After Google announced that they will count the page loading time of a website for search engine positioning of a website, after that still date, thousands of unique blog posts have been written on this topic only. We will discuss how much fast your website should be, what Google expects, why Google expects a website to load faster and yes, some myths and realities. These obsessive compulsive disorder to make your website loaded at .0001 seconds must be dissected well to give the webmasters a bit of mental peace.
Why and how much page loading speed Google expects
There are millions of sites need to be crawled daily, simply they are relevant for Google search results and if your site is slower for loading important parts, frankly Google bot will visit less. Google has the fastest crawler bot and server than any other search engine today has. Still it hangs because of undue loads. Keep in mind, Google do cache each and every page that is crawled.
Another reason is, if a website takes 30 second to load, visitor might close it before the page actually loads. Google’s algorithm simply can not understand why the visitors are closing the pages so soon (is it a spam site or not). So to equalize this factor to some extent and make their work easy, they implemented this factor too.
So, all you need to do is, create a faster loading page by means of creating a site with graphically lower burden, low http requests. This is it. Expected mean of loading time preferably be < 6 seconds (analyzed by WebPageTest as Google suggests). Do not yell on us for saying ” < 6 seconds”. People usually test only the homepage, not each and every individual pages, but yes, inner pages takes slightly more time to load. It is nice to have a webpage that loads within 3 seconds and very bad if it takes 20 seconds to load.
Analyze by using above mentioned tool to correct your problems. Use Yahoo’s Yslow Firefox plugin (Firebug is also required to be installed) and Google’s page speed tool to analyze the bottlenecks. If you can not afford a faster server, CDN etc, do not cry, read ahead.
A slow site is not a irrelevant site
It seemed that so far Google is trying to establish its positioning algorithm formulas to gauge its opinion of the relevance of a website and classify it based on that. Honestly, that site is slow does not necessarily imply that its content is irrelevant, and vice versa.
This is a test from hundreds
In Google Webmaster tool it is clearly written it is based on hundred website’s performance. This is a vague gauge to make you scared. If you know basics of statistics, you know very well, comparing with only hundred can prove or establish nothing and among these hundreds there are big websites like Google itself, Yahoo etc; who can afford several thousand Dollars everyday to make the page loading faster. Not to say, they use their own dedicated servers, CDNs etc.
Penalty of 1% of queries
In the article on this topic posted on the Google Webmaster Central Blog, Amit Singhal and Matt Cutts says that only 1% of queries are affected. Basically, this is not huge, but still. It also foreshadows a consideration significantly larger in the future. This also brings me to another question appears:
What are the sites those are penalized?
If Google says that this criterion has an impact of 1% of the requests the fact is that Google does not take into account this criterion for all websites. So, it would be more of a penalty on the sandbox or duplicate content (and less violent we guess), intended for websites on the death row.
As you (probably) or me do not sell products like large companies, to afford the cost of faster server (plus CDN for some luckier) we need to run Advertisements. These in turn increases the page loading speed. So, practically, after a certain point, you or me can do nothing. Personally, we like Yahoo’s guide for making a website faster. It is comprehensive, if you can deploy only 50% “rules” written there, your web site will load within 6 seconds (by WebPageTest tool, with others, like Pingdom, it would be <4 seconds) that is great.
Follow the Author of this article :