By Sam Haakman
At Fastrack, we strive to ensure that our websites present the best user experience possible, guiding visitors to book directly with minimum friction. User experience is affected by many factors, most of which are highly subjective, such as text appearance and placement, use of colour and overall layout. However, there is one major factor of user experience which is objective and measurable - page load speed. Various studies over the years have highlighted that users will leave a site if it takes too long to load. In order to expose this data in an actionable way, Google has recently released an update to their TestMySite tool.
TestMySite is a tool that represents the changing ecosystem of the web. Speed testing tools have been around for a long time, but they mostly focused on the test case of desktops with high-speed connections, the ‘best case scenario’. However, we know Google is putting more and more weight on the mobile experience of sites, even starting to base desktop site rankings in search results on the mobile experience of a website, beginning to roll out their ‘Mobile First Index’ in October 2017.
Many sites will report poor results in the TestMySite tool, even when testing very well in traditional speed tests, including Google’s PageSpeed Insights. This is because TestMySite simulates loading websites with a low-powered device on a regional 3G connection - the ‘worst-case scenario’. Ideally, a website can perform well on both desktop and mobile, however, optimising websites for slow mobile connections will yield results for both platforms.
Although statistics show that around 70% of mobile devices globally will be connecting to the internet using these slow networks until at least 2020, it is always important to consider your specific demographic. For the hotel industry, this is not the case. Google's statement of 70% is taking into account the increasing number of devices in rural areas and developing nations, most of whom are not among the target market for accommodation products.
We don't have any water-tight data on this, but what we do have shows that amongst our clients about 85% of mobile visitors experienced load times much better than TestMySite's reported load time for the site. Whilst people visiting your accommodation may not be on the latest, fastest phone with 4G coverage everywhere, they will have better average connections than the test case presented by TestMySite. Regardless, this is a Google tool for measuring user experience on mobile devices at a time where mobile experience is becoming a very important search ranking factor, so it shouldn't be dismissed.
Sites that were built with best performance practices in mind even 6-12 months ago struggle in these tests as the sheer weight of functionality that a modern website requires will fail to load quickly on a slow 3G connection. When seeing poor performance in a speed test like this, many will want to blame the server technology, but in this case, the limiting factor is the connection speed being simulated - upgrading your server will likely make little to no difference to performance in these mobile tests.
In order to perform well in these tests, the only solution is to transfer less data. Doing this without compromising on the features of a site can be very difficult. In order to achieve this, most sites will need to restructure the way scripts and images are loaded so that browsers will be able to display critical content in a timely manner, ‘deferring’ non-critical scripts and images to load after the page is loaded. Other techniques, such as serving lower resolution images to small screens, or even simplifying functionality on mobile can also prove valuable.
At Fastrack, we have a commitment to user experience and maximising the value of our clients' websites. We will be strongly recommending all our clients invest in mobile website speed optimisation in order to reduce their visitor loss and improve bounce rate moving forward.