For a long time now Google have used load speed as an SEO ranking factor, but I thought it would be useful to delve a little into how they actually use load speed and what impact it can have on your website.
First Google use averages when discovering page speed. This is the same principle as web spidering and content discovery. There is a program that Google use called the Googlebot, but the Googlebot comes in a range of shapes and sizes.
Google use many different IP addresses, geographical locations and computer environments to analyse a website, that is one of the reasons why they are so good as a search engine – they store an absolute wealth of information about every website in their index.
This is also the reason why Google are able to quickly discover black hat websites – if a website blocks Google or redirects them to ‘SEO’ content Google would be foolish to trust that content because it is obviously for them and potentially very different from the content that a real user will see. So Google use loads of different ways to ‘see’ a website.
When it comes to load speed Google actually major not on their parsing of the website but on the information gathered from users – actually for users who enable the Google Toolbar. The Google Toolbar (when enabled) sends data to Google on a variety of user searching habits and website specifics. In the instance of load speed Google will collect data about the users environment such as type of operating system, browser, broadband connection speed, geographical location etc. They will also measure the speed at which the website loads and compare that across their network.
Load speed as a ranking factor
So how is that data then users across rankings, the answer is sparingly. Imagine for an example where you have website.com and 10,000 people a month use it. The website has a load speed in USA of 1000ms (mili-seconds) and in UK of 50ms. Someone makes a search for the keyphrase in the title. Is that website’s ranking affect? No! Now imagine that same website and a search for the keyphrase in the title of a third level page from a user based in USA, this can potentially significantly affect the ranking position, if you compare in to the equivalent position of a searcher in the UK.
The point is that whilst load speed certainly is a ranking factor it isn’t a predominant one: title tags, anchor text etc. The depth of information gathered is very interesting because it means that Google have the potential to really specify search results based on a huge amount of user criteria; what about rankings based on low speed broadband connections, or specific rankings based on whether you are using Windows or Mac?
What do you think, will rankings get more targeted based on the data stored for the user or is a democratic ranking best for the search results?