⭐ G2 Reviews Scraper - Seller URLs, Star, Date and Keyword Filters, No Login

Export published G2 software reviews for any number of products or sellers in one run: star rating, title, what the reviewer likes, dislikes and which problems the product solves, publish date, reviewer name, role, company size, industry and country, G2's sub-ratings (ease of use, setup, admin, support, meets requirements), love and hate themes and the product the reviewer switched from. Star, date, company-size and keyword filters work on product links and seller links alike. Export to JSON/CSV/Excel, run on a schedule, call via API, or connect to Make, Zapier or n8n.

Price: $0.003 per review on the Free plan, $0.0028 on Bronze, $0.0026 on Silver and $0.0025 from Gold up ($3.00 to $2.50 per 1,000 reviews), plus $0.00005 per run start. A product without matching reviews, a link that leads nowhere and a filter that matches nothing cost nothing beyond the start. Until November 10, 2026 the Actor still carries the earlier price of $0.007 per review; the prices above apply from that day.

🧭 What users of other G2 review scrapers run into, and what this Actor does

Seen on the most used G2 review Actors	This Actor
One URL per run; the keyword search works on product URLs only	startUrls takes a list of product links, seller links and plain slugs. Keyword, star, date and company-size filters apply to every one of them, seller links included.
A block or a renamed product ends the run with an error	A request that fails is tried again three times, then over an Apify datacenter and then a residential connection. A link that cannot be read is reported in the status message and skipped; the run ends as succeeded with the rows it has.
"Session pool empty — all requests returning 403"	No pages, no browser, no cookies: the Actor reads the search service G2's own site uses. 2,500 reviews took 26 requests and 9 seconds in a test, without a refusal.
Star filter plus keyword can return nothing without saying why	The status message says per product how many reviews G2 has for exactly these filters ("Slack: 1 reviews saved of 1 matching"), so an empty result is explained.
Only "published after"; at least 10 reviews per run	publishedAfter and publishedBefore (a date or a span like 30 days), any limit from 1.
4 GB of memory for a review list	256 MB.

🔎 What is the G2 Reviews Scraper?

G2 publishes reviews of business software, and its site search runs on a search service that returns those reviews 100 at a time. This Actor asks that service for exactly the reviews you filtered for, newest first or in another order, and saves one row per review. Only reviews G2 has published and approved are read; reviews still in moderation or rejected are filtered out in the query itself. There is no page limit: all 40,121 published Slack reviews can be read.

What data does it extract?

• Review: id, link, star rating (1-5), likelihood to recommend (0-10), title, full text, likes, dislikes, problems solved, publish, submit and update time (UTC), helpful votes, review source (organic, invited and so on), video flag
• Reviewer: name as G2 shows it ("Verified User in Computer Software" when the reviewer stays anonymous), role (User, Administrator, Consultant...), company size, industry, country, region
• Ratings: ease of use, ease of setup, ease of admin, quality of support, meets requirements, ease of doing business with (G2's 1-7 scale; empty when the reviewer skipped one)
• Themes and switching: G2's love theme and hate theme, whether the reviewer switched from another product and why
• Product: name, id, slug, seller name, review page link

⬇️ Input

Field	What it does
startUrls	G2 product links (https://www.g2.com/products/slack/reviews), seller links (https://www.g2.com/sellers/salesforce) or plain slugs (slack)
maxReviewsPerProduct	Reviews saved per product (default 100)
maxProductsPerSeller	For seller links: how many of the seller's products are read, most reviewed first (default 10)
starRatings	["1", "2"] for the critical reviews only; empty means all
searchQuery	Only reviews whose title or text contains all of these words
publishedAfter, publishedBefore	2026-01-31 or a span like 30 days; the end date includes the whole day
companySize	small (50 or fewer), mid (51-1000), enterprise (over 1000)
sortBy	newest (default), oldest, most_helpful, g2_default

Example input

{
  "startUrls": [
    {"url": "https://www.g2.com/sellers/salesforce"},
    {"url": "https://www.g2.com/products/slack/reviews"}
  ],
  "maxReviewsPerProduct": 50,
  "maxProductsPerSeller": 3,
  "starRatings": ["1", "2"],
  "searchQuery": "support",
  "publishedAfter": "2025-01-01",
  "companySize": ["mid", "enterprise"]
}

⬆️ Output

{
  "reviewId": 11462200,
  "reviewUrl": "https://www.g2.com/products/slack/reviews/slack-review-11462200",
  "productName": "Slack",
  "sellerName": "Salesforce",
  "starRating": 2,
  "title": "Our primary corporate communications channel",
  "likes": "Basically nothing. Its not the best application. I find teams far better",
  "dislikes": "Very poor interface and unorganized messages. Not very feature rich. Poor customer support. Easy to implement",
  "problemsSolved": "It helps with video conferencing and cross functional communication",
  "reviewText": "Basically nothing. ... cross functional communication",
  "publishedAt": "2025-08-08T06:41:43Z",
  "reviewerName": "Abhinanth G.",
  "reviewerRole": "User",
  "companySize": "Enterprise (>1000 emp.)",
  "industry": "Computer Software",
  "country": "India",
  "helpfulVotes": 0,
  "reviewSource": "organic",
  "likelihoodToRecommend": 4,
  "easeOfUse": 3,
  "easeOfSetup": 5,
  "qualityOfSupport": 4,
  "meetsRequirements": 4,
  "switchedFromOtherProduct": true,
  "switchedReason": "Competitive pricing",
  "productId": 3437,
  "productSlug": "slack",
  "productUrl": "https://www.g2.com/products/slack/reviews",
  "input": "https://www.g2.com/products/slack/reviews",
  "scrapedAt": "2026-10-27T09:13:10Z"
}

Empty fields are left out of a row. The run's key-value store holds RUN_SUMMARY with the per-product report.

🤖 For AI agents

One input to copy: {"startUrls": [{"url": "https://www.g2.com/products/slack/reviews"}], "maxReviewsPerProduct": 20, "starRatings": ["1", "2"]}. No credential is needed. A link that is not a G2 product or seller link is skipped, and the status message says why, so an agent can retry with a corrected link.

💡 Use cases

• 🥊 Competitor battlecards: the 1- and 2-star reviews of a rival product that mention "pricing" or "support", with the product the reviewer switched from.
• 📈 Review monitoring: a daily schedule with publishedAfter: "1 day" returns only new reviews of your products and your competitors'.
• 🧪 Product research: likes and dislikes by company size and industry for a whole seller portfolio in one run.
• 🤖 LLM and sentiment pipelines: clean review text with stars, themes and reviewer context, ready for classification or retrieval.

⚠️ Limits, stated plainly

• Likes, dislikes and problems solved come as separate fields when G2 stores exactly three answers, which is most reviews written in recent years (94% of 200 recent ones in a test, 38% of reviews from 2016-2017). Otherwise reviewText holds the whole review and the three fields are empty.
• Reviewer job titles, profile links and LinkedIn data are not returned. The reviewer name is what G2 shows publicly.
• A seller link reads products whose seller name equals the seller's name on G2; products G2 lists under a sub-brand ("Salesforce Labs") need their own link or the sub-brand's seller link.
• G2's AI pros-and-cons summaries and video review files are not included.
• The search service is G2's own and may change. If it does, runs end with zero rows and a message, not with a charge.

❓ FAQ

Do I need a G2 account or cookies? No. Nothing is logged in, and no proxy is needed in normal runs.

How many reviews can I get? All published reviews of a product. Set maxReviewsPerProduct as high as you need; every review is one charged row.

Why did a run return 0 reviews? The status message names the reason per link: no published review matches the filters, the product or seller does not exist on G2, or the link is not a G2 link. Each case ends as a succeeded run and costs only the start.

Can I combine stars, keywords and dates? Yes, on every link type. All filters are applied together by G2's search service, and the status message shows how many reviews match.

What does a seller link return? The reviews of that seller's most reviewed products, each row with productName and sellerName. Raise maxProductsPerSeller to read more of them.

How is the star rating related to the 0-10 score? starRating is G2's 1-5 star value; likelihoodToRecommend is the reviewer's 0-10 answer to "how likely are you to recommend".

What happens if a run is interrupted? A migrated or resurrected run reads the reviews it already saved from its dataset and skips them, so nothing is saved or charged twice.

Can I switch from another G2 Actor? url (a single link) and products (slugs) are read as well as startUrls. Field names follow common conventions (likes, dislikes, problemsSolved, reviewUrl, companySize).

Is it legal? The Actor reads reviews G2 shows publicly. You are responsible for how you use them: respect G2's terms of use, and treat reviewer names as personal data under GDPR and similar laws. Do not use the data for unsolicited email (CAN-SPAM).

🔗 You might also like

• Booking.com Reviews Scraper
• Google Maps Reviews Scraper
• Reddit Search Scraper
• Tech Stack Detector

Keywords: G2 reviews scraper, G2 scraper, G2 reviews API, export G2 reviews, G2 reviews to CSV, G2 review monitoring, scrape G2.com, G2 seller reviews, software reviews scraper, B2B reviews dataset, G2 competitor analysis, G2 negative reviews, G2 pros and cons, SaaS review data, G2 crowd scraper, review sentiment analysis, G2 reviews JSON