Web search engine

Web search engine
Type of site	Search engine
Available in	Multiple
Owner	Various
Created by	Various
Commercial	Varies (mixed)
Registration	Optional (varies by engine)
Launched	1990 (Archie)
Current status	Active
Written in	Various (C, C++, Java, Python, Go, Rust)

A web search engine (or simply search engine) is a software system that searches the World Wide Web for content matching a user's query and returns a ranked list of web pages, images, videos, news, or other items.^[1] Search engines are among the most-used services on the internet: as of 2024, Google Search alone handled more than 90% of global desktop and mobile search traffic.^[2]

Also known as: internet search engine, web search, search engine.

What is a web search engine?

A web search engine is a program or service that lets a user enter keywords or a question and returns a ranked list of results drawn from an index of web pages.^[1] Unlike a web directory, which is curated by hand, a search engine builds its index automatically by crawling the web. Most engines also offer vertical search — dedicated tabs or modes for images, videos, news, maps, shopping and knowledge panels — alongside the default "All" or "Web" view.

The core components are:

a crawler (or spider) that fetches pages by following links;
an indexer that parses fetched pages and stores them in a searchable inverted index;
a ranker that scores results against the query;
a query processor and front end that accept the user's search and display results.

What it means

The term "search engine" distinguishes these systems from web directories such as early Yahoo! or the Open Directory Project, which were edited by humans. Modern engines are largely automated, though they apply SEO-aware heuristics, spam filtering and ranking algorithms to decide what appears first.

Who made it

No single inventor created the web search engine. Several systems appeared independently between 1990 and 1994:^[3]

Archie — created in 1990 by Alan Emtage, Peter Deutsch and Bill Heelan at McGill University; indexed FTP archives rather than web pages.
Veronica and Jughead — Gopher-era indexers, from 1992–93.
Wandex and Aliweb — 1993, among the first Web crawlers.
WebCrawler — 1994, the first full-text search engine for the Web.
Lycos — 1994, a major early commercial engine.
AltaVista — 1995, the first engine to gain mass adoption for full-text web search.
Google — 1998, founded by Larry Page and Sergey Brin; introduced PageRank and became the dominant engine.
Bing — 2009, Microsoft's current engine, successor to MSN Search and Live Search.
DuckDuckGo — 2008, emphasising privacy and no personalisation.
Brave Search — 2021, an independent index with its own crawler.

When it started

The first search tools appeared in 1990 with Archie for FTP archives.^[3] Web-specific search engines followed from 1993–94, and the modern ranking-based era began with Google's PageRank in 1998.^[3]

Types of search engine

General web search engines — Google, Bing, Yahoo Search, DuckDuckGo, Brave Search, Yandex, Baidu, Naver, Seznam.
Vertical search engines — specialised in one domain, such as Google Scholar (academia), Google News, PubMed (biomedical), Kayak (travel) or Amazon (shopping).
Metasearch engines — combine results from several engines, e.g. Dogpile, MetaGer, SearXNG.
Privacy-focused engines — no logging or personalisation, e.g. DuckDuckGo, Startpage, Mojeek, Qwant, SearXNG.
Peer-to-peer search engines — index built and shared across users, e.g. YaCy.
Enterprise search engines — search an organisation's own content, e.g. Elasticsearch, Apache Solr.
Knowledge engines — return structured answers alongside links, e.g. Wolfram Alpha, and the "knowledge panel" style of Google, Bing and AWFL.

How it works

Crawling — a bot fetches pages by following links from seeds such as sitemaps and known URLs.
Parsing and indexing — the page is stripped of markup and its words are stored in an inverted index, mapping each term to the pages containing it.
Query processing — the user's query is tokenised, spell-corrected, expanded with synonyms, and matched against the index.
Ranking — results are scored by relevance signals such as term frequency, link popularity, freshness, page quality and the user's location or language.
Serving results — the ranked list, often with knowledge panels, ads and vertical results, is returned to the user.

A simplified ranking formula historically used is tf–idf; modern engines use hundreds of signals, often combined with machine-learned ranking models.

Features

Keyword and natural-language search — engines increasingly accept questions rather than just keywords.
Instant answers — Knowledge panels, featured snippets, calculators and unit converters.
Vertical tabs — Images, Videos, News, Maps, Shopping, Books, Flights, and more.
Filters and operators — exact-phrase quotes, site:, filetype:, intitle:, date ranges.
Personalisation — location, language and (for some engines) search history.
Privacy modes — no-logging and no-tracking options (DuckDuckGo, Startpage, Brave).
APIs and syndication — programmatic access, often paid (Google Programmable Search, Bing Web Search API).

Market share

As of 2024, Google Search holds roughly 90% of worldwide search traffic, followed by Bing, Yahoo, Baidu, Yandex and DuckDuckGo.^[2] Share varies sharply by region: Baidu dominates mainland China, Yandex in Russia, Naver in South Korea and Seznam in the Czech Republic.

Privacy and regulation

Search engines log queries, IP addresses and click histories, raising concerns about profiling, filter bubbles and the right to be forgotten.^[4] Regulations such as the EU's GDPR and the Digital Markets Act impose logging and interoperability duties. Privacy-focused engines compete explicitly on not retaining query histories, and several regulators have ruled on search manipulation and self-preferencing.

Advantages and disadvantages

Advantages:

Immediate access to a large slice of the web's content.
Very low cost to the user; most engines are free.
Constantly refreshed index and increasingly useful direct answers.
Rich verticals (images, video, news, maps) from one query box.

Disadvantages:

Dominance by a few providers reduces competition and user choice.
Personalisation can create filter bubbles and reduce serendipity.
Privacy risks from query logging and tracking.
Low-quality and spam results can still rank well; SEO gaming is persistent.
Uneven coverage of the deep web and paywalled or non-English content.

History

1990 — Archie indexes FTP archives.^[3]
1993 — Wandex and Aliweb crawl the early Web.
1994 — WebCrawler introduces full-text web search; Lycos launches.
1995 — AltaVista brings full-text search to a mass audience.
1998 — Google launches with PageRank.^[3]
2000s — Google overtakes AltaVista and Yahoo; verticals (Images, News, Maps) mature.
2009 — Microsoft launches Bing.
2008–2011 — Privacy-focused DuckDuckGo gains traction.
2010s — Knowledge Graph, featured snippets and mobile-first indexing.
2021 — Brave Search launches an independent index.
2023 onwards — LLM-powered answers and AI search assistants reshape result pages.

References

↑ ^1.0 ^1.1 "Web search engine". Wikipedia. Retrieved 13 November 2026.
↑ ^2.0 ^2.1 "Search engine market share". StatCounter. Retrieved 13 November 2026.
↑ ^3.0 ^3.1 ^3.2 ^3.3 ^3.4 "Timeline of web search engines". Wikipedia. Retrieved 13 November 2026.
↑ "Search engine privacy". Wikipedia. Retrieved 13 November 2026.

[intro-1] 1.0 ^1.1 "Web search engine". Wikipedia. Retrieved 13 November 2026.

[statcounter-2] 2.0 ^2.1 "Search engine market share". StatCounter. Retrieved 13 November 2026.

[history-3] 3.0 ^3.1 ^3.2 ^3.3 ^3.4 "Timeline of web search engines". Wikipedia. Retrieved 13 November 2026.

[privacy-4] "Search engine privacy". Wikipedia. Retrieved 13 November 2026.

[1]

[2]

[3]

[4]