The practice of SEO – tweaking a site to make it rank higher in search results – has thus far evolved alongside developments in search. There has been some discussion, however, that search as we know it will change so much that it could make SEO obsolete.

The source of the evolution is usage data – information about the way people interact with a website – which could be used to determine a site’s ranking in search results. This change will happen because while SEO and search engines are interlinked, they don’t necessarily have the same goals. Rob has brought this up a few times, and Shoemoney wrote a great post on it as well, so I thought it would be worth discussing the idea here.

Search engines need to deliver results that are as relevant to their users as possible. Not only does this provide their reason to be, but it supports the main revenue source of search – targeted advertising. 

Google‘s system currently mainly uses links to rank sites. The more links there are to a site, especially from reputable sources, the more relevant your site must be. These links essentially act as votes of legitimacy.

Search Engine Optimisation has evolved alongside search engines, and by focusing on link building and a number of other tactics (use of meta tags, keyword optimised copy, html and server tricks), can ensure that a given site ranks highly in SERPs.

Arguably, then, a particularly effective link building campaign could make your site rank higher than another site that might offer more value to a user. The traditional Google system is therefore obviously not tamper proof, and arguably not the best way to determine relevancy – which is their main goal. 

The best way to figure out how useful a site is to a user is to go to the source. Bounce rates, search histories and click through rates can all tell you something about a site. If most users are spending only a second on the homepage, it’s likely that it’s not the most relevant result for the search term they entered. If someone clicks through from a result and looks through a number of pages on the site, it should arguably rank higher than the one a user didn’t stay on.

Google has developed a number of services and products that can provide the information they need to find an answer. Google AdWords, AdSense, Analytics, Google Reader and Tool Bar all provide Google with usage data about websites. The launch of the Analytics Benchmarking service a short while ago suggests that information about users’ time on site, page views, pages per visit and new visitor percentage can all be used to affect the way search results are ranked. While the benchmarking service is opt in, we can hardly think that Google could ignore information from these products.

Based on this, it stands to reason that SEO as we know it will have to change. If your site is ranked based on how users interact with it, you’ll have to build a better site to rank well. Optimising a site might have little to do with link building or Meta data, and as Rob pointed out at a recent presentation, everything to do with user value.
The end of spam? I can hardly believe that. Achieving results truly based on relevance can only make the Web a better place. SEO will become less important, mean something different to what it does currently, or die out completely.