Page Ranking

Tutorial Directory

Are the First Results Always the Best Results?

When you send a query to a search engine do you assume that the first results listed will be the most relevant? Do you persist and look beyond the first few pages of returns? Market research indicates that 85% of casual search engine users don't go beyond the first page of results. This makes Webmasters and marketing experts very interested in placing their sites in the first 10 returns of a search engine. Understanding the mechanics of this relevancy placement will help you evaluate the quality of the results, and may make you more determined to dig beyond the top 10.

Every search engine strives to return relevant web pages that will satisfy your requests. Each search engine uses a proprietary 'ranking algorithm' that attempts to instantly build a list of highly appropriate responses to your query. Since each search engine applies its own formula to a unique database of information, results and relevancy rankings will always vary from search engine to search engine. (This is one reason serious searchers always consult multiple search engines.)

All search engines use the same basic ingredients to determine relevancy, it's just the mix that varies. The keywords of your query start the process. All ranking algorithms consider how often keywords appear in a document (frequency). They also measure keywords in relation to each other within a document (proximity). Another measure considers the location of keywords in a document. Keywords occurring at the beginning of a page, in the titles of pages, and in the URLs of the pages, are all given more 'weight' as relevancy is determined.

Another important element is the website's link popularity ranking. Many search engines, most notably Google, consider links to a Web page as a vote of confidence and popularity. The number of links to a resource then becomes a major factor in determining relevancy. It is assumed that a page that is linked to by other sites has been judged valuable and worthy. The greater the 'link popularity', the higher the potential relevancy of the site. Extra weight is assigned if the pages linking to the resource also have 'link popularity'.


cartoon: lady at a computer with question marks swirling about!

What is web page ranking?

Search engines display their results as an ordered list. The list is 'ranked' with the most relevant websites (as determined by the search engine's ranking algorithm) highest on the first page of returns. There are many factors involved in determining the rank order of search results. It is important for the serious searcher to understand the underlying dynamics of the process.

How do search engines determine the relevancy of a web page?

Each search engine determines the relevance of a page as it relates to a query by using a ranking algorithm. The ranking algorithm is a computerized formula designed to match highly relevant pages with a user's query. The exact nature of each search engine's ranking formula is a closely guarded secret. In general, search engines use a combination of factors that always include keyword frequency and page popularity. If a query is well formed, the results, while imperfect, often satisfy the searcher. (See the IMSA modules on Keywords, Operators, and Queries for more information on how to form effective queries.)

Do all search engines use the same formula when determining ranking?

Each search engine has a unique ranking algorithm that parses its database of web pages to determine relevant responses to your queries. Relevance is determined by weighing both keyword and website popularity factors. The same query done on different search engines will yield different rankings. Each search engine will weight the relevancy of a page according to its own special algorithm. Additionally, each search engine indexes information in a different way. This means that the relevancy rankings of each site will be unique.
graphic of a colorful checklist

How are keywords used to determine ranking?

The keywords of your query start the process. All ranking algorithms consider how often keywords appear in a document (frequency). They also measure keywords in relation to each other within a document (proximity). Another measure considers the location of keywords in a document. Keywords occurring at the beginning of a page are considered most important. Additionally, keywords that appear in the titles of pages, and in the URLs of the pages, are given more 'weight' as relevancy is determined.

What is Link Popularity?

Websites that are 'linked to' by other websites are said to have 'link popularity'. This concept is fundamental in the Google Pagerankª algorithm. Google considers each link to a web page a market driven vote for the website. The more links to a website, the more votes and the higher the Pagerankª that site achieves with Google. However just being linked to by a large number of other websites won't guarantee higher placement on Google. The popularity of the page displaying the link is also given weight. If the page linking to a website is also highly popular, the Pagerank of the linked to site improves even more. By contrast, creating links to other pages does not improve a site's ranking. Otherwise, it would be easy for any site to boost rankings by creating links to highly popular sites like

How does page popularity affect relevancy rankings?

Determining ranking and relevancy by page popularity is based on the concept that quality sites with good information draw more visitors. This is a market driven premise that assumes a site is popular because the information provided is of high quality. If a site has a great many 'visitors' and is relevant to your query, it is placed higher on the relevancy list than less popular sites.

What is click-through popularity?

One measure a search engine might include in its ranking algorithm is how often users actually visit or 'click-through' the ranked results of a query. The more often a site is chosen from the query results list, the higher its 'click-through popularity', and the higher it climbs in the rankings. Click-through popularity is a factor in the ranking algorithms of Ask Jeeves, Teoma and others. The time spent at the 'click through' site before returning to the ranked list is also measured. This measure of 'stickiness' can be factored into the ranking algorithm.

Does a site's presence in subject indexes affect its relevancy ranking?

Subject indexes like the Open Directory Project (deprecated) and LookSmart rely on human editors to review websites and categorize worthy sites into subject indexes. Search engine's ranking algorithms use subject index listings as evidence of resource quality. A site that has been reviewed and included in a subject index will be ranked higher than one that has not been listed in a subject index.

Is popularity ranking good for the researcher?

In most cases popularity ranking will improve your search results. Google pioneered this approach and won considerable market share because consumers noticed a significant improvement in the relevancy of returns. Most search engines now use some form of popularity ranking to determine relevancy. When popularity ranking is applied to millions of web pages you get a fairly unbiased measurement of a page's value. Exceptions are when you are researching unpopular opinions or seeking information from newly published resources on the web. A new website might have highly relevant and accurate information, but suffer in the popularity ranking simply because it is knew. Of course as the site ages and awareness of its qualities spreads we can assume the popularity of the site will grow.

How Do Paid Placement and Paid Inclusion Effect Search Result Rankings?

Don't assume that inclusion in the top 10 or 20 results is based on content merit alone. Be aware that sites ranked in the top 20 returns, while relevant, may have paid to be there. A dedicated researcher will look beyond the first pages and not be biased against results that are 'lower on the list' Intense competition for the top 20 spots associated with a keyword has created an income opportunity for search engines. Webmasters can buy their way into the top twenty by paying a placement premium.
Many search engines sell accelerated inclusion or placement. These Search engines include Google, Yahoo and Bing. Ad placements show up in the search results (as shown here) or may be marginalized to one side or the other. When ads are mixed in the results without a label, there is no way for the user to distinguish which sites have paid for high placement. Paid Inclusion is another common practice of popular search engines. Since it can take weeks or even months before a search engine includes a web site in its index, some webmasters are willing to pay to be quickly included in a search engine's database. Paid inclusion does not guarantee any precedence in the rankings; it just expedites the process of being indexed by the search engine.

What are Webmaster tricks? Do they affect page ranking?

Search engine ranking algorithms are constantly being revised to improve performance and screen out 'webmaster tricks' that attempt to unfairly skew page ranking. These tricks, sometimes called spamdexing, are attempts to fool search engines. Flooding a home page with keywords or loading the html meta-tag feature with keywords are examples of spamdexing.
Search engines have counter programmed their indexing systems to identify illicit attempts to 'spam' the relevancy rating of a page. In some cases offending websites are banned and removed from the search engine index altogether. Google no longer considers 'meta-tags' because this feature of HTML that provides descriptive keywords to search engine 'crawlers' is frequently abused.


Content authored by Dennis O'Connor 2003 | modified 2018