Although the text-box and search button is fairly common-place, the type of search—often described in terms of the scope of content the search engine has indexed—is not always evident.
a search that can only be used to find content on a single website (or intranet or extranet); for example the Motive search, at the top-right of each page, can only be used to find pages on the Motive website that contain the keyword entered
a search that can be used to find content on any website, anywhere on the web; for example Google (also see details below on search engine registration)
a search engine that uses the indexes of other search engines to find content, anywhere on the web; for example Dogpile
Only webpages that can be navigated to from the URL submitted to the search engine are indexed.
To index a website, a search engine must first be told where to ‘find it’. Notifying a search engine of a new website is referred to as search engine registration.
The registration process involves submitting an entry-level webpage address usually the address of the homepage or sitemap. Depending on the search engine, classification information may also be requested, such as a short description of the website, topics covered, and owner.
Most public search engines have an ‘Add URL’ or ‘Submit URL’ link (often in the homepage footer) providing information on how to register a website.
Once the website has been registered, the search engine owner runs an indexing program (spider). The indexing program follow hyperlinks from the URL(s) submitted to other webpages under the same domain, and in turn follows hyperlinks on those webpages, ‘crawling’ the entire website, to build a content index.
A search engine results page (SERP) lists webpages in order of their relevance to the query entered. The webpage listed at the top of the results has been determined (by the search engine) to be the most likely of the webpages indexed to provide the content the user is seeking.
Each search result listing usually features the destination webpage meta title (as the link text), followed by a description and/or an excerpt showing the query highlighted in the context of the webpage content (concordance).
Each search engine has its own method for determining relevance, usually based on an analysis of the content of the destination webpage, including:
Each of these aspects of the webpage is scored (primarily according to the quality of the match to the query) and then weighted. For example, a search engine may assign a greater weighting to meta title text than other aspects of the webpage. In this case, a webpage that includes the query in its meta title text may then be ranked higher than a webpage where the meta title does not include the query.
The calculation each search engine uses to rank webpage relevance is often a closely-guarded secret.
The scores for each aspect of the webpage are combined to determine the overall relevance of the webpage.
The calculation (algorithm) each search engine uses to rank webpage relevance is often a closely-guarded (and patented) secret. This is both to prevent websites from artificially inflating their rankings; and also because the quality of the search results translates directly into user-loyalty, traffic and revenue generating opportunities.
The Nielsen//NetRatings MegaView Search reporting service measures the search behavior of more than a million people worldwide. These web surfers have real-time meters on their computers which monitor the sites they visit. This metered information is compiled to produce NetRatings results. Below are statistics about searching from NetRatings provided to Search Engine Watch:
The chart below shows the percentage of online searches done by US home and work web surfers in July 2005 that were performed at a particular search engine. Internal site searches, such as those to find material within a particular web site, are not counted in these totals. The activity at more than 60 search sites makes up the total search volume upon which percentages are based -- 4.5 billion searches in this month.
Note that the figures are search-specific but probably not necessarily web-search specific. More about this is explained on the comScore search engine ratings page. Also note that some companies own more than one search site. This means searches at different sites may be combined into one overall figure for the company's entire network. The notes below provide more information of what's in each share. Remember, in all cases, only activity by those in the US is measured, even if those in the US go to a site run by a company outside the US, such as Google UK.
The chart below shows how the share of searches has changed over the past few months, for those search sites with a share of 5 percent or higher: