Glassdoor Company Reviews Scraper

Scrape company reviews and ratings from Glassdoor — one row per review, carrying the company context.

Glassdoor sits behind Cloudflare Bot Management (the Just a moment… interstitial), which is why plain HTTP scrapers get blocked instantly. This actor runs a real browser with anti-detection, reads reviews straight from Glassdoor's own Next.js __NEXT_DATA__ Apollo GraphQL cache, and rotates a fresh IP and a fresh browser fingerprint whenever a page is challenged — so it returns reviews or an honest BLOCKED, never silent-empty.

Input

Field	Notes
mode	sample (default) returns a few built-in example reviews instantly — no proxy, no charge — so you can preview the output shape for free. Switch to live to scrape real Glassdoor reviews.
companyUrls	Glassdoor reviews URLs, e.g. https://www.glassdoor.com/Reviews/Google-Reviews-E9079.htm. The employer id is read from the E{id} in the URL.
companyNames	Company names to resolve to an employer first (e.g. Netflix) when you don't have the URL.
maxReviews	Reviews to collect per company (paginates by 10s).
proxyConfiguration	Optional. Runs on rotating datacenter addresses; metered groups (RESIDENTIAL/GOOGLE_SERP) are not offered and are replaced with datacenter. Supply your own servers in proxyUrls to use those instead.

Quick start: the default sample mode lets you try the actor for free and see exactly what a review row looks like. When you're ready for real data, set mode to live.

Provide companyUrls, companyNames, or both. Submitting neither returns an uncharged BAD_INPUT diagnostic row (not a crashed run).

Output

One row per review, each carrying its company context:

• Company: companyName, employerId, overallRating, reviewCount, companyUrl
• Review: ratingOverall, summary, pros, cons, advice, jobTitle, employmentStatus, isCurrentJob, location, reviewDate, recommends, ceoApproval, businessOutlook, and sub-ratings (ratingWorkLifeBalance, ratingCultureAndValues, ratingCareerOpportunities, ratingCompensationAndBenefits, ratingSeniorLeadership, ratingDiversityAndInclusion).

Nullable fields: Glassdoor reviewers fill these in optionally, so most review-level fields can be null when the reviewer left them blank — advice, jobTitle, location, employmentStatus, isCurrentJob, recommends, ceoApproval, businessOutlook, and any of the sub-ratings. Every row always carries the company context plus at least one of summary / pros / cons / ratingOverall (rows with none of these are dropped, never charged).

Each successful review is charged as one review unit. Blocked, bad-input, or empty runs return a coded diagnostic row (BLOCKED / BAD_INPUT / NO_RESULTS, all with ok:false) and are not charged.

Pricing

$2.50 per 1,000 reviews ($0.0025 each), with no run-start fee. Flat rate — no volume tiers, no plan gates — and you are charged only for reviews actually returned.

That single per-review price is the whole bill. The real browser, the anti-detection work and the IP rotations Glassdoor's Cloudflare layer forces are all absorbed in it — there is no separate compute or proxy line item, and no charge at all for sample mode, so previewing the output shape is free. A run that is challenged all the way to BLOCKED costs nothing.

Notes

Data is read from Glassdoor's embedded GraphQL cache (props.pageProps.apolloCache), not brittle HTML selectors. Cloudflare challenges trigger a full browser + IP rotation (up to 7 times) before giving up with a BLOCKED diagnostic.