Glassdoor Jobs Scraper

TL;DR

Pull every Glassdoor job listing for a search you'd run by hand: title, company, location, numeric salary range, employer rating, posted date, employment type, experience level. Returns ~30 jobs per page of any Glassdoor search URL, normalized to a stable schema with ISO 8601 dates and typed enums. Built for recruiters sourcing engineers in their metro, lead-gen agencies building outreach lists, salary-benchmarking analysts, and AI agents pulling daily job market signals.

Try it on a small dataset, then let us know what you think in a review.

---

What does Glassdoor Jobs Scraper do?

This actor takes a Glassdoor search URL (the exact one you'd paste from your browser after running a search on glassdoor.com) and returns every job listing on that page as a clean, structured record. Each record contains:

• Identity — job ID, listing URL, job title, hiring company
• Location — city/state string, plus a derived isRemote boolean
• Money — raw salary text ("$120K - $180K") plus numeric salaryEstimateLowUsd and salaryEstimateHighUsd fields. Per-hour salaries are automatically annualized at 2080 hours so you can sort across hourly and salaried roles in one column.
• Recency — both the relative posting age ("2d", "30+d") and an ISO 8601 postedDate plus integer daysPosted, so you can filter daysPosted < 7 directly in SQL or pandas.
• Classification — typed enums for employmentType (FULL_TIME / PART_TIME / CONTRACT / INTERNSHIP / ...) and experienceLevel (INTERN / ENTRY / MID / SENIOR / STAFF_PRINCIPAL / EXECUTIVE) inferred from the title and salary text.
• Employer signal — companyRating (Glassdoor's 1.0-5.0 score) so you can filter for top-rated employers in one query.
• Description — the first few sentences of the listing as it appears in Glassdoor's preview.
• Drop-into-LLM card — a 300-500 character agentMarkdown field per record. Paste any row into Claude, ChatGPT, Slack, or your CRM and it renders as a clean summary with title, company, location, salary, rating, and link — no extra formatting needed.
• Self-filter signal — a fieldCompletenessScore (0-100) per record so you can drop rows that are too sparse for your downstream pipeline.

The actor does not require you to bring an API key. Paste a Glassdoor search URL, set maxItems, run.

Why scrape Glassdoor?

Glassdoor's job inventory is the broadest single source of US job postings outside LinkedIn — millions of active listings across every industry, with the unique addition of employer-rated context (rating, review count, salary signal) that no plain job board carries. The listings themselves are public, but the data is locked inside JavaScript-rendered search results behind anti-bot defenses tuned to block naive scraping. Plain HTTP doesn't work — Glassdoor's bot detection blocks curl and basic headless Chrome. This actor handles that for you with a stealth-Chromium backbone running on a private Oracle Cloud egress, so you get the data without managing your own proxy infrastructure.

The dataset is valuable because it sits between three buyer workflows:

1. Lead enrichment for cold outreach — every company hiring an engineer in your metro is a warm sales target.
2. Salary intelligence — numeric salaryEstimateLowUsd / salaryEstimateHighUsd lets you benchmark roles in minutes.
3. Hiring market timing — daysPosted plus a daily watchlist run tells you which companies are scaling, freezing, or quietly hiring.

Who needs this?

• Technical recruiters sourcing engineers, PMs, and designers in specific metros, who today open Glassdoor and copy-paste 30 listings into a spreadsheet by hand.
• B2B sales reps and lead-gen agencies building outreach lists for ATS vendors, payroll, dev tools, and recruiting SaaS — every company hiring an engineer is a warm prospect.
• Salary benchmarking analysts at HR consulting firms who need fresh numeric salary ranges for compensation surveys.
• Competitive intelligence teams monitoring which companies in their space are scaling teams, opening new offices, or pivoting to new tech stacks (revealed via job titles).
• AI agents and LLM-powered workflows that need a structured, schema-stable view of the job market — agentMarkdown plus a fixed schema means your LangChain / CrewAI / Claude-Code pipeline doesn't break when Glassdoor tweaks their HTML.
• Solo founders and indie hackers validating product ideas by counting how many companies post specific job titles week-over-week.

How to use Glassdoor Jobs Scraper

1. Open glassdoor.com and run the search you want (keyword + location + filters).
2. Copy the URL from your browser's address bar. It looks like https://www.glassdoor.com/Job/new-york-software-engineer-jobs-SRCH_IL.0,8_IC1132348_KO9,26.htm.
3. Paste that URL into the jobsSrpUrl input on this actor.
4. Set maxItems (start with 10 for a trial; production runs commonly use 100-500).
5. Click Start. Each page returns ~30 listings. The actor stops when maxItems is reached, when maxPages is exhausted, or when a page returns zero new jobs.
6. Download the dataset as JSON, CSV, or XLSX, or pipe it directly into your destination via Apify's API, webhook, or one of the 100+ integrations (Slack, Make, Zapier, Google Sheets, Airtable, etc.).

How much will scraping Glassdoor cost?

Pay-per-result tiered pricing. The apify-actor-start fee is $0.001 per run (a fraction of a cent — fires once whether you save 5 records or 5,000). The per-job-listing event price varies by your Apify plan:

Apify plan	Per-job price	Cost per 1,000 jobs
FREE	$0.0003	$0.30
BRONZE	$0.0005	$0.50
SILVER	$0.0006	$0.60
GOLD	$0.0007	$0.70
PLATINUM	$0.0007	$0.70
DIAMOND	$0.0007	$0.70

A typical daily watchlist run pulling 100 jobs costs $0.03–$0.07 in event fees plus a few cents of Apify platform usage (compute + bandwidth, billed by Apify directly). You only pay for records actually saved to the dataset — runs that return zero results cost only the $0.001 start fee.

Free-tier Apify users get $5 in monthly platform credit, which yields roughly 1,600+ free trial records on this actor before any out-of-pocket spend.

Is it legal to scrape Glassdoor?

This actor scrapes only public-facing job listing pages that Glassdoor serves to any visitor without login. It does not bypass authentication, does not access private review or salary endpoints, and does not interact with any rate-limited API. The actor honors a reasonable request cadence and is designed for personal lead-gen, market intelligence, and analytics workflows.

That said: public data is not the same as "free to redistribute". Glassdoor's Terms of Service restrict commercial republication of their content. If you plan to package this data into a downstream product you sell, get an opinion from your own counsel — not from us. For internal recruiting, sales prospecting, salary benchmarking, and AI-agent automation, the pattern is common industry practice. For building a competing job board or selling Glassdoor's listings to third parties, it isn't.

Examples

Software engineers in NYC

{
  "mode": "jobs",
  "jobsSrpUrl": "https://www.glassdoor.com/Job/new-york-software-engineer-jobs-SRCH_IL.0,8_IC1132348_KO9,26.htm",
  "maxItems": 50,
  "maxPages": 2
}

Use case: a recruiter at a YC company tracking which NYC startups are hiring backend engineers this week.

Product managers in San Francisco

{
  "mode": "jobs",
  "jobsSrpUrl": "https://www.glassdoor.com/Job/san-francisco-product-manager-jobs-SRCH_IL.0,13_IC1147401_KO14,29.htm",
  "maxItems": 100
}

Use case: a PM tools vendor (Productboard, Linear) building a weekly outreach list of companies that just opened a PM role.

Remote data scientist roles (US-wide)

{
  "mode": "jobs",
  "jobsSrpUrl": "https://www.glassdoor.com/Job/united-states-remote-data-scientist-jobs-SRCH_IL.0,13_IN1_KO14,35.htm",
  "maxItems": 200
}

Use case: a data-tools company (dbt, Hex, Mode) sourcing leads for self-serve outreach.

Sales reps in Chicago

{
  "mode": "jobs",
  "jobsSrpUrl": "https://www.glassdoor.com/Job/chicago-sales-representative-jobs-SRCH_IL.0,7_IC1128808_KO8,28.htm",
  "maxItems": 50
}

Use case: a CRM reseller building a list of Chicago-area companies expanding their sales teams.

Junior dev roles for early-career sourcers

{
  "mode": "jobs",
  "jobsSrpUrl": "https://www.glassdoor.com/Job/united-states-junior-developer-jobs-SRCH_IL.0,13_IN1_KO14,30.htm",
  "maxItems": 100
}

Use case: a bootcamp's career services team aggregating entry-level openings for cohort placement.

Marketing managers across major metros

{
  "mode": "jobs",
  "jobsSrpUrl": "https://www.glassdoor.com/Job/marketing-manager-jobs-SRCH_KO0,17.htm",
  "maxItems": 300,
  "maxPages": 10
}

Use case: a marketing agency benchmarking salary ranges and rating distributions across competitors.

Input parameters

Field	Type	Required	Default	Description
mode	enum	no	"jobs"	Mode of operation. v0.1 supports "jobs" only. "reviews" and "both" are reserved for a future build.
jobsSrpUrl	string (URL)	yes for mode=jobs	—	A pre-resolved Glassdoor search URL. Run the search you want on glassdoor.com first, then paste the URL from your address bar.
companyReviewsUrl	string (URL)	no	null	Reserved for v0.2 reviews mode.
maxItems	integer	no	10	Cap on total records saved. Use 10-25 for trials; bump to 100-500 for production.
maxPages	integer	no	1	Pagination cap. Each page returns ~30 jobs.

Glassdoor output format

job_listing record

Field	Type	Description
recordType	enum	Always "job_listing" for this record type.
outputSchemaVersion	string	ISO date of this schema version ("2026-05-13"). Bumps on any breaking change.
recordId	string	Stable across runs. Format: gd-jl-<jobId>.
jobId	string	Glassdoor's internal job ID.
listingUrl	string (URL)	Direct link to the job detail page.
title	string	Job title.
company	string	Hiring company name.
location	string	City, state, or country.
isRemote	boolean	Derived from the location string.
postedRelative	string	Relative age as shown on Glassdoor ("2d", "30+d").
postedDate	string (ISO 8601)	Computed YYYY-MM-DD.
daysPosted	integer	Days since posting.
employmentType	enum	FULL_TIME / PART_TIME / CONTRACT / TEMPORARY / INTERNSHIP / UNKNOWN.
experienceLevel	enum	INTERN / ENTRY / MID / SENIOR / STAFF_PRINCIPAL / EXECUTIVE / UNKNOWN.
salaryEstimateText	string	Raw salary text as Glassdoor displays it.
salaryEstimateLowUsd	number	Numeric low end. Per-hour values annualized at 2080 hours.
salaryEstimateHighUsd	number	Numeric high end. Same normalization.
companyRating	number	Glassdoor employer rating (1.0-5.0).
description	string	Snippet from the listing's description.
agentMarkdown	string	Drop-into-LLM ready summary card (~300-500 chars).
fieldCompletenessScore	integer	0-100. Filter sparse rows.
scrapedAt	string (ISO 8601)	Timestamp the record was saved.

Example output

{
  "recordType": "job_listing",
  "outputSchemaVersion": "2026-05-13",
  "recordId": "gd-jl-1010130768473",
  "jobId": "1010130768473",
  "listingUrl": "https://www.glassdoor.com/job-listing/php-software-engineer-cybercoders-JV_KO0,21_KE22,33.htm?jl=1010130768473",
  "title": "PHP Software Engineer",
  "company": "CyberCoders",
  "location": "United States",
  "isRemote": false,
  "postedRelative": "2d",
  "postedDate": "2026-05-12",
  "daysPosted": 2,
  "employmentType": "UNKNOWN",
  "experienceLevel": "MID",
  "salaryEstimateText": "$100K - $120K (Employer provided)",
  "salaryEstimateLowUsd": 100000,
  "salaryEstimateHighUsd": 120000,
  "companyRating": 3.9,
  "description": "Bachelors degree in computer science or a related field, coding bootcamp, or equivalent practical experience...",
  "agentMarkdown": "**PHP Software Engineer** @ CyberCoders\n- 📍 United States\n- 💰 $100K - $120K (Employer provided)\n- 🏷 UNKNOWN · MID\n- 🕒 2d\n- ⭐ 3.9\n- 🔗 https://www.glassdoor.com/job-listing/...",
  "fieldCompletenessScore": 90,
  "scrapedAt": "2026-05-14T00:59:56.188Z"
}

During the Actor run

The actor reads your input, validates the search URL, then walks pages serially (one fetch at a time) through a stealth-Chromium backbone. Each page takes ~20-30 seconds end-to-end including the bot-defense challenge resolution. A AGENT_BRIEFING.md digest is written to the run's key-value store at finish, summarizing record count, salary distribution, remote share, and the top 10 most-complete records — useful for daily Slack drops or LLM context windows.

If the actor returns zero records on a given page, pagination stops automatically — this prevents wasted runs on URLs that resolve to a redirect or filter-collapsed result.

FAQ

How does Glassdoor Jobs Scraper work?

You provide a real Glassdoor search URL (one you'd paste from your own browser). The actor fetches each results page using a stealth-Chromium backbone designed to clear Glassdoor's anti-bot defenses, parses the ~30 job cards on each page using current selectors (not legacy ones that broke months ago), and saves a typed record per job. Repeat for as many pages as you configure.

Where do I get the Glassdoor search URL?

Open glassdoor.com, run the exact search you want (e.g., "software engineer in New York"), then copy the URL from your address bar after the page loads. The URL will contain codes like SRCH_IL.0,8_IC1132348_KO9,26.htm — those are Glassdoor's internal location and keyword IDs. The actor passes them through verbatim.

Can I monitor only new records since the last run?

Not yet in v0.1 — but the record IDs are idempotent (gd-jl-<jobId> is the same across runs), so you can dedupe downstream by storing seen IDs in a key-value store, database, or Apify's own dataset. A watchlistMode flag that auto-emits only new records since last run is planned for v0.2.

Can I integrate this with Make, Zapier, n8n, Slack, or my CRM?

Yes — Apify provides 100+ native integrations including Slack, Make, Zapier, n8n, Google Sheets, Airtable, Salesforce, HubSpot, Webflow, and a generic webhook output. You can also call the actor directly from any HTTP client using your Apify API token; the dataset is queryable as JSON, CSV, XLSX, or RSS.

Can I use this from Python or Node?

Yes — Apify ships an Apify Python client and Apify JS client for actor runs and dataset retrieval. Both are pip / npm installable and authenticate with the same API token you use elsewhere.

Why does this cost more than the free Glassdoor scrapers on the Store?

Free scrapers run on minimal infrastructure with no anti-bot defense and break monthly when Glassdoor tweaks their HTML or rate-limits their proxy pool. Their daily auto-test silently fails, their dataset goes empty, and your downstream pipeline starts returning zero rows without warning. The per-result cost on this actor ($0.0003–$0.0007) pays for an actively maintained stealth backbone, a versioned schema with bump-on-breaking-change policy, and a daily reliability test that catches breakage in 24h instead of weeks.

Will Glassdoor block me?

The actor uses a stealth-Chromium backbone on a low-traffic egress, so individual buyers don't carry their own IP reputation risk. Glassdoor's defenses target naive curl and obvious headless Chrome — this actor passes their commercial-grade fingerprint checks (verified on bot.sannysoft.com and browserscan.net at "Normal" / no detection). For very high-volume runs (10k+ jobs/day), throttle with smaller maxItems per run instead of one mega-run.

Why doesn't v0.1 include reviews or salaries?

Glassdoor's /Reviews/ and /Salaries/ paths sit behind a stricter Cloudflare layer than the /Job/ path. The current cloak backbone clears the jobs path cleanly but doesn't yet auto-solve the interactive Cloudflare challenge on reviews. v0.2 will add reviews using a humanized cloak profile.

How fresh is the data?

Each run fetches live HTML from Glassdoor at run time — there is no caching layer between the source and your dataset. Jobs posted in the last hour appear with daysPosted: 0.

What is agentMarkdown?

It's a pre-formatted ~300-character markdown summary baked into every record. Drop any row into Claude, ChatGPT, Slack, a CRM note field, or an email draft and it renders as a clean card with title, company, location, salary, rating, and link — no string-templating on your end. It exists because AI agent buyers were the loudest "this is the best output schema we've seen" feedback on our other actors; we ship it on every skootle actor now.

Can I run this on a schedule?

Yes — Apify has native scheduling: pick a cron expression, set the input once, and Apify will run the actor automatically and pipe results into your destination via integrations or webhooks. Most buyers run this as a daily watchlist.

Why choose Glassdoor Jobs Scraper

• Numeric salary mirrors — every salary field has both the raw text and a typed numeric (per-hour values annualized) so you can sort, filter, and average without writing parser code.
• ISO 8601 everywhere — postedDate and scrapedAt are real ISO timestamps. Sort by daysPosted directly in pandas, BigQuery, Postgres, or Airtable.
• Typed enums for classification — employmentType and experienceLevel are stable enum values, not raw source strings. Your downstream code doesn't have to handle 12 variants of "Senior" vs "Sr.".
• Versioned schema — outputSchemaVersion: "2026-05-13" literal on every record plus a bump-on-breaking-change policy. Your pipeline doesn't silently break when we ship improvements.
• Idempotent record IDs — recordId is stable across runs for the same job. Dedupe downstream without state.
• Field completeness score — fieldCompletenessScore: 0-100 per record so you can filter sparse rows in one query.
• AGENT_BRIEFING.md per run — run-level markdown digest written to the key-value store. Pipe it into Slack or feed it to an LLM context window for a daily summary.
• Currently maintained selectors — most free Glassdoor scrapers grep legacy DOM classes that broke months ago and now return zero parsed cards on real pages. This actor targets the current JobCard_* taxonomy and is verified against a fresh page on every release.

Your feedback

Hit a bug or want a feature? Open an issue on the Issues tab rather than the reviews page, and we'll fix it fast (typically within 48 hours).

Other Skootle actors you might want to check

• 🏢 Wellfound Jobs Scraper — startup hiring inventory with company funding signals.
• 📋 Greenhouse Jobs Scraper — every role on any company's Greenhouse board.
• 🌱 Lever Jobs Scraper — same but for Lever-hosted boards.
• 🇺🇸 SAM.gov Federal Contracts — federal contract opportunities with USAspending award history.
• 📰 HackerNews Watchlist — daily HN signal filtered for your keywords.
• 🐙 GitHub Trending Repos — daily/weekly/monthly trending across languages.

Support and contact

• 🐛 Open an issue for bugs or feature requests
• 📝 Leave a review after a successful run
• 🌐 skootle on Apify for the full actor catalog