I’ve occassionally heard that SEO is the red headed step child of interactive marketing, which in and of itself is the red headed step child of marketing processes. Even so, I’ve always tried to stress the importance of search marketing to the end-user, not just as a means of building traffic..

However, when I see a community dedicated to interactive marketing processes fundamentally lose sight of the end-user while keeping tunnel vision on crawlers, I would tend to agree that maybe, just maybe – we are just a bunch of red headed step children.

This month, Matt Cutts made an announcement (don’t panic) that many people in the SEO industry have long suspected, that Google incorporates site speed and page load time into its search it’s scoring and search rankings.

In a blitzed whirlwind of activity, I would imagine that SEO’s around the world have been splicing and dicing their analytics, camping out in their web developer’s offices, and looking for solutions to common causes of slow site speed – such as inefficient code structure, un-optimized database tables, and the widespread use of CMS infrastructure.

In fact, most of the "SEO blogs" that I’ve been reading this past month have had multiple postings on this topic.

It is a strange fact indeed that the SEO industry is only now starting to grasp the importance of this factor, and unfortunately they are doing it for all of the wrong reasons. In the past, they couldn’t have been less concerned about optimizing page load time for the users of their websites, but now that they’ve confirmed Google Bot is running around their properties with a stopwatch – it’s a matter of the utmost urgency.

Continuing this ’bot centric’ theme, I was inspired to go back and review Google’s patent on Information retrieval based on historical data that was filed all the way back in 2003 and, coincidentally, mentions load time playing a part in scoring and ranking processes within their algorithm.

Here’s are a few other elements that are listed in relation to website metrics and, according to the patent, have a right to determine scoring and ranking for a particular website with Google:

I. Date of Domain Registry
Over the years Google has drawn many parallel qualities between doorway, or "spam" sites - one of these is registry information for the domain itself. The longer a domain is registered out for - the less likely it is to be a doorway site.

DNS records and physically correct addresses - and how both of these change over periods of time are also analyzed to determine legitimacy, which could result in a re-scoring of a particular website.

Nameservers, as well as their 'neighborhoods' are also analyzed by Google. If a website is hosted on an NS having distinct and high quality websites, it might be viewed more favorably than a website hosted on a NS consisting of high volume, low quality sites of similar "bottom feeder" context.
 

II. Link Behavior
The rate and time-varying behavior of link acquisition plays a big part in your link profile. Spam sites tend to attract hordes of low-quality links instantaneously through spam networks - and their scoring is a reflection of this. Conversely - a slow and upward trend of link acquisition is considered to be natural. Google actually goes on to say that ideal link acquisition rate would reflect a logarithmic curve:

Anchor text changes and link migrations during website rebuilds are also positive changes , since one would assume that "a good link would not go unchanged."

Link behavior can also be aggregated with domain-related information. For instance, if a domain registry was recently updated (or sold) - Google might devalue the links pointing towards that website - especially if the on-site text has little correlation with the anchor text of backlinks pointed to it. Which makes perfect sense. If you have a 10 year old website about horses, and I buy it from you and make a website about cows - why should the backlinks you've acquired relating to horses have any algorithmic properties in the scoring and ranking methods for the website in question?
 

III. Freshness of Content and Block-Style On-site Analysis
Google can discount certain on-site content and deem that content to be unimportant. This can include content that rarely changes - such as JavaScript, comments, advertisements, navigational elements, and boilerplate (content that appears everywhere).

Keywords within these elements would likely have very little impact within Google's '200 signals' that are used to determine the ranking/scoring of your website.
 

IV. Query Analysis
Prior rankings on both the domain and page level can have an impact on your scoring. If you're penetrating a new query space, Google might see your website as being "topically fresh" and award you higher rankings than if you've previously been in query spaces and have eventually fallen out of favor.

CTR (click-through-rate) also plays a part in determining your positioning within Google search engine results pages. If you rank number #3 in a given query space but have a higher CTR than the #1 occupant, Google might award your website as being more relevant within the given space and adjust your rankings accordingly.
 

V. Traffic Patterns and User Behavior
Google can analyze the frequency of browser (user) favorites and/or bookmarks relating to your website, and adjust your scoring and ranking accordingly. The more browsers that have you contained within their favorites and/or bookmarks - the more likely your website is to positively adjusted within Google's scoring methodologies.

Bounce rate, time spent on site, and page views are also available to Google for analysis, and as such - can all be comparatively analyzed against other web properties to determine search engine rankings.

It's important to remember that this patent was filed 7 years ago, and just how much of it was integrated into the actual algorithm is unknown. I will say that if you were evaluating Google patents based on the number of engineers involved, this one would indeed be the mother lode of all publicly available Google patents to date.

Another reason why I'm mentioning the age of this patent is to illustrate just how advanced Google is likely to be in their scoring methodologies. And if you believe that Google has stopped innovating in search quality, then that's a flawed assumption.