Search engine market share history, year by year

Google's share, year by year, depends on who was counting

Google's share of US search rose from about a third of searches in late 2004 to about two-thirds by 2009 on comScore's panel, and stayed between roughly 64 and 67% until comScore stopped publishing monthly rankings in early 2016. Measured as search queries rather than panel searches, the federal court in United States v. Google found Google already at 80% of US general search queries in 2009 and at 89.2% by 2020.

Worldwide, StatCounter Global Stats has never shown Google below 88.73% in any month since its series began in January 2009. Google fell under 90% in October 2024 for the first time since March 2015 and has sat either side of that line since. Microsoft's engine (MSN Search, then Live Search, then from June 2009 Bing) has never held more than about 11 to 12% of US queries by the court's measure.

Every figure on this page carries its method because there is no single true share for any year. comScore projected searches from a panel of people with metering software installed. StatCounter counts referrals (visits arriving at a website it tracks from a search engine's results page). Courts and regulators count queries, using data compelled from the companies themselves. A 2017 figure from one is not comparable with a 2013 figure from another.

This page is the history. For the current worldwide split, see the site's search engine market share guide, which carries the latest month. For current figures in individual countries, see the companion guide to search engine market share by country.

Who measured search share, and what each one counted

• comScore qSearch (United States, about 2004 to early 2016). A panel measure: comScore recruited internet users who installed its metering software and projected their behavior to the US population. It counted searches at five search entities only (Google Sites, Yahoo! Sites, Microsoft Sites, Ask Network and AOL), so a share was a share of those five, and it included searches made inside the Yahoo and Microsoft portals. From about January 2015 the releases were explicitly desktop-only.
• Nielsen and Experian Hitwise (United States). Two rival panels. Nielsen gave Google 66.1% for October 2009, a month it warned could not be compared with earlier ones because its method changed; Hitwise, sampling about 10 million users, ran higher, at 72.25% for the four weeks ending 2 January 2010.
• StatCounter Global Stats (worldwide and by country, January 2009 to the present). A tag-based measure built from the tracking code on its member sites. It counts referrals, not searches.
• NetMarketShare (Net Applications). A tagged-site network with country weighting. It stopped publishing after October 2020, according to a VentureBeat report.
• Pew Research Center. A survey of stated preference, not usage. Asked which engine they used most often, 47% of US search users said Google in 2004 and 83% in 2012.
• Courts and regulators. Query counts from the companies' own data, published at fixed points. The strongest numbers available, covered below.

United States, 2004–2016: the comScore years

comScore's monthly US search engine rankings were the figures the trade press quoted for a decade, and they are the answer to most searches for a "comScore search engine market share" figure in a given year. December releases are shown where found; December 2004 and 2005 releases were not located, so November figures stand in.

Three method changes sit inside that table, and each one moved the numbers.

• qSearch 2.0, 20 August 2007. comScore began counting partner searches (queries started on a partner site that hand off to the engine) and each results tab such as Images or News as a separate search. It said at the time that "Google benefits disproportionately from affiliate searches and multi-tab searches," and restated Google at 55.2% for July 2007 and 46.2% for July 2006. Part of the jump from 47.3% in December 2006 to 58.4% in December 2007 is therefore a change of ruler, not growth.
• Explicit core search, from July 2010 data. comScore added a view that excluded contextual searches, which it described as searches arising inside non-search browsing rather than from a deliberate query. The gap was material: for July 2010 Google was 65.8% explicit against 61.6% total core, and Yahoo 17.1% against 20.1%.
• Desktop only, from about January 2015. The releases were explicitly labeled desktop-only from then on, which matters because desktop is where Google is weakest.

Microsoft Sites passed Yahoo! Sites in December 2011, 15.1% to 14.5%. Yahoo jumped 1.6 points in December 2014. comScore never showed Google at 70%; its December peak was 67.3%, in 2013.

What happened to comScore's rankings after 2016

The last monthly US ranking found in this research is the February 2016 desktop release, dated 16 March 2016: Google 64.0%, Microsoft 21.4%, Yahoo 12.2%, Ask 1.6% and AOL 0.9%. No later release and no comScore statement explaining the stop were found; the European Commission's 2017 Shopping decision cites the same release, consistent with it being the last. Searches for "comScore search engine market share 2017" or "2018" therefore have no comScore answer. The honest substitutes are StatCounter's referral series below and the court's query figures.

United States, 2009–2026: StatCounter's referral series

StatCounter's US series answers the year-specific questions comScore cannot, including "US search engine market share 2017" and "search engine usage 2017". The figures below are annual averages, all platforms unless marked, downloaded from StatCounter Global Stats on 27 October 2026. StatCounter restates its history, so a figure printed in a 2015 news story may not match what it serves today.

* Bing for 2009 is a partial year: the Bing label begins in June 2009, and Windows Live (1.24%) and MSN (0.93%) appear as separate lines that year.

• The 2016–2017 step. Google's all-platform share rose from about 81 to 87%, while its desktop-only share barely moved (77 to 80%). Most of that step is the shift to mobile inside StatCounter's sample rather than a change on desktop.
• The Firefox–Yahoo period. Yahoo ran above Bing on StatCounter's US series from December 2014 to March 2015, at 10.02 to 10.21%, while Google fell to 78.58 to 78.79%.
• Bing on desktop. Bing's desktop-only share has risen from about 6% in 2009 to 17.18% in 2025, more than double its all-platform figure of 7.94%, which reflects Windows and the Edge browser rather than a wider preference.
• Lows and highs. Google's lowest US month is 76.68% (March 2009); January 2026 at 83.91% was its lowest since January 2016. Bing's highest US month is 10.73% (June 2013), with February 2026 reaching 10.48%.

StatCounter's US figures sit close to the court's (79.5 against 80% for 2009, 88.0 against 89.2 for 2020). That is useful context, not proof that either is right.

Worldwide, 2009–2026

The worldwide series is the one most often quoted and the one that moves least. Annual averages, all platforms, from StatCounter on 27 October 2026; Yandex combines StatCounter's separate "YANDEX" and "YANDEX RU" lines.

Three readings of this table are commonly wrong.

• "Google hit an all-time low in 2024." It did not. Google dropped to 89.34% in October 2024, the first month under 90% since March 2015, and has been under 90% in 14 of the 24 months from October 2024 to September 2026, in a range of 89.34 to 91.32%. The series low is October 2013, at 88.73%. The accurate description is "about 90%," not a collapse.
• "Baidu has about 1% of world search." StatCounter's member sites sit mostly outside China and Russia, so worldwide Baidu and Yandex rows say more about the sample than about either engine. StatCounter's own China series puts Baidu at 55.84% in 2009, 78.75% in 2015 and 54.64% in 2025. Yandex's last annual filing with the US Securities and Exchange Commission, using its own Yandex Radar tool, gave it 61.9% of Russian search in 2022 against Google's 37.0%; that is the company's own measurement.
• "The AI-powered Bing took share in 2023." Microsoft launched the new Bing with chat on 7 February 2023. No jump followed: worldwide Bing was 3.03% in December 2022 and 2.81% in February 2023, while Google rose to 93.37% that month. Bing's later rise, to 5.29% in September 2026 (its highest worldwide month), came gradually through 2024 to 2026.

A desktop-versus-mobile split for the same era survives in an unusual place. The European Commission's 2017 Shopping decision preserved NetMarketShare's global averages for February 2017: on desktop, Google 80.47%, Bing 7.15%, Baidu 5.59% and Yahoo 5.55%; on mobile, Google 94.87%, Yahoo 2.94% and Bing 0.89%. NetMarketShare weighted its sample by country, which is why its Baidu figure is far above StatCounter's.

Why comScore and StatCounter disagree

In December 2013, comScore put Google at 67.3% of US searches; StatCounter's US average for 2013 was 80.23%. Both were honestly produced. They count different things, and the gap comes from four structural differences.

• Panel versus tag. comScore projected from a sample of people. StatCounter counts traffic arriving at a sample of websites, roughly 2 million member sites in 2019–2020 by the UK Competition and Markets Authority's description, with no reweighting to correct for how representative they are.
• Searches versus referrals. StatCounter records a search only when a user clicks through to a tracked site. The CMA noted that referral shares will not count searches answered on the results page itself, such as instant answers, so an engine whose users click out more often scores higher.
• Portal-inclusive versus web-wide. comScore's base was five "Sites" entities, including searches inside Yahoo's and Microsoft's portals and, until 2010, contextual searches.
• Desktop-only. From 2015 comScore measured desktop only. The court found Google's 2020 share at 84% on desktop against 94.9% on mobile, so a desktop-only panel understates Google by design.

The clearest demonstration that the metric alone moves the answer is the CMA's 2020 market study, which set five measures side by side for the United Kingdom. On StatCounter page referrals, Google held between 89 and 93% from January 2009 to April 2020, with Bing between 1 and 8%. By number of searches, using data the CMA compelled from the engines, Google held a band of [90–100] percent in 2018–2019 and Bing 5% by the end of 2019; the CMA publishes the Google figure only as a redacted band. Measured by comScore page views, a different comScore product from the US qSearch series, Google's UK share was only 59 to 81% between June 2015 and February 2020. The CMA judged those page-view shares "inconsistent with the shares presented in all of the other measures" and concluded they "do not provide a reliable indication." A thirty-point swing from the choice of metric is why no figure here appears without its method.

The CMA's Appendix C sets out all five UK measures in one document.

What the Yahoo–Microsoft alliance did to "Yahoo share"

Before 2010 a Yahoo share was the share of an independent search engine with its own crawler and index. After 2010 it is the share of a brand whose results come from Bing. Any series that runs across that line without saying so is comparing two different things under one name.

• 3 June 2009. Bing is fully deployed worldwide, the date Microsoft gave in its launch release of 28 May 2009, replacing Live Search. Earlier versions, as the European Commission records, were called Live Search, Windows Live Search and MSN Search.
• 4 December 2009. Yahoo signs its Search and Advertising Services and Sales Agreement with Microsoft, under which, in the words of Yahoo's FY2009 annual report, "Microsoft will be Yahoo!'s exclusive platform technology provider for algorithmic and paid search services." The deal had been announced earlier in 2009.
• 18 February 2010. The US Department of Justice closes its review, finding the deal "not likely to substantially lessen competition in the United States."
• 23 February 2010. Implementation begins, market by market.
• April 2015. The agreement is amended to make it non-exclusive, letting Yahoo route some traffic elsewhere, with Bing keeping most of it, according to a PCWorld report of 16 April 2015.

The court describes the result plainly: by 2009 Yahoo "had stopped crawling the web and producing its own search results," and going forward its results would be delivered by Bing. By 2020, it found, Yahoo and DuckDuckGo both syndicated (licensed and re-served) their results from Bing.

When the question is whose index answered the query, add Yahoo to Bing. comScore did exactly that from late 2010 with a second line reporting which index supplied the organic results:

So in December 2013, the month comScore showed Microsoft Sites at 18.2%, Bing's index was answering 27.1% of US searches in its panel. The Firefox episode shows defaults at work from the other side: the court found that while Yahoo was Firefox's default from 2014 to 2017, Google's share of Firefox queries fell from 80–90% to 60–70% and Yahoo's rose from about 10 to 30%, reversing when Mozilla switched back. Those are shares within Firefox only.

The court and regulator figures

Court and regulator figures rest on data compelled from the companies rather than samples. They exist only at fixed points, appear years late, and are often released as bands.

United States v. Google (2024)

The memorandum opinion of the US District Court for the District of Columbia in United States v. Google LLC, No. 1:20-cv-03010-APM, found that by 2009 80% of all US general search queries, desktop and mobile, flowed through Google, and that the figure had risen to 89.2% by 2020. For 2020 the court found Google at 94.9% on mobile, with Bing at 5.5%, Yahoo at 2.2%, DuckDuckGo at 2.1% and all other rivals at 0.9%.

Two points about sourcing are often misreported. First, the opinion does not name comScore, StatCounter or any other public data vendor as the source of these shares; the exhibits it cites are an expert demonstrative and Google's internal documents, so describing the court's figures as "comScore data" is wrong. Second, the opinion states Bing's ceiling two ways: "Bing's market share has never risen above 12%" in the findings of fact (paragraph 25), and Bing "has never held a market share above 11%" in the conclusions of law. About 11 to 12% is the honest summary, and it rebuts the recurring claim that Bing once reached 20%, a figure found only in desktop-only, portal-inclusive comScore data. The judgment is on appeal and the shares describe 2009 and 2020.

The European Commission's Google decisions (2017 and 2018)

The Commission's 2017 Google Shopping decision found that, apart from the Czech Republic and Slovenia, Google's shares in every European Economic Area country had been above 85% since 2008, that Google had been above 70% in those two countries since 2011, and that Bing's share had never exceeded 10% in any EEA country since 2009. Those findings rest largely on StatCounter referral data, not query logs.

The same decision records Nielsen page-view data for 2010 putting Google at 81.3% in the United Kingdom and 85.3% in Germany, and AT Internet site-visit data for November 2014 putting it at 92.6 and 94.1% in the same two countries. The 2018 Android decision quotes the same AT Internet barometer slightly differently (Germany 94, Spain 96.7, UK 92.9%), so even a regulator's secondary figures drift between documents.

When this page was last checked

Every StatCounter figure on this page was downloaded from StatCounter Global Stats on 27 October 2026, as annual averages with September 2026 as the latest complete month. StatCounter's worldwide chart can be set to yearly granularity to re-check any row. The comScore figures come from comScore's releases of 2006 to 2016. The court figures describe 2009 and 2020, and the judgment is on appeal.

Historical StatCounter rows can be revised when StatCounter reclassifies traffic, and its brand labels have shifted over time, so a figure quoted from this page should carry both the source and the retrieval date.

Frequently asked questions

Sources

• US District Court for the District of Columbia — Memorandum Opinion, United States of America v. Google LLC, No. 1:20-cv-03010-APM (Redacted Public … (2024, PDF)
• European Commission — Commission Decision of 27.6.2017, Case AT.39740 Google Search (Shopping), Table 1 (recital 188) and footnote 114 (NetMarketSha… (2017, PDF)
• European Commission — Commission Decision of 18.7.2018, Case AT.40099 Google Android, recital 814(3) (2018, PDF)
• UK Competition and Markets Authority — Online platforms and digital advertising market study, Appendix C: Market outcomes, paras 13–15 and 41 (2020, PDF)
• Yandex N.V. (filed with US SEC) — Yandex N.V. Annual Report on Form 20-F for the fiscal year ended December 31, 2022, pp. 34 and 47 (2023, PDF)
• sec.gov/Archives/edgar/data/0001011006/000119312510043149/d10k.htm
• justice.gov/opa/pr/2010/February/10-at-163.html
• ir.comscore.com/node/14506
• comscore.com/Insights/Press-Releases/2014/1/comScore-Releases-December-2013-…
• comscore.com/Insights/Rankings/Comscore-Releases-February-2016-US-Desktop-Se…
• gs.statcounter.com/search-engine-market-share/all/worldwide
• news.microsoft.com/source/2009/05/28/microsofts-new-search-at-bing-com-helps…