Skip to content

How to Scrape Glassdoor: Job Titles, Salaries, and Company Ratings

Glassdoor combines job listings, company profiles, employee reviews, salary figures and interview reports in one place. Because much of that content loads dynamically and is guarded against automation, a plain HTTP request usually returns an incomplete page. This guide explains how to target each data type, validate selectors before running a job, and export clean structured results.

Quick start: listings with extraction rules

A practical route is: confirm the page type and URL pattern, test each selector in browser developer tools, request the page with JavaScript rendering enabled, extract into structured fields, then save the result. The example job-search URL in module code0006 shows the listing pattern to start from. For listings, the extraction rules in module code0007 collect title, company, location, salary and rating from each job card, and module code0009 shows the request that returns that data as JSON. Before running a full job, test a selector such as the one in module code0008 in the browser console; if it prints the expected text, the selector is valid.

Map the site before you scrape

Glassdoor organises content by company and search term, and the company ID in the URL identifies a specific employer. Overview pages describe the company, Reviews pages hold employee feedback, Job pages hold search results, and Salaries pages hold pay data. Keeping the company ID separate from the company name matters when you build URLs programmatically, because the same name can appear with different IDs. The four URL patterns listed in the source under this heading should be treated as templates, not as guaranteed live destinations.

Set up Python and the client

Install Python and confirm the version with module code0002. Create an account, copy your key from the dashboard into {{OWNER_API_KEY}}, and initialise the client as shown in the private module code0004. The SDK handles authentication, request formatting and response parsing, so you can focus on selectors. When content loads on scroll, a JavaScript scenario can click, scroll, dismiss pop-ups or fill forms before extraction; module code0005 gives a scroll-and-wait example useful for incrementally loaded reviews and interviews.

Choose the right extraction for each data type

Reviews: module code0010 targets each review element and returns title, rating, position, date, pros, cons and management advice, with a page parameter for pagination. Salaries: module code0011 returns job title, salary range, base pay, additional pay and sample size per row. Interviews: module code0012 returns the question, job title, difficulty and interview experience text. A useful decision check is to match the page type first — reviews, salaries and interviews each need their own URL template and selector set, so reusing a listings rule on a reviews page will return empty fields even if the request succeeds.

Export to JSON or CSV

JSON is the default response format and suits programmatic processing; module code0013 saves it directly to a file. CSV suits spreadsheets and charting; module code0014 writes Title, Company, Location, Salary and Rating columns for the listings example. Decide the export format before the run, because field names in CSV headers should match the extraction output keys exactly.

Reliability checklist: loading, paging and blocks

Waiting for elements: use wait_for with a selector such as the one in module code0015 so extraction starts only after listings appear. Pagination: basic listings can be paged by modifying the URL, as in module code0016, while deeper review or salary history may need scrolling scenarios or GraphQL extraction. Responsible use: space out requests, collect only what you need, check robots.txt, and consider an official API where one is available. Anti-bot handling combines user-agent rotation (module code0017), IP rotation with optional country selection (module code0018), and full JavaScript rendering (module code0019).

Why collect this data

Salary benchmarks help compare pay across roles and locations, review sentiment reveals culture themes, benefits data supports perk comparisons, and interview reports show how hiring is run. Recruiters can track employer reputation, satisfaction trends and recurring praise or complaints, and benchmark against competitors. Investors may treat sentiment, hiring volume, executive approval and pay growth as signals alongside financial analysis. A concrete example: a startup comparing salary ranges for a role before setting its own pay can use the salary extraction in module code0011, then export the rows with module code0014 for comparison in a spreadsheet.

Start responsibly

Scrape responsibly: respect Glassdoor's servers, check robots.txt, and confirm legality for your use case. To start, create an account, use {{OWNER_FREE_CREDITS}} free API credits, and run the examples in this guide for listings, reviews and salaries. Setup takes less than five minutes, and you save the hours you would otherwise spend building and maintaining your own scraping infrastructure.

Frequently asked questions

Is scraping Glassdoor legal?

Scraping publicly available information is generally treated as legal, but responsible use matters. Avoid republishing content verbatim, respect Glassdoor's terms of service on data use, and seek legal advice for your specific use case.

Why does Glassdoor block my scraper?

Glassdoor protects its data and servers by detecting patterns such as many requests from one IP, missing cookies or headers, and bot-like behaviour. A browser-rendering scraping service helps by mimicking real user behaviour through proxies, headers and JavaScript rendering.

Can I get salary info from Glassdoor listings?

Yes, when salary information is present in a listing it can be extracted. Not every listing includes it, but the .salarySnippet selector shown in the examples targets that field.

Does it work for logged-in Glassdoor pages?

Some login-protected content can be handled with cookies extracted from a browser and passed with the request. For more complex cases, a JavaScript scenario can automate the login process.