How to Run a Website Speed Test?

Speed is a relevant concern for any website owner and information on the topic may diverge greatly depending on sources. Loading time, server response time, hardware efficiency… Several elements need to be measured when carrying out a website speed test and there exists a certain level of confusion as to the why and how of these assessment procedures.

Why Run a Website Speed Test?

The impact of poor website performance may be felt on several levels. From user experience to return on investment, speed is at the heart of several key elements in a website’s overall success and should not be underestimated.

The Influence of Performance on Business Statistics

A website’s performance is directly correlated to its profitability. Speed is indeed cited as one of the main factors which affect bounce rate, conversion rate, shopping cart behaviour and advertising revenue. From a search engine optimisation perspective, slow loading time can lead to penalisation (as was the case for 1% of websites in 2010) and has an immediate influence on rankings (as higher-ranked sites typically load faster than the competition). From a user experience standpoint, a faster website is practically a requirement as 67% of users demand a loading time of 4 seconds or less. Speed also conditions loyalty and reputation, since users often share their experience on social media. High performance is generally expected regardless of the platform (desktop browser or mobile). Running a website speed test is therefore an important step towards addressing these numerous issues.

Stand Out from the Crowd

In a context where a website’s performance directly impacts user experience, behaviour and, as a result, the business’ return on investment, having a website with a fast response time constitutes an undeniable advantage over the competition. A good rule of thumb is to observe how your immediate competitors perform and to strive to do even better. Having a clear idea of your website’s performance over time is the only way to ensure consistent, reliable progress.

Loading Time is Subject to Change

There is no real loading time consistency for any website, as many factors may influence performance. Which is why running a website speed test, or even individual page speed tests, while not irrelevant, does not provide an accurate account of the situation. Performance may indeed be impacted by such alterations as a new functionality being added to a website, the adjunction of more content, new partnerships being implemented, etc.

To avoid any setbacks, carrying out frequent speed tests is therefore paramount and several metrics need to be considered in order to provide a reliable performance assessment.

Why is Response Time Not Sufficient to Assess a Website’s Performance?

TTFB (Time to First Byte), or the time elapsed between the moment a user makes an HTTP request and the moment the first byte from the page displays on their browser, is often misconstrued as the only performance indicator. As a result, if the server your hosting service relies on is quite powerful, you may assume that there is no reason to worry about page performance. However, this is unfortunately not true.

Performance-Impacting Factors

A website’s performance as it is experienced by Internet users depends on a variety of factors.

How your content is organised and classified within your website’s architecture, for instance, accounts for part of its behaviour. File types and sizes also matter a great deal: the larger and the more numerous the files are on a given page, the longer it will take to load. Compressing or opting for optimal formats is therefore very important.

The coding languages, their balance and their use across the website should also be considered in a website speed test. HTML, JavaScript, CSS, etc. all have their own characteristics.

Plugins – which are particularly valuable for CMS-based websites – have a “weight” of their own and can significantly slow down a website if they aren’t optimised or if there are too many.

Connection speed impacts performance dramatically, which is why carrying out a page speed test within the confines of a single case scenario is not sufficient to get a full picture of the situation. A user relying a dial-up connection or who accesses your website through a 2G mobile device will experience a slower response time than someone using DSL. A cable connection will provide even faster results, and a fibre-optic connection combined with an ethernet cable remains unbeatable.

The choice of browser, especially when not using the latest version also impacts page loading speeds, as they may not be caching elements as effectively.

Hardware makes a huge difference, which many website speed tests fail to take into account. RAM issues, in particular, tend to affect how quickly the cache makes items available to reduce load time and prevent it from playing its role.

When measuring a website’s performance, traffic context needs to be considered. Indeed, if the bulk of your traffic originates in the UK, carrying out page speed tests from the US won’t paint a reliable picture.

Finally, traffic volume, especially if it suddenly exceeds the amount of bandwidth set by your host, may lead to a more sluggish website or even a temporary shutdown. Which is all the more frustrating, as it proves your website is attracting more and more customers.

The many factors which may influence website performance not only illustrate how important it is to run a website speed test on a regular basis, but show that several criteria should be assessed in order to guarantee reliable readings.

Performance Metrics

It is necessary to acknowledge that no single performance metric is capable of providing an overall insight as to a website’s speed. While Speed Index does stand out as a particularly relevant option, each performance indicator has a well-defined role and should be utilised if you wish to truly optimise the behaviour of your pages.

Why Should You Use a Dedicated Tool to Test Your Website Speed?

Testing your page’s responsiveness straight from a web browser is theoretically possible (using Chrome Dev Tools, for instance). However, it is more than likely that the results won’t be reliable and, as such, not exploitable.

For starters, as stated above, you may not be in the same geographical location as your users and your own hardware probably won’t reflect theirs (as web developers typically work with more powerful machines than average users do). In addition to these elements, matching the quality of their Internet connections is impossible, both because they vary greatly between users and because your own will probably fluctuate even over the course of the website speed test.

Relying on a dedicated tool provides an effective way to address all these issues by putting several functionalities at the disposal of web developers, such as:

• Customisable context – to emulate various circumstances users may experience
• Stable, isolated environment – to guarantee reliable measurements
• Transparent information regarding performance indicators, all of which are relevant – including Speed Index, TTFB and many more. No vague “load time” obtained through unclear methods.

Dareboost provides reliable website speed test solutions and gives you 5 reports a month with the free version so you can stay on top of your website’s performance.

The pro version of Dareboost features monitoring functionalities thanks to which regular automated website speed tests are carried out to alert you any time your pages are not performing as well as they should.