Duration is necessary for all the elements of a web page to load and display in the browser, the loading time is a concept present in the mind of every Internet user, since the development of the use of the web. In web marketing, when we talk about load times, we commonly measure KPIs that you may be familiar with: Time to First Byte, the start of the display, Time to Interactive, Speed Index, etc. Everyone agrees that loading time is an important concept for the activity of a site. It is even, some will tell you, a capital concept in SEO, which Google uses as a positioning criterion. Truly?
LOADING TIME, AN SEO CRITERION?
Google has been communicating about the loading time for a long time. The communication of the web giant has indeed gradually integrated the notion of loading time as a web and particularly mobile uses have developed. There are Google press releases from over ten years ago on the issue of loading time and the importance of a smooth browsing experience. Cupertino experts quickly understood that if a domain was more fluid to load, this would greatly facilitate the work of its bots and therefore reduce the machine times (and therefore the costs) necessary for the crawl of a site.
WEIGHT OF PAGES AND MOBILE
However, Google has not pushed developers and advertisers to streamline loading times for its own benefit. A long time ago, the oracles of Google predicted two phenomena to come: an increase in the average weight of web pages on the one hand, and the explosion of web consumption via a mobile device on the other hand.
On the weight of the pages, it is clear that in ten years, it has increased considerably.
Increase in the average weight of desktop and mobile pages from November 2010 to September 2020 (source https://httparchive.org/ )
On mobile use, Google had anticipated the explosion in the use of smartphones, with a symbolic milestone reached in 2018, when the world now had more mobile users than Internet users. With increasingly heavy pages, coupled with the increase in the number of mobile users, further slowed down by voluminous resources, we would run the risk of seeing mobile browsing remain significantly degraded compared to that on Desktop.
This explains Google’s policy on loading times. From simple recommendations in the early 2010s, through the threat of losing positions in the SERP to the Mobile-First Index where the Mobile version of a website becomes the benchmark version for bots, Google has long maintained the idea that a slow site would not perform well in SEO.
However, this is not exactly the case, since if we test on any competitive query, the first results are often far from the ideal of mobile performance.
ONE CRITERION AMONG MANY OTHERS
Indeed, the sites that perform in SEO are far from being fighter planes:
Above: The performance of the first three results in the SERPs on the query “Men’s summer shirt” (source: Dareboost and Google Lighthouse )
Through the few KPIs presented above, we see that the differences between the different pages of First Contentful Paint First Byte times (TTFB), the start of the display, and Time to interactive, are marginal: the pages, according to the criteria of Google, are slow to load. And yet the sites in the example are E-Commerce sites very well positioned and very present in the SERPs, on a multitude of keywords. So, are Google’s speed measurement criteria too demanding? Without a doubt. Because what we see is that the loading speed of a page is not a major factor in defining its position on a keyword. The positioning of a page depends on more traditional SEO factors.
LOADING TIMES: A UX PROBLEM?
Improving load times is often pushed into SEO optimization strategies. However, the sometimes heavy and costly efforts of the development teams to improve the fluidity of loading, are not rewarded by increases in positions. This is due, on the one hand, to the fact that despite its slowness, Google will still manage to browse your page, and on the other hand that the positions are not the right KPI to look at. Because most of the time, the loading speed remains a problem of user experience. And it is by looking at the behavior of Internet users on our site that we can see if the latter really has slowness problems.
TEST THE COMPETITION
To find out if your site requires effort in terms of loading time, you can test your market competitors and your SEO competitors. It’s a safe bet that they too are severely judged by performance testing tools. Of course, compare what is comparable: home page versus home page, product page versus a product page. You will see if any important differences appear. On the contrary, are their sites ultra-efficient? So, loading time may be a criterion, not for SEO, but for your Internet users. Otherwise, the loading time is very likely not to affect your positioning.
THE FIRST KPI, THE BOUNCE RATE
A simple method to check if the load time is impacting the user experience on mobile is to measure the bounce rate of a page compared to the average load time and follow the evolution over the months. By doing this exercise, you will be able to see if there is a correlation between loading time and user impact.
Above we see that fluctuations in load times are not necessarily followed by changes in bounce rate. The loading time is therefore not the element that detracts from the user experience.
In fact, if an increase in loading time is followed by an increase in bounce rates, then your loading speed becomes problematic because it directly leads to a rejection of mobile users, known to be put off by the slightest too long wait.
A SECOND KPI, CONVERSIONS
The conversion volumes and the conversion rate of a page in relation to inbound traffic is the second key point to check. Indeed, if we do the same exercise, namely comparing load times and this time conversions, we can shed light on the positive impact of a few seconds less loading on conversions.
Above, the comparison between loading time and conversion rate. In this example, when the load time drops by one second, the conversion rate goes from 65% to 83%.
With the conversion rate, we can therefore measure the real impact of the loading time on your objectives, and if linked to business objectives, the net profit that a few seconds less can bring.
LOADING TIME, A CONCEPT ABOVE ALL UX BUT WITH A REAL SEO ISSUE
Natural referencing is unfortunately not only measured by the good positions that we manage to consolidate over the course of our efforts. It is also necessary that our users click on our links in the SERPs and interact with our site to take all its value. This raises the question of measuring conversion and the real issue it imposes. Being able to demonstrate that SEO is profitable is indeed an important aspect of a SERP position optimization strategy. If too long loading times increase the bounce rate and push an Internet user captured in the SERPs to leave the page without interacting with the Call to Actions, then the loading time becomes an SEO problem.
SHOW YOUR USERS
If your site is having load-time problems that impact the user experience, then you must weigh the elements according to what is important to display for the Internet user, depending on the objective of the template. These are varied: Are you aiming for an interaction with a Call-to-Action? Do you want your content to be viewed? That a form is completed? So what should the page show as a priority to meet this objective? A content? A price offer? A solution? An answer to a question? Prioritize loading the determining elements, so that the first thing that will help you achieve your objective is displayed. These must be upstream of the items to be loaded, but also up above the waterline.
Likewise, the visuals must not negatively impact the rendering of the page. Prefer the most recent image formats like WebP or AVIF format. Finally, if these elements require action, the time between the moment when the element is called and when we can interact with it must be as short as possible. By prioritizing load time optimizations on things that matter to your users, you’ll see the difference in behavior results.
LOOK AT THE RIGHT KPIS
Performance KPIs are good indicators to highlight technical issues, and each provides valuable clues as to which links in the chain are causing difficulty. If we had to retain a single index to take into account in SEO, it would be the Time To First Byte, or in other words, the download time of the DOM. In the event that your site is slow to respond, you will inevitably have an SEO impact as Google will have difficulty accessing and crawling your pages.
However, the drop in the bounce rate, the increase in the conversion rate, the increase in clicks on CTAs, the lengthening of the reading time for editorial content… Here are the KPIs that really measure the added value of an effort to improve the user experience by optimizing loading times. In your performance audits, always keep in mind that performance for performance is useless if it does not allow you to achieve your objectives, and that what matters above all is the way your Internet users look. will relate to your site.
GOOGLE, ENCORE.
This very significant notion of user comfort is also once again supported by Google’s policy and its Core Web Vitals project, announced in May this year. The idea is to provide precise indicators that would measure the relationship between user behavior and fluidity of loading. To summarize, this project is based on three factors: loading time, interactivity, and visual stability of the content of the page. The objective, and the real benefit for us, is here to provide concrete KPIs on real positioning factors concerning the relationships between user experience, mobile ergonomics, and loading times. These three elements combined represent the heart of Core Web Vitals and promise to be essential for the SEO strategies of tomorrow.
