Measuring page speed – why and how
By Timmy Fredriksson
Most website visitors won't even think about page speed, not until a site is too slow or takes too long to load a page.
It's not just user experience that is impacted by page speed. Search engines also use this information when ranking content. This means that you have to consider its effect on your website's visitors, but also on those searching for you or content related to you. In this article, we would like to discuss why you should measure page speed and some tips on how to do it.
The capital market and fast loading times
Webranking by Comprend compares what corporate websites provide for their users with stakeholder expectations, be they analysts, investors, business journalists or jobseekers. Although Webranking's primary focus is content, page speed has been a part of our scoring criteria since 2015. We felt that it was important to make a note of it, since site speed was a topic repeatedly brought up by stakeholders – around 94% of the capital market claims that this is a highly important factor in their user experience.
What stakeholders see as a fast page is, essentially, a page that shows them the content they're looking as fast as possible i.e. perceived page speed or perceived performance. That is what the end user experiences, which we believe is the most relevant way to look at page speed.
To achieve a satisfactory perceived page speed, the backend and the fronted of a website need to work in unison. Content needs to be delivered fast and displayed fast, and this needs to be done in the right order. Luckily, there are several tools that can be used for free, which means that anyone responsible for a website can keep track of their website's performance. Page speed tools vary in what they can measure, with common elements including the speed of servers, content delivery, and response time.
Transparency, fairness and authenticity
For a tool to be relevant when measuring perceived page speed, there are a number of metrics you should take into consideration.
Of course, it should measure how fast servers are, i.e. how fast content distribution is. But to really get to the core of perceived performance, there is another set of metrics that can be used to increase the feeling of speed. For example, how long does it take before the page is ready for user interaction? How fast do the pictures load? While there are no shortcuts to a high performing and fast loading website, perceived performance can be optimised. For example, by styling the content the user can see on the screen when the page renders and prioritising visible images when initially loading the page.
A tool should also be transparent, fair and authentic. Transparent about what metrics are measured, so that a backlog can be populated with possible improvements. Fair in a way that is consistent when measuring: location, device or internet connection specifications shouldn't matter. Authentic, as in measuring in a way that reflects the target group's circumstances.
Lighthouse Audits vs. Page Speed Insights
If we wanted to be transparent, fair, and authentic to the 900+ corporate websites we have ranked this year as part of Webranking, we needed to explore our options. As part of our process, we evaluate our tools on a yearly basis.
Our choice was to switch from Page Speed Insights to Lighthouse Audits. It was the only tool that allowed us to rank the perceived page speed of all 900+ sites against the exact same metrics, at the same time, while also providing a full disclosure of the metrics used. It's also available by default in Chrome browsers.
For added authenticity, Lighthouse Audits lets the user mimic the world standard of Internet usage by focusing on mobile viewing on a 3G connection. As a website owner, knowing how to run Lighthouse Audits on your website can be a valuable way to get insight on what can be done to improve your page performance.