Web Usage Metrics

We all know that SEO is continuously evolving, mainly because search engines keep improving their own algorithms relentlessly aiming to give better results for every individual search query. It’s no news that we’re getting many more vertical results in serps for example, such as images, video clips, news headlines and local business listings (to mention some). What’s surprising to some extent is finding out that Google is already using web usage metrics to determine ranking positions especially for localization. I have strong reasons to believe that what was once one of the many SEO myths or considered to be a factor of future SEO, is today a reality and will be sharing my findings and insights here with you today.

Google has been collecting user behavior statistics for years now, through normal search of course and its own tools such as Analytics, Chrome and the famous Google Toolbar. These tools are free and definitely useful, but in exchange we’re indirectly supplying Google with invaluable data about us and our online behavior. Google sees our local IP address, what search queries we use, which result we click, whether we bounce back to the search engine to choose another result, whether we refine our query to search again, how long we stay on a particular website or page and much more.

Question: Are web usage metrics a part of Google’s ranking algorithm today?
Answer: Yes!

I do create a lot of websites on regular basis and that’s a fact! This gives me the opportunity to analyze the way search engines interact with new sites and identify procedures required for immediate successful rankings. So couple months ago I decided to create a particular website (mainly as an SEO experiment) and target Google US, details following:

• registered .com using a popular domain registrar in US
• bought a hosting package from same company
• optimized the website for en-US
• done basic link building, all from US websites
• and waited (yes, all SEO professionals have to be patient, hehe)

My new site got indexed pretty quick, a week or so on Google and that’s perfectly normal. Rankings are ‘usually never’ impressive immediately, unless I’m targeting some obscure keywords, but that’s normal too. I kept working on the website, adding content and what not and rankings started to improve gradually as anticipated. Important to remind you that being of Maltese nationality, I work mostly from Malta, unless I’m traveling. Although we do have Google.com.mt, many locals still use Google.com, especially in this case since I wanted to target the US market.

Whilst monitoring the website through Google Analytics I noticed that I’m still monopolizing the stats, since I did not exclude my own IPs and traffic is minimal. Looking at the data in point form I listed above, any SEO guru will confirm that my website should in theory rank best on Google US than any other localized version, but it’s not the case! Let’s say the keyword I’m trying rank for is ‘seo myths‘, I find my website:

• ranking #20 on Google.com (US)
• ranking #5 on Google.com.mt (Malta)

I’m obviously not checking rankings while logged in a Google account or from my own IPs, cause personalized search might be the reason. Website was optimized to rank in US, yet it’s ranking better on Google.com.mt. Is this a case of Google using user behaviour analysis to understand where a website is currently getting traffic from and boost rankings depending on the geolocation source of traffic? I definitely think so, give me one reason why my website is ranking better on Google.com.mt!

This is just one simple example, but being a new website where I had full control of its optimization I could exclude any other factor that could have effected rankings. Absolutely nothing points to Malta, apart from me searching for it on serps and working on it constantly, sometimes even from different IP locations in Malta. It’s very clear that Google is using already web usage metrics as part of its ranking algorithm and we should use this aspect to maximize our search engine optimization efforts.

In my opinion this also proves that:

• traffic generated to websites through offline advertising does effect rankings
• running PPC campaigns, does help Google ranking (not because we’re giving money to Google through Adwords, but simply because we’re attracting more traffic)

Search engines are always preaching that we need to make websites for the user, so attracting traffic alone is not enough if people just bounce away. Aren’t they saying indirectly that the bounce rate (a user behavior stat) is very important too? There you go! SEO’ing a website the traditional way is not enough any longer, we should replicate an offline business model to succeed online too, i.e. get the most clients in any creative way possible and provide the best user experience.

On seomoz.org they say that web usage metrics is a factor of the future of seo, but I honestly think it’s an important part of Google’s ranking algorithm already!