What I am going to share with you is supposedly the secret sauce of Gurujada's super fast page speed scores. May be not all of it, but definitely enough to boost your website speed by a few hundred milliseconds if not more. And that could be the difference between winning or losing a few thousand bucks on your website revenue.
For starters check any website we built, for that matter, our own website using any of your favorite website speed test tools. You will find that all of them load in under Two seconds flat, from anywhere in the US. This is only to establish the fact that we have both understood what takes to attain these speeds, and have been successful in getting the results.
Now, every kid out there who has ever done SEO will tell you site speed matters. What most people don't tell you are:
Let's hit the road with why you should care about how fast your website really is. Most recently, studies have shown that your website speed has a direct correlation with your search rankings. From abandonment to reduced pageviews, lower time spent on page, to increased bounce rates, pagespeed plays an important role in making or breaking your marketing plans. See this infographic for case studies on how companies like Walmart, Mozilla, and AutoAnything have seen benefits of lowering their page load times.
If you are used to browsing on your mobile you would have already seen some of the initiatives from Google to make browsing websites faster and lighter with initiatives like AMP. The year 2016 witnessed for the first time a phenomenon of mobile traffic exceeding desktop traffic. Google itself has revamped its pagespeed analytics tools and taken out the test for mobile friendliness into a new tool.
All this points to two interpretations:
Most of us refer to the overall loading time of a website as the speed. However, there are various metrics developed by engineers over time that help define the speed of a website. Notice how these technical terms can easily be correlated with how a human would expect a website to load.
Time To First Byte: (TTFB)
When you open a website, the server gives you back a HTML page. Time To First Byte is the time taken for the first little chunk of website code to reach your browser. Ideally Time To First Byte should be under 0.5 sec.
The browser turns the HTML page into a visual display on your screen. Start render is time taken for the first visual part of a website to appear. Ideally Start Render time should be under 1 sec.
The browser has completed turning the visual part of the site. The user can see the full web page. Ideally Visually Complete time should be under 1.5 sec.
Most websites load some background scripts and they continue to load even after the visual part is loaded. Server tells us the website completed loading. Ideally Document Complete time should be under 2 sec.
Some of the background scripts may load more elements and they take some time. This metric is measured when all loading activity has ceased for 2 seconds. Ideally Fully Loaded time should be under 2.5 sec
Number of file requests
Websites are made of multiple files - CSS, JS, and images. Each file results in a request to the server and that impacts the speed. Number of file requests should ideally be less than 50.
These metrics are measured for the first time load and are compared with the results for subsequent loading. Meaning, when you open a website for a second time it is only natural that it takes lesser time than the first time around. See if that is the case.
My purpose is not to turn this into a technical discussion of website speed, but to show that speed is a function of multiple aspects of how your website is constructed. Knowing about these is only the beginning of your journey to the high speed zone of the web. Let's cover what you need to do to drive the results.
The HTTP1.1 (commonly known as http and seen as http:// in URLs) A fundamental overhaul of the web's fundamental protocol means you can get a speed that was possible never before. HTTP2 brings in a host of advantages over HTTP1.1 like servers being able to handle multiple data requests simultaneously, server push, prioritising which elements should load first etc making it an obvious choice for speed improvements.
Get rid of bloat
Most Content Management Systems and Websites builders dump a ton load of code no matter if the page really needs it. Take Wordpress for example. Wordpress websites use a number of plugins and not all of them are optimized for speed. By being selective about what plugins are being used and if they are all required for the job at hand, you could start on a stronger note.
First fold optimization
Much of the speed problem is set in the context of how long does the first fold (the visible part of a website without having to scroll down) take to load. Including too many elements or effects has an impact on the performance and scores. For instance, a feature from SumoMe loads an opt in form even before the first fold of the website is visible. On slightly slower connections, this works in a lousy way. The first fold load first, the website freezes for a moment making the user wonder as to what is wrong with your website and then the welcome mat (as it is called) loads. Such tactics may offer higher conversions for a while, but could also beat your user experience. So be careful as to what you choose to keep on your first fold.
There is a myth around HTTPS connections being on the slower side. See the responses from Google, Facebook and Twitter when asked about SSL (aka TSL) slowing the websites down.
Getting your website SSL certified does a ton of good to your website and does not come in the way of your efforts to speed up your website.
Hopefully this information helps you in having a meaningful conversation with your website developer and avoid tactics that are not going to help you anyway. May the speed be with you!