How Google Is Using Website Speed To Impact Search Results
I have always been impressed with the history of Google. In 1996, two students at Stanford created their own search engine called BackRub. By 1998, the search engine had been renamed Google and was able to return search results more relevant than any other engine.
Instead of relying only on in-page metrics, Google used “a better system that analyzed the relationships between websites.”
This revolutionary idea essentially catapulted the company into what it is today. Showing the most relevant search results is still incredibly important to Google. They update their search algorithm almost daily in a quest for perfection.
The importance of speed
In 2010, it was announced that Google would begin to use site speed as a piece of the search algorithm. In the blog post that revealed this news, Google discussed how speedier websites keep users happier than slower websites. This idea is easy to accept and is a public evidence of Google placing an importance on user experience for search results.
Google’s search algorithm is complicated and secret. The world only receives bits and pieces of what the algorithm is made up of. But, we know speed is one piece of the pie.
How does Google define speed?
Site speed can be defined in quite a few ways: the time it takes to load the entire page being requested, the time it takes to load the aspects of the page that are immediately viewable, and the time it takes for the first byte of the requested page to be received.
Which ones are used by Google? This is a tough question to answer. But, search engine experts have performed correlation experiments to decipher what is going on and have posted about their results.
In the publicly posted experiment, two thousand search terms were selected. Some of the terms were one word queries, others were four to five words long. The top fifty results were selected from each search term, resulting in 100,000 web pages (2,000 x 50 = 100,000). Site speed measurements were kept of all 100,000 pages.
The only identifiable correlation between site speed and search results was found to be time to first byte.
There was no correlation found between time to fully load the page or time to partially load the page.
If you are looking to optimize site speed in order to impact search results, look to first focus on the time to first byte. You can improve this metric by decreasing the amount of processing that is required to deliver the first bit. I suggest looking into caching your pages, which makes them static. This minimizes the amount of back end processing required before delivering the first byte.
Matthew Schmoldt is an SEO expert who has been featured at Moz. He is an Internet entrepreneur and is the founder of Cool Things Inc.