Google recently announced Google Speed search ranking algorithm to be rolled out in July 2018, after which page load times on a mobile for the first time will be a ranking factor for Google.

People want to be able to find answers to their questions as fast as possible — studies show that people really care about the speed of a page.

Although speed has been used in ranking for some time, that signal was focused on desktop searches.

The “Speed Update,” as we’re calling it, will only affect pages that deliver the slowest experience to users and will only affect a small percentage of queries. It applies the same standard to all pages, regardless of the technology used to build the page. The intent of the search query is still a very strong signal, so a slow page may still rank highly if it has great, relevant content.

As mobile traffic accounts for about 45% for my blog, with this upcoming Google Speed update I decided to make the most, and most probably to boost overall search rankings for mobile search. A time ago I already implemented Google AMP, and now about 20% of search traffic coming to my blog are served by AMP.

One of the bottlenecks for a really fast website is the so-called third party external javascript files (Social Share, Google AdSense, Amazon OneClick and more).

in my journey on a faster blog, one of the first elements I've decided to eliminate from the blog social share buttons, counters and like boxes. Despite they look cool, they don't drive a lot traffic for my blog, but takes a lot of page resources. 

Let's make a test from tools.pingdom.com. I will test two pages - a frontpage and one of the most popular articles on blog.

Let's start with testing a frontpage:

Testing frontpage

Testing frontpage

Now, this seems already a nice result, but I would love to see could I push performance above 90 and decrease request count under 50

Testing article page with social sharing buttons enabled

Testing article page with social sharing buttons enabled

Depending on how you measure, I still got a lovely result with performance grade at A 91 - it seems awesome.  Now, there are two things I don't like from this test (actually 3)

  • Requests - 293 requests seems way too much
  • Page size 3.4MB - well, this is a hefty image loaded article, but still, I would love to see this under 2.5MB
  • Load time - 3.76 - despite seems the page is still 43% faster than other (tested) - I would love to keep this under 3 second

Now, I have a couple of external JavaScript libraries serving social share buttons, it's time to remove them and retest.

First, by removing any social share modules (in my case it was a social share counter module for Drupal)

Retesting frontpage

Retesting frontpage

Now, here are some improvements already - page size has decreased by 0.2MB, and 4 fewer requests

After performing a test for the popular article, there was no much difference, except I noticed another bottleneck caused of external javascript

Requests by domain

Requests by domain

Amazon JavaScript takes makes more than 170 requests or about 60% of total requests, just to display recommended amazon products and redirects to .co.uk or .de store users. That's too much of performance, and actually is the next thing I'm planning to eliminate once I will finish with social share codes.

I have Facebook pixel (which I'm not actually using and Facebook SKD served on all pages just to power Facebook like button, now it's time to get rid of them.

Test results after completely removing Facebook SDK and Tracking Pixel

Test results after completely removing Facebook SDK and Tracking Pixel

Now, this is some serious performance gains - page size is down from  1.8MB to 1.1MB (I just saved 0.7MB by simply removing social share javascript), Total requests have decreased from 66 to just 40. One thing strangely, but page load time has increased in fact (might be related to empty cache for CDN)

Artilcepagetestingafterremoving social share buttons

Article testing after removing social share buttons

Here are some performance improvements too, but as I already noticed above, I have problems with Amazon Associates third-party scripts.

After a brief re-thinking, I decided to remove Amazon's OneClick JavaScript too and was able to boost page loads under 2 seconds, and having just 110 requests.

Amazon OneClick? No, Thanks - Improve Page Load times by removing Third Party scripts