Glassdoor Reviews Scraper

Scrape Glassdoor employee reviews — pros, cons, ratings, job titles, and sub-scores — using AI-powered extraction that handles layout changes automatically.

30+ reviews/run · Company ratings · Pros & cons · Sub-scores · Search by name or URL · JSON & CSV

---

What You Get

• ✅ Company overall rating (out of 5) and % who recommend to a friend
• ✅ Total review count for the company
• ✅ Per-review: headline, star rating, pros, cons, date, job title, location
• ✅ Employment status (current or former employee)
• ✅ Advice to management section (when present)
• ✅ Sub-ratings: Work/Life Balance, Culture & Values, Career Opportunities, Compensation & Benefits
• ✅ Search by company name or provide direct Glassdoor URLs
• ✅ Automatic pagination across review pages

---

Use Cases

🏢 Employer Brand Research

Understand how employees across the industry rate work-life balance, compensation, and career growth — benchmark yourself against competitors.

{
  "companyUrls": [
    "https://www.glassdoor.com/Reviews/Google-Reviews-E9079.htm",
    "https://www.glassdoor.com/Reviews/Microsoft-Reviews-E1651.htm"
  ],
  "maxReviews": 50
}

📊 HR & People Analytics

Collect reviews from your own company listing to track sentiment over time — identify patterns in cons before they drive attrition.

🔍 Pre-Interview Research

Gather employee perspectives on company culture, management style, and interview process from a specific employer before applying.

🤖 NLP & Sentiment Analysis

Build labeled datasets of workplace sentiment from structured pros/cons text — ideal for training aspect-based sentiment models for HR tech.

💼 M&A Due Diligence

Assess acquisition targets by analyzing employee satisfaction, leadership trust, and culture scores across hundreds of reviews.

---

Quick Start

1. Click Try for free above
2. Add company URLs from Glassdoor in Company URLs (e.g., https://www.glassdoor.com/Reviews/Google-Reviews-E9079.htm)
3. Or search by Companies (enter company names like ["Google", "Stripe"])
4. Set maxReviews to control how many reviews to collect per company
5. Optionally set minRating to filter out reviews below a star threshold
6. Run and download as JSON or CSV

---

Input

Field	Type	Default	Description
companyUrls	string[]	—	Direct Glassdoor review page URLs
companies	string[]	—	Company names to search on Glassdoor
maxReviews	number	30	Max reviews to collect per company
minRating	number	—	Minimum star rating to include (e.g., 3 to skip 1-2 star reviews)

At least one of companyUrls or companies must be provided.

---

Output JSON Examples

Company record:

{
  "type": "company",
  "companyName": "Stripe",
  "overallRating": 4.1,
  "recommendRate": 82,
  "totalReviews": 1847,
  "url": "https://www.glassdoor.com/Reviews/Stripe-Reviews-E671923.htm"
}

Review record:

{
  "type": "review",
  "companyName": "Stripe",
  "title": "Great engineering culture, high expectations",
  "rating": 4,
  "pros": "Incredibly smart colleagues, excellent compensation, real ownership of projects.",
  "cons": "Fast pace can be stressful. Lots of context switching.",
  "date": "2025-01-08T00:00:00.000Z",
  "jobTitle": "Software Engineer",
  "location": "San Francisco, CA",
  "employmentStatus": "Current Employee - 2 years",
  "advice": "Invest more in onboarding for new engineers.",
  "ratingWorkLifeBalance": 3,
  "ratingCultureAndValues": 5,
  "ratingCareerOpportunities": 4,
  "ratingCompensationAndBenefits": 5
}

---

Pricing

Volume	Estimated cost
1 company, 30 reviews	~$0.05
5 companies, 30 reviews each	~$0.25
10 companies, 50 reviews each	~$0.80

Apify free tier includes $5/month of compute. Note: this actor uses Stagehand AI extraction as a fallback, which may incur small LLM costs when embedded data is unavailable.

---

Why This Scraper

AI-powered extraction that self-heals. Glassdoor changes its page layout frequently, breaking CSS-selector-based scrapers within days. This actor uses StagehandCrawler with an LLM fallback — when the embedded JSON is absent, AI reads the page and fills the schema automatically, regardless of how Glassdoor's layout changes.

Apollo/Next.js state extraction first. The primary extraction strategy reads Glassdoor's embedded React hydration data — zero LLM cost, zero CSS selectors. The AI is only used when the embedded state is incomplete.

Cloudflare bypass. Glassdoor is protected by Cloudflare. The actor detects Cloudflare challenges and waits for them to resolve before proceeding.

Search by name or direct URL. Provide a company name and the actor finds the correct Glassdoor profile via search — or pass direct URLs to skip search entirely.

Sub-scores extracted. Work/Life Balance, Culture & Values, Career Opportunities, and Compensation & Benefits are all captured as separate numeric fields — not just the overall rating.

---

FAQ

Q: Do I need a Glassdoor account? A: No account is required. The actor accesses public review pages using residential proxies.

Q: Why does this cost more than other scrapers? A: Glassdoor is one of the hardest sites to scrape — Cloudflare protection, frequent layout changes, and login popups all add overhead. The AI fallback ensures you always get data even when the embedded JSON is absent.

Q: What is the difference between companyUrls and companies input? A: companyUrls takes direct Glassdoor review page URLs (fastest). companies takes company names and runs a Glassdoor search first to find the correct page — adds one extra step but is easier to use.

Q: Are reviews from anonymous employees included? A: Yes — Glassdoor allows anonymous reviews. The jobTitle may be present even when the author name is not.

Q: How accurate is the AI extraction fallback? A: The Stagehand AI uses GPT-4.1-mini with a strict Zod schema, so extracted fields are validated types. In testing, accuracy is high for structured fields like ratings and dates. Pros/cons text is captured verbatim.

---

⭐ If this actor saves you time, please leave a review — it helps other developers find it.