Filtering YouTube videos by creation date, analyzing tweets of users included in a certain list, searching strictly by one subreddit - there are hundreds of examples of using them.
Submint a link to video and get it metada:
Snippet (the exact date and time of publication, description, thumbnails links)
Statistics
Geolocation
Status
Content details
(also it's possible to get channel and playlist metadata)
This query search all sql-files, which contain "ivanov" in #Github
(2/12)
Sometimes in this way it's possible find databases with information about orders, customers or employees. Inattentively, site owners forget to close them from indexing in robot.txt.
Such SQL-dumps may contain names, surnames, dates of birth, addresses, phone numbers.
Now look for interesting fragments in Google or Yandex
(by the way, Yandex has a built-in cropping)
It happens that the country can be identified by the model of public transport. If the photo shows "PAZ 4234", it is most likely Russia (but maybe other CIS countries).
Check the details of the company in the official registry of the country in which it is registered (it can often be found on the website of the federal tax inspectorate)