So ... The next important question ...

How to test your site to identify the major performance problems?

You already know that your page speed is literally crucial if you want to keep your impatient online customers engaged and satisfied. And this is why identifying your performance bottlenecks is so important to your success.

The good news: slicing off large chunks of page load time from slow loading websites is a relatively easy if the aforementioned bottlenecks can be successfully detected using standard website performance testing methods, and the speed optimization tactics are implemented in the right direction.

The bad news: the situation tends to become really complicated when those performance bottlenecks are unknown, too complex or simply too vast and diverse to be identified.

But, usually the situation is not that bad ... According to Steve Souders, the Google’s Head Performance Engineer:

80-90% of the end-user response time is spent on the front-end.

And I can tell you from my own experience: your effort will be worth the cold! An efficient web performance optimization will generate almost instant results in the form of higher conversion rates and sky-rocketing sales. In fact, the statistics are simply staggering. According to a famous O’Rielly report you'll get:

• +3% conversions per second with a page load time improvement from 15 seconds to 7 seconds
• +2% conversions per second with a page load time improvement from 7 seconds to 5 seconds
• +1% conversions per second with a page load time improvement from 4 seconds to 2 seconds

Now let's see the how ... The very first question ...

What tool to use?

There are dozens of reliable free and paid tools out there. Just a few free examples:

Google Page Speed

https://developers.google.com/speed/pagespeed/insights/

GTMetrix

https://gtmetrix.com/

Pingdom

https://www.pingdom.com/

WebpageTest

https://www.webpagetest.org/

Obviously, each tool has its own advantages and disadvantages, identifying different issues and testing different performance parameters in different ways, from different browsers and geographic locations.

In other words: no single tool identifies every single performance bottleneck. And this is why you should always use at least three different tools to detect the major performance problems and to identify the available improvement areas. Obviously, you'll get slightly different performance ratings and test results and that could be a bit confusing sometimes, but testing across multiple tools, analyzing different design elements and comparing multiple different test results will help you to perform a truly efficient performance optimization.

One last piece of advice: you need to consider where your main target audience is located! You'll have to measure your website speed from those geographical locations!

OK. Now let's see the really important question ...

What to test?

In order to accurately identify the performance bottlenecks you need to understand the most important performance indicators. Here they are:

• initial page speed; in many cases your visitors only seek some sort of primary information in form of pure web content that is free of any graphics or media elements; this is why you should always keep an eye on parameters such as time to first byte (TTFB) and performing Ping requests from website servers; they will accurately describe the initial page load times and page speed statistics; if they are too high, you'll have to perform a server-side and website core optimization at the back-end of your site
• full page load time; basically, it's the time needed to render the entire page onto the requesting browser (with every single media element, script, add-on, etc included); it's an indicator that can carry many valuable information in terms of page size and content type; you can leverage that information to efficiently cache your static content in different geographical locations (using content delivery networks)
• geographic performance; the biggest advantage of on online business is the ever-expanding global market reach; but, this reach is not always equally effective across the globe due to various factors that can limit the efficient client-server communication (government policies, bandwidth and technology limitations, etc); in plain English: you should perform speed and availability tests across disparate geographical locations to get a better overview on your actual world-wide business reach
• load tolerance; there is no such thing as constant web traffic; you'll always have fluctuations and hosting servers handling normal load at peak performance may lose page speed under excessive loads; if you want to find out how much traffic your website can handle before losing performance, you'll have to perform periodical stress tests, ramp tests and load tests; if you compare the results with your regular usage patterns, you'll be able to determine the point where your site needs a hardware upgrade to efficiently handle the peak loads
• web server CPU load; as I already said, a mediocre hosting service can kill your business; shared hosting providers that are running thousands of websites on the very same server usually will fail to deliver high performance even if your site is well-designed and well-optimized
• website database performance; to put it simply: it's a vital factor especially for websites that are handling and maintaining dynamic content pieces pulled from back-end databases; the most common problems: inaccuracies in data transmission and insufficient free disk space on servers; the potential data losses eventually will affect your website performance

OK. Now that you already know how to identify the most important performance bottlenecks, we can move forward to expand our website performance checklist with 7 more factors that can affect the website performance.