|
|
All guides
AramaBaşlangıç

Google Dorking: OSINT with Advanced Search Operators

OSINT TR9 min read

Typing a few words into a search engine and looking at the first page shows you a tiny part of the web. With advanced search operators (dorks) you can narrow a search to a specific site, file type, page title or date range. This guide covers the most useful operators, example queries and the limits to keep in mind.

Basic operators

  • Quotes ("..."): Searches for the exact phrase. "Ahmet Yılmaz" returns pages where the two words appear side by side.
  • Minus (-): Excludes a word. jaguar -car leaves the results about the animal.
  • OR: Searches for either option. "port" OR "terminal"
  • site: Limits the search to a site or domain ending. site:gov.tr, site:reddit.com
  • filetype: Returns specific file types. filetype:pdf, filetype:xlsx
  • intitle: Requires the word to appear in the page title.
  • inurl: Requires the word to appear in the address.
  • before: / after: Narrows Google results by date. after:2024-01-01 before:2024-06-30
Tip: Do not put a space between the operator and the word: site:gov.tr works, site: gov.tr does not.

Combining operators

The real power comes from using operators together. Each new operator cuts the number of results and brings you closer to what you are looking for.

  • site:gov.tr filetype:pdf "faaliyet raporu" 2025: 2025 annual reports of Turkish public bodies.
  • site:linkedin.com/in "OSINT" "Istanbul": public profiles of people working on OSINT in Istanbul.
  • "username" -site:instagram.com: places where a username appears outside Instagram.
  • intitle:"press release" "company name" after:2025-01-01: a company’s recent press releases.
  • site:example.com -inurl:www: a good start for discovering a site’s subdomains.

Dork types: a ready-made query library

The queries below are grouped by type of research. Replace the example values in quotes (Company Name, example.com, username) with your own. Treat each query as a starting point; add or remove operators based on the results.

1. Documents and files

  • site:example.com filetype:pdf: every PDF on a site (reports, brochures, presentations).
  • site:example.com (filetype:xlsx OR filetype:csv): spreadsheets and data files.
  • "Company Name" filetype:pptx: presentations that mention the company.
  • site:gov.uk filetype:pdf "annual report" 2025: annual reports of public bodies.
Tip: Metadata inside documents, such as author name, organisation and creation date, is a clue too. Open documents you download in a metadata viewer.

2. Companies and organisations

  • "Company Name" (site:linkedin.com/company OR site:linkedin.com/in): the company page and public employee profiles.
  • "Company Name" (tender OR contract) site:gov.uk: public tenders and contracts.
  • "Company Name" (fraud OR complaint OR lawsuit): adverse media and complaints.
  • "Company Name" -site:example.com: what others write about the company outside its own site.

3. Usernames and social media

  • "username" (site:reddit.com OR site:github.com OR site:medium.com): a username’s traces on forums, code and blogs.
  • site:x.com "username": posts on X that mention the username.
  • site:t.me "keyword": public Telegram channels and groups.
  • site:instagram.com "place name": public posts about a place.

4. Location and images

  • "place name" "city" (site:tripadvisor.com OR site:foursquare.com): visitor photos and reviews of a place.
  • site:flickr.com "street name" "city": photos of a street from different angles.
  • "text on the sign" "city": the fastest way to locate a sign seen in a photo.
  • site:wikimapia.org "place name": buildings and facilities that are not named on maps.

5. Time and news

  • "event" after:2025-03-01 before:2025-03-08: what was written in the first week of an event.
  • "event" site:reuters.com: searching a news agency’s archive.
  • intitle:"breaking" "place name" after:2025-01-01: breaking news about a place.
  • "quoted sentence" before:2024-01-01: finding when a quote or claim first appeared.

6. Domains and infrastructure

  • site:*.example.com -site:www.example.com: discovering subdomains.
  • "example.com" -site:example.com: pages elsewhere that mention the site.
  • site:example.com inurl:blog OR inurl:news: the site’s blog and news sections.
  • site:example.com intitle:"our team" OR intitle:"about us": pages about the team and the organisation.

7. Maritime

  • "IMO 1234567": every source that mentions a ship by its IMO number.
  • "ship name" (site:marinetraffic.com OR site:vesselfinder.com): the ship’s pages on tracking sites.
  • "ship name" filetype:pdf: port reports, inspection records and accident reports.

For more on researching ships, see our ship tracking guide.

A type to know about but not to use

One kind of dork is aimed at finding file directories, configuration files and admin panels left open on the internet by mistake. Security researchers use these queries only on systems they are authorised to test. We do not include such examples in this guide; if you want to check your own organisation’s site, do it together with your IT team.

Do not rely on one search engine

Each search engine has a different index and different operator support. What you cannot find on Google you may find on Bing, Yandex or DuckDuckGo. Bing and Yandex support site: and quotes; Yandex can give better results for Russian and Eastern European sources. Repeat an important search in at least two engines.

Read the results correctly

  • The result count is an estimate; “about 1,200 results” is not the real number.
  • Search engines can change results by person, language and location. Searching in a private window gives more neutral results.
  • A page may have dropped out of the index but still be online, or appear in the index but be gone. In the second case, check the web archives.

Ethical and legal limits

Dorks sometimes turn up files, admin panels or lists of personal data that were left open by mistake. A file showing up in a search engine does not give you the right to download and use it. In Türkiye, accessing an IT system without permission is a crime under Article 243 of the Turkish Penal Code, and unlawfully obtaining and spreading personal data is a crime under Articles 135 and 136; most countries have similar laws.

Do not force password-protected areas, and do not download or share personal data left exposed. If you find something like that, report it to the organisation concerned.

Use search techniques to research documents, organisations and events, not to monitor people. For the general framework, see our What is OSINT guide.

Put it into practice
Solve OSINT missions built from real photos and climb the leaderboard.
Go to missions

More guides

ArşivDoğrulama

Finding Deleted Content: A Guide to the Wayback Machine and Web Archives

Find old versions of deleted or changed web pages: the Wayback Machine, archive.today and national archives. Reading a site’s history, archiving your findings and the limits.

6 min read14 November 2026
RaporlamaBaşlangıç

How to Write an OSINT Report: Findings, Evidence and Confidence

Turning open source research into a readable, verifiable report: report structure, evidence logging, archiving, confidence language and tips for writing CTF write-ups.

6 min read14 November 2026
DenizcilikGEOINT

Ship Tracking OSINT: IMO, MMSI, AIS and the Shadow Fleet

Tracking ships with open sources: IMO vs MMSI, AIS data and its limits, verifying a ship’s identity, measuring ships on satellite imagery, the shadow fleet and sanctions databases.

6 min read13 November 2026
Due DiligenceFININT

What Is Due Diligence? Vetting Companies and Candidates with OSINT

What is due diligence and when is it done? A step-by-step guide to vetting suppliers and companies with open sources, spotting red flags and running lawful pre-employment screening.

6 min read13 November 2026
OSINTBaşlangıç

What Is OSINT? A Beginner’s Guide to Open Source Intelligence

What is OSINT (open source intelligence), where is it used and how do you learn it? The intelligence cycle, core skills, ethical and legal limits, and a roadmap for getting started.

4 min read12 November 2026
GEOINTGeolocation

How to Geolocate a Photo: A Step-by-Step Guide

How to find where a photo was taken: checking metadata, reading clues in the image, reverse image search, narrowing down with maps and street view, and verification.

4 min read12 November 2026
Görsel AnalizIMINT

Reverse Image Search Guide: Google Lens, Yandex, Bing and TinEye

What is reverse image search and what is it good for? The strengths of Google Lens, Yandex Images, Bing Visual Search and TinEye, with practical tips for better results.

3 min read12 November 2026
Görsel AnalizKişisel Güvenlik

What Is Photo Metadata (EXIF)? How to View and Remove It

What is the EXIF metadata hidden in your photos and what does it carry (location, date, device)? How to view it on phone and computer, remove it before sharing, and use it in OSINT.

3 min read12 November 2026
SOCMINT

Tracing a Username: An Introduction to SOCMINT

How to research open sources starting from a single username: username scanning, verifying matches, pivoting techniques, archives and ethical limits.

3 min read12 November 2026
GEOINTUydu

How to Read Satellite Imagery: A Guide to Free Satellite Sources

Free satellite imagery sources (Sentinel, Landsat, NASA Worldview, Google Earth) and the basics of reading satellite images: resolution, dates, shadows, false colour and change detection.

3 min read12 November 2026
DoğrulamaGörsel Analiz

Fake News and Image Verification: A 5-Step Checking Guide

Check in 5 steps whether a photo or video you see on social media is real: source, first publication, location, time and context. Tips for spotting AI-generated images.

3 min read12 November 2026