SEO Search Engine Optimization

Understanding Google's indexing algorithms has long been the domain of the search marketing professional.  Here we take a look at latent semantic indexing, local inter-connectivity and link analysis amongst others.

Latent Semantic Indexing (LSI):

Ranking search results based on local inter-connectivity:

Google's site and link analysis:

This is the analysis of websites, pages, links and paterns as patented by Google in 2005.  Google looks at the following paterns, amoungst others:

• The length of domain registration
• Domain ownership changes
• WHOIS data and physical address information
• C-Class IP information
• Keyword and non-keyword domains
• The discovery date of new domains/pages
• Document change frequency and the amount of change
• The number of linked internal documents
• Link anchor text
• Link discovery date
• Link changes and deletions
• External link growth patterns
• The authority of external links
• Link quality ratios
• The distribution of links
• The lifespan of links
• Link patterns (new vs old and old vs new)
• Anchor text variety

In a specific topic, link spikes can be measured against search volumes, if these spikes are not facilitated by search volume trends, Google can discount or even penalise websites.  If the linking pattern is out of the ordinary, Google will initiate other algorithms to establish whether the flood of links are spammy or natural.

Topic Sensitive PageRank (TSPR):

TSPR is based on the topical relationship betwen the documents and the query.  This relationship is determined by the Open Directory Project, user input and search history.  By matching the topic to the Open Directory Project listings, Topic Sensitive PageRank can provide more refined results.  TSPR determines relevancy based on topical communities rather than pure link power.

TSPR shows us that Google places high relevance to topicly related links, make sure that you focus on these sorts of links and get a link from the ODP, even if you have to buy one.

Topic sensitive Trust Rank:

It is imperative that you have a good understanding of Trust Rank to understand Topic Sensitive Trust Rank.

Using the same principles, Topic Sensitive Trust Rank relies on seed pages which help establish the topics of websites.  Your high quality directories like the Yahoo Directory and DMOZ can assist with this.

Google's supplimental index:

Google places pages and websites that it does not trust into its supplimental index.  While the regular index houses the results that we see in the SERPS.  Only if there are insufficient results for a particular query, willn Google call results from the supplimental index.  There is no way of telling if your pages are in the supplimental index, here are a view reasons that they might be:

• Too many low quality links
• Not enough links
• Too many low quality outbound links
• Duplicate content
• Too many pages for your pagerank (low PR, many pages)
• Your site is new and has a low link profile.

By default, Google places all new sites in the supplimental index, until they have aged and gained authority.  If your website is in the supplimental index, Google will not crawl all of your URL's.  Ensure that you do not have broken links and 404 pages, as this will definitely get you into the supplimental index.

Spam detection:

Quality websites always get loads of spam links, Google will compare inbound and outbound links to your pages, one authority link will offset about 50 spammy links.  If you link to bad neighbourhoods, Google will see you as being part to the spam network, so be careful with whom you link.

There are claims that Google has about 10 000 human reviewers that check search results for integrity and review thos websites that have been flagged by there algorithms.  These human reviewers are also taking behavioural aspects into account in algorithms.

Human quality raters:

Someone at Google published this document highlighting the Quality raters guidelines.  Google uses human quality raters to review search results,t his shows us that human reviewers are very real.  When writing copy for your pages make sure that you write naturally, if one of Google's quality raters flags your page, it will be very difficult to gain back your position.

Behavioural data and personalized searches:

Behavioural data analysis includes click through rates, time spent on a site, which pages were visited and so on.  Google monitors and analysis this data, in addition, it is likely that Google monitors users with a long history of specific searches.

Personalized search:

If you have a google account and use Google search on a regular basis, it adjusts search results based on the sites you visit most, search queries you use frequently, blogs you monitor with Google reader and other content that you find and or read using Google services.  These results are customized specifically for each of us and will be different from the standard, unadjusted search results.

Bearing this in mind, we can assume that once Google has this information about a pool of related users, they can cross compare to see if they have a similar site visiting patterns.  For example, if ten thousand users, interested in SEO all read our Search Marketing News feed with Google reader, then we can assume that that site must be an important authorative site.

Conclusion:

With the focus on behavioural data, now more than ever, we need to ensure that we build sites for humans, that is the point of Google's algorithms, to return sites that human searches will find relevant.

 Your Google algorithm comments are welcome...