web Store reviews

1. Paste one or more extension URLs — https://chromewebstore.google.com/detail/<slug>/<id>, the old chrome.google.com/webstore/detail/... form, or just the bare 32-character id. Up to 100 extensions per run.
2. Set max reviews per extension, pick a language (or all) and a sort order.
3. Press Start. Results land in the dataset and export to JSON, CSV or Excel, or sync to Google Sheets, Zapier, Make, a webhook or the Apify API.

Example input

{
    "extensionUrls": [
        "https://chromewebstore.google.com/detail/adblock-block-ads-across-the-web/gighmmpiobklfepjocnamgkkbiglidom",
        "kbfnbcaeplbcioakkpcpgfkobkghlhen"
    ],
    "maxReviewsPerExtension": 2000,
    "language": "en",
    "sort": "newest"
}

• extensionUrls — Chrome Web Store listing URLs or bare extension ids. Anything containing a 32-character id works; entries without one are skipped with a warning.
• maxReviewsPerExtension — cap per extension (not per run). With language: "all" the cap covers all languages together.
• language — en by default. One of 39 codes, or all to sweep every language and merge. The Store returns only reviews written in the requested language.
• sort — newest (default), most_relevant, lowest_rating, highest_rating.
• proxyConfiguration — off by default; the Store answers plain requests.

Output fields

Field	Type	Description
reviewId	string	Stable review UUID — use it to diff runs and deduplicate
extensionId	string	32-character Chrome extension id
extensionName	string | null	Listing title
extensionUrl	string	Link to the listing
extensionIcon	string | null	Icon URL
extensionSummary	string | null	Short description shown under the title
extensionCategory	string | null	Store category, e.g. productivity/workflow
extensionVersion	string | null	Version currently published
extensionUpdatedAt	string | null	ISO date the listing was last updated
extensionRating	number | null	Average star rating (full precision, e.g. 4.4736)
extensionRatingCount	number | null	Total star ratings, including ratings with no text
extensionReviewCount	number | null	Written reviews available in the requested language
extensionUserCount	number | null	Users reported by the Store
extensionDeveloper	string | null	Publisher name
extensionWebsite	string | null	Developer website
rating	number | null	This review's rating, 1–5
text	string | null	Full review text, newlines preserved
author	string | null	Reviewer's display name
authorAvatar	string | null	Reviewer's avatar URL
publishedAt	string | null	ISO 8601 UTC timestamp the review was posted
updatedAt	string | null	ISO 8601 UTC timestamp of the last edit
edited	boolean	true when the review was edited after posting
helpfulCount	number	Users who marked the review helpful
notHelpfulCount	number	Users who marked it not helpful
totalVotes	number	helpfulCount + notHelpfulCount — the "N out of M" the page shows
version	string | null	Extension version the reviewer was running
language	string | null	Language the Store detected for the review
requestedLanguage	string	Language bucket this review was fetched from
developerReply	string | null	Developer's public reply, when there is one
developerReplyAuthor	string | null	Name the developer replies under
developerReplyAt	string | null	ISO 8601 UTC timestamp of the reply
scrapedAt	string	ISO 8601 UTC timestamp of this run

A run summary (per extension, per language, reported totals and any errors) is stored in the run's key-value store under SUMMARY.

What can you do with Chrome extension review data?

• Competitor teardown. Pull lowest_rating on a rival extension and you have its bug backlog, its churn reasons and its angriest users, ranked, in one table.
• Release-quality tracking. Group by version and watch average rating per release — the fastest way to prove which build broke things.
• Monetization backlash monitoring. Extensions that go freemium light up in the review feed within hours; schedule daily runs and alert on 1-star spikes.
• Support triage. developerReply is null on every review nobody answered — filter for unanswered 1- and 2-star reviews and you have a work queue.
• Market research. Scrape the whole category, aggregate extensionRating, extensionUserCount and review sentiment, and find the underserved niche.
• LLM training and sentiment sets. Multilingual, labelled with a star rating, dated, version-tagged — with language: "all" you get real coverage instead of an English-only sample.

Integrations

Every Apify dataset exports to CSV, Excel, JSON, XML or HTML with one click, or syncs to Google Sheets, and the Actor plugs into Zapier, Make, webhooks, MCP and the Apify API:

# Python
from apify_client import ApifyClient

client = ApifyClient("<YOUR_API_TOKEN>")
run = client.actor("fabrikit/chrome-web-store-reviews-scraper").call(run_input={
    "extensionUrls": ["gighmmpiobklfepjocnamgkkbiglidom"],
    "maxReviewsPerExtension": 1000,
    "language": "all",
    "sort": "lowest_rating",
})
for review in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(review["rating"], review["version"], "—", review["text"][:80])

// Node.js
import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: '<YOUR_API_TOKEN>' });
const run = await client.actor('fabrikit/chrome-web-store-reviews-scraper').call({
    extensionUrls: [
        'https://chromewebstore.google.com/detail/adblock-block-ads-across-the-web/gighmmpiobklfepjocnamgkkbiglidom',
    ],
    maxReviewsPerExtension: 2000,
    sort: 'newest',
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items.length, 'reviews');

Performance & limits

• Plain HTTP JSON, no browser. A verified benchmark: 6,000 reviews from 3 extensions in 37 seconds (AdBlock, Grammarly, Honey — 2,000 each).
• 200 reviews per request, token pagination with no depth limit. Pagination ends cleanly: an extension with 347 English reviews returns exactly 347, zero duplicates.
• Reviews are deduplicated by reviewId within a run, including across languages when language: "all" (the Store aliases pt-BR onto pt and zh-TW onto zh-CN; both are handled).
• A missing or unpublished extension never kills the run — it is logged, recorded in SUMMARY, and the remaining extensions continue.
• Requests are paced at ~3/second. No proxies needed; the endpoint accepts anonymous requests.
• Written reviews are a subset of ratings. AdBlock has 290,307 star ratings but 37,963 written English reviews — extensionRatingCount vs extensionReviewCount tells you which is which before you set a cap.

Tips for best results

• Use language: "all" when the extension is global. English-only runs silently miss the majority on many extensions: Raindrop.io returns 347 English reviews but 489 across all languages.
• sort: "lowest_rating" is the money setting for competitor research and bug mining — it front-loads every complaint instead of making you page through praise.
• For daily monitoring, use sort: "newest" with a small maxReviewsPerExtension (50–200) and diff on reviewId. You only pay for reviews actually delivered.
• Set the cap from extensionReviewCount, not extensionRatingCount — the second number is 5–10× larger and counts silent star ratings you can never fetch.
• Track releases by grouping on version; note it is null on ~7% of reviews (the Store simply did not record it), so treat it as optional in your pipeline.
• Batch up to 100 extensions in one run to amortize the start fee across a whole category.

Pricing — US$0.60 per 1,000 reviews

This Actor uses Apify's pay-per-event model: US$0.0006 per review delivered (event review-result) — US$0.60 per 1,000 reviews — plus a US$0.00005 run-start fee, on top of standard Apify platform usage. No subscription, no rental fee. You only pay for reviews actually pushed to your dataset.

FAQ

Is there an official Chrome Web Store API for reviews? No. Google publishes no public reviews API, and the reviews are not in the page's HTML either — the Store is an Angular app that loads them from an internal JSON-RPC endpoint. This Actor calls that endpoint the same way the page does, and re-reads the endpoint's identifier from the live page on every run, so a Google-side rotation does not silently break it.

Can I get all reviews of an extension, not just the first page? Yes. Pagination is unlimited and verified deep into extensions with tens of thousands of reviews. Set maxReviewsPerExtension high and language: "all" for full coverage.

Why does the review count not match the number of ratings? Most people rate without writing. extensionRatingCount counts every star rating; extensionReviewCount counts written reviews in the requested language. Only written reviews can be fetched.

Does country or region change the results? No — the Store varies reviews by language, not country. A gl (country) parameter has no effect at all, so the Actor does not pretend to offer one.

Are the dates real timestamps? Yes: publishedAt, updatedAt and developerReplyAt are true ISO 8601 UTC timestamps from the Store's own data, not parsed "3 months ago" strings.

Do I need proxies or a Google account? Neither. The endpoint answers anonymous requests with no cookies. proxyConfiguration is available if a very large run ever hits rate limits.

Does it collect personal data? It collects only what the Chrome Web Store publishes on a public product page: the display name and avatar the reviewer chose to show alongside a public review. No emails, no accounts, no private data. How you process it is your own compliance call.

Which extensions work? Any public listing — extensions, themes and apps. Unpublished or removed items are skipped with a warning rather than failing the run.

Support

Missing a field the Store publishes, or need a filter that isn't there? Open an issue in the Issues tab — issues are typically answered within 24 hours, and new fields are cheap to add because the Actor reads structured JSON, not HTML.

More scrapers by Fabrikit

• Google Play Reviews Scraper — Android app reviews with rating, version, helpful votes and developer replies, any country and language.
• App Reviews Scraper — App Store & Google Play — both mobile stores in one normalized dataset.
• ATS Jobs Scraper — Greenhouse, Lever, Ashby & SmartRecruiters — job postings from the four biggest ATS platforms.

Technical details

• Data source: the Chrome Web Store's own batchexecute JSON-RPC endpoint (/_/ChromeWebStoreConsumerFeUi/data/batchexecute), plus one HTML request per extension to read the listing metadata and re-discover the RPC identifiers the page declares for itself.
• No browser, no CSS selectors, no HTML parsing of reviews — a visual redesign of the Store changes nothing.
• Timestamps arrive as epoch seconds and are converted to ISO 8601 UTC.
• helpfulCount / notHelpfulCount reproduce the page's "N out of M found this helpful" exactly.
• The dataset schema in .actor/actor.json is generated from the scraper's own item builder and verified in both directions before every push, so no item is ever rejected for a field the code does not produce.