Search engines are designed to find relevant webpages for users based on:
• the keywords the user types into the engine's box (or boxes)
• the search logic the user may use
• and the engine's own search logic

When a user executes a search, the engine does not actually go out on the entire web and search through every document that is out there - rather it searches its own database of keywords and associated webpages. When the user's results are found, the search engine uses its own ranking logic to try to order the results so that (hopefully) the most useful and relevant results are placed at the top and beginning of the results page.

NOTE: The search engine's database is constantly being updated by software searching tools called spiders that regularly go out and scour the internet for new pages. The contents of new pages are then scanned for all significant words (not words like is or the) which are used to categorize the new page for the engine's directories AND to create files in the engine's database that will reference keywords to the associated webpage.