Definition:

A search engine is a system that locates information in response to a query. On the web, it discovers documents, processes them into an index and orders the answers it considers most relevant for display on a results page.

When someone types or speaks a query, the search engine does not scan the entire web at that moment. It consults information collected earlier, interprets the search intent and retrieves content that may answer it.

How a search engine works

Although each service uses its own technology, web search generally relies on four connected processes:

Crawling and discovery. A crawler visits known addresses and follows links or sitemaps to find new or updated content.
Processing and indexing. The system analyses the document’s text, images, language, links and other elements. If the document can be included, it stores a representation in its index.
Retrieval and ranking. When the query arrives, different algorithms select candidates and assess signals such as relevance to the search, quality, freshness and the user’s context.
Presentation. The results page may combine links, images, videos, maps, news, direct answers and other formats. Its composition changes with the query.

Crawling a page does not guarantee that it will be indexed, and indexing it does not secure a particular position. These are separate stages of the process.

Types of search engines

Search engines do not all cover the same information. Image and video tabs can also be parts of a general search engine rather than independent engines. Three common groups can be distinguished by scope:

General search engines. They cover a broad portion of the web and provide several result formats for many different queries.
Specialised or vertical search engines. They focus on a field such as academic publications, products, jobs, images or videos.
Internal search engines. They search within a website, application, catalogue or an organisation’s documentation.

Differences from a browser and a metasearch engine

A browser, such as Firefox, Chrome or Safari, is the application that opens and renders web pages. A search engine helps people find them. The two are often confused because browsers accept searches in the same bar used for web addresses.

A metasearch engine sends a query to several search engines or sources and brings their responses together. It may apply its own filters, but it does not need to maintain a complete web index like a general search engine.

Examples of search engines

Google and Bing are general search engines with broad indexes and several result formats. Yahoo provides a search portal whose result infrastructure depends on agreements with other providers. Baidu is primarily aimed at the Chinese market and language, while Yandex has a strong presence in Russian-speaking markets. DuckDuckGo focuses on privacy and combines its own and external sources to produce results.

These services do not crawl, index or rank information in the same way. A page may therefore appear in one engine, be absent from another or occupy different positions for an equivalent query.