SEO

On-Page Factors

On-Page factors are related directly to the content and structure of the website. This normally consists of pages written in the HyperText Markup Language but also applies to other document formats that are indexed by search engines, for example Microsoft Word or PDF formats. On-page optimization involves modifying keyword frequency in the URL, Title, Headings, Hypertext Links and Body text. It may also involve reducing redundant HTML codes (aka cruft) produced by Web page authoring tools and restructuring the site to produce better linked and focussed page content.

Many search engines now discount the weight given to on-page factors because they give too much scope for abuse by SEO experts. In theory the visible parts of a web-page are less prone to manipulation as they have to make sense to readers. However doorway pages with redirections and clever use of style sheets enable different content to be served to search engines and end users.

Each page should target between two and four keywords directly related to the contents. If you feel the need for more keywords then consider splitting your content into separate pages. The Uniform Resource Locator (URL) should contain keywords, separated by hyphens without being too long, around 128 characters is probably a sensible upper limit for the entire URL. The Title tag should contain the keywords with no stop words but arranged to make sense.

This should be the first tag in the Head section of the page. There is evidence that search engines give more weight to factors higher up the page. The content should be properly structured with the use of Heading (H1, H2, H3 etc) tags containing relevant keywords. Search-engines will only index a limited amount of text in HTML tags and using too many keywords will dilute the focus. Don't spam any of these tags, this won't be effective and could result in a penalty.

Many website designers spend a lot of time creating Keyword and Description meta tags. Although these may be read by search engines, for example the description tag is used by Yahoo! to provide a short description of the site in the Search Engine Results Pages, they are not used for ranking pages.

Personally I don't bother with them as they bulk out pages for little real benefit. Both Google, Yahoo! and MSN Search will use the text they find on the page as a description so make sure your first header and sentence describe the contents. However some search engine watchers say that the new Microsoft search engine, currently in beta tests, puts some weight on meta-tags. There is also evidence to suggest that search engines give more prominence earlier in the page and some engines will only index a limited amount of body text so making the first paragraph punchy is a good idea.

Image alternate-text tags (ALT tags) are only indexed where the image is part of a hyperlink. However ALT tags are useful for non-graphical browsing and should be employed correctly.

Comments are not indexed. Use bold/strong/italic attributes where appropriate.

Write natural copy aimed at the end user and not search engines. Don't worry too much about keyword density for the contents but take the opportunity to include keywords combined in different phrases and orders and create anchor text to related internal pages. Keep the number of links to fewer than 50, and probably less and don't repeat identical outbound-links. Theme related pages should be at the same level in the site hierarchy and be linked through the site's menu structure and site map. At least one page at the same level should link back to the home page so that search engines that have traversed a deep-link can index the rest of the website.

For any other document format, e.g. PowerPoint, Adobe PDF etc make sure you at least have a descriptive document title. Try to avoid formats that search engines find hard to understand, even where a search engine can index a format it will carry less information than plain old HTML. Avoid using images to replace text, except occasionally in hyperlinks. Avoid formats such as flash, shockwave and sitemaps where there is no alternative text. Avoid HTML Frames which some search engines find hard to navigate, use Style Sheets (CSS) instead. Style Sheets should also be used to reduce the amount of formatting within documents. Keep pages to less than 100 kilobytes and preferably not much more than a screen full of text. Where Javascript or Flash menus are used include plain-text links at the bottom of the page. These will ensure all search engines index the rest of your website.

Other factors directly under the control of the website is the amount of content. Large websites generally rank better than small websites for a number of reason. Search engines also like fresh content and will spider this more frequently. A regularly updated news page, even a blog, can provide deep links to the rest of the website.

Off-Page Factors

Due to the ease with which on-page factors can be manipulated search engines now place more weight on so called off-page factors. Google made this form of ranking famous with its patented PageRank algorithm but researchers discussed using ideas such as link anchor text as far back as 1994. Off-page criteria are obtained from sources other than the website.

Probably the most important of these criteria are the quantity of inbound-links using anchor text containing your target keywords. These should come from pages covering similar topics, preferably from large, long established authority sites. In the case of Google the higher the PageRank the better. All other things being equal, a link from a PR6 site is worth around eight to ten times that of a PR5 site. Remember that Google uses all inbound-links in its ranking process, not just those shown by the 'backlinks' option of the Google toolbar. If the page has the luxury of many inbound-links then mixing target keywords to cover different queries is a good idea. Since Yahoo! acquired Overture and dumped the Google index it has published an index called WebRank. This is based on the number of incoming links to a site. It would seem likely that this is also a factor in their ranking process.

Some search engines also rank sites based on usage information. This is called Click Density and was pioneered by the DirectHit engine. The search engine monitors the results pages to see which links users actually follow. This is a kind of quality control although is affected by on-page factors such as the Description META tag and other summary information the search engine uses to describe the page.

PageRank

PageRank is the concept that enabled Google to leap-frog over other search engines during the late 1990s in terms of relevant search engine results pages and speed of search. PageRank holds US patent number 6,285,999 granted on the 4th September 2001 to Larry Page, one of Google's founders. Page's idea is based on established practice for scientific papers where the importance of a paper is based on the number of citations made to it by other papers. The Google algorithm interprets a link from one page to another as a vote. The Google algorithm and the configuration of web pages means that page rank can be fed back so that pages that have more votes are then deemed to be more important and the votes they cast (outbound-links) are subsequently given more weight.

Many web designers' efforts to optimize their sites for search engines, if they make any efforts at all, stop at acquiring inbound-links and exchanging links with other websites in the hope that this will boost their site's PageRank. This focus on PageRank is probably due in part to the Google toolbar that can be installed on Microsoft's Internet Explorer web browser. This shows a PageRank value for the current site on a scale of 0 to 10.

Getting a good site ranking is not quite as simple as getting a good score on the Google toolbar and decisions about outbound-links and site structure can have significant effects on PageRank.

Fortunately because of the patent the theory behind PageRank is known. The formula is relatively simple:

Where

∂ damping factor, given as 0.85
n total number of inbound links
PRⁿ page rank of inbound page n
Cⁿ number of outbound links from page n

So the PageRank (PR) for any page is equal to some constant value (0.15) plus the sum of all the page ranks from inbound links divided by the number of links (C) on each corresponding page multiplied by the damping factor (0.85).

Don't worry if that is too hard to visualize, we will look at some real examples below. One more thing before we do. As pages can, and usually do, feedback to other pages by way of hyperlinks the calculation has to be made iteratively. Every time Google visits a page it recalculates the PageRank based on the weights of the inbound links. Assuming no other changes it takes around 40 iterations for the figure to converge to a stable value although Google itself uses linear algebra to reduce the number of calculations. The damping factor (∂) is set at 0.85 for optimum performance.

SEO

Search engine optimization (SEO) is the process of improving the volume and quality of traffic to a web site from search engines via "natural" ("organic" or "algorithmic") search results for targeted keywords. Usually, the earlier a site is presented in the search results or the higher it "ranks", the more searchers will visit that site. SEO can also target different kinds of search, including image search, local search, and industry-specific vertical search engines.

As a marketing strategy for increasing a site's relevance, SEO considers how search algorithms work and what people search for. SEO efforts may involve a site's coding, presentation, and structure, as well as fixing problems that could prevent search engine indexing programs from fully spidering a site. Other, more noticeable efforts may include adding unique content to a site, ensuring that content is easily indexed by search engine robots, and making the site more appealing to users. Another class of techniques, known as black hat SEO or spamdexing, use methods such as link farms and keyword stuffing that tend to harm search engine user experience. Search engines look for sites that employ these techniques and may remove them from their indexes.

The initialism "SEO" can also refer to "search engine optimizers", a term adopted by an industry of consultants who carry out optimization projects on behalf of clients, and by employees who perform SEO services in-house. Search engine optimizers may offer SEO as a stand-alone service or as a part of a broader marketing campaign. Because effective SEO may require changes to the HTML source code of a site, SEO tactics may be incorporated into web site development and design. The term "search engine friendly" may be used to describe web site designs, menus, content management systems, URLs, and shopping carts that are easy to optimize.