We’re Not There Yet: Images, Scripts and Redirects Are Slowing Down the Top Mobile Travel Sites


It was inevitable: travelers are utilizing their smartphones to research and make arrangements for their trips, as smartphone and data network penetration continues its inexorable climb toward establishing a truly “mobile-first” world.

This will increase as a shared experience among the world’s travelers, with a quarter of the globe’s online bookings for the $2.36 trillion dollar travel and hospitality industry expected to occur via mobile devices by 2019, according to projections by U.K-based market research firm Euromonitor. The United States is ahead of the curve for mobile-based bookings, with half expected to originate on smartphones by 2016.

It makes sense: travel is, by its very definition, already a mobile experience, and now you can research your trip and handle bookings from hotels to planes, trains and automobiles. All this can be done from a device you always have on your person.

But is the Performance of Mobile Travel Sites Going Anywhere?

And yet, while the industry is working to facilitate travelers’ desires to be able to handle details on their smartphones, the user experience is suffering due to poor mobile website performance practices.

We wanted to gauge the real-world experience of modern travelers, so utilizing some of the most popular smartphones, we tested in November 2015 the performance of the top 100 mobile travel sites as ranked by information technology company SimilarWeb, calculated by data aggregation and based on traffic and engagement levels.

Tested sites were accessed using actual mobile devices connected to the AT&T 4G/LTE data network. Each tested site was loaded three times on the iPhone 5 and 6 as well as the Samsung Galaxy S5 and S6. The data from the median of the three test runs was used to report and compare site performance for the following metrics for comparison:

  • Load Time (Fully Loaded)
  • Time to First Byte
  • Start Render
  • Requests (Fully Loaded)
  • Bytes In/Total Size in KB
  • Content Breakdown by MIME Type
  • Redirects

As we’ve discovered in testing leading ecommerce sites for our last annual State of the Union for Mobile Ecommerce Performance, the usual suspects were in play: page bloat, increasing complexity, image size and too many redirects. These all have an effect on latency and slow down page load times.

Key Findings from our Report

Compared with typical ecommerce sites, travel sites for mobile typically are more form-heavy, meaning the initial elements presented are fields for data entry, as opposed to rotating pictures of destinations.

Therefore, the page composition of these mobile travel sites will generally differ from the mobile retail sites we’ve previously tested, although there are plenty of opportunities to clean up the slow sites on an internal basis to increase the performance and improve the user experience.

IMPORTANT: These tests are not intended to be taken as performance reviews of the devices we tested on. Rather, they demonstrate how broadly performance can vary across mobile devices. Site owners need to be aware that this degree of variance exists among their users so that they can prioritize performance testing their pages across a range of devices and connection types.

Among the top 100 mobile travel websites:

  • On all four devices, median load times exceeded the four-second target: The median page took 6.7 seconds to load on the iPhone 6 and 5.5 seconds to load on the iPhone 5. On the Samsung Galaxy S5, the median load time was 5.7 second, with the S6 scoring the lowest median load time at 4.1 seconds.
  • While the median page had 62 requests (such as images, CSS, and JavaScript files), 18 percent of the mobile pages we tested contained 100+ resource requests. Each of these requests incurs latency, which adds up to slower load times.
  • As a percentage of the sites’ total size, images accounted for about a quarter of a page’s weight on average (22.8 percent). Ten percent of the sites had images making up over half their size, going up as high as 86 percent.
  • The mobile travel sites exceeded the ideal total payload of under one megabyte, averaging 1.2 MBs.
  • The median number of redirects was 8, with 43 percent of the sites including over 10 redirects.
  • In general, the slowest sites had more JavaScript requests than the fastest sites, taking a performance hit from third-party trackers. The average JavaScript payload was around 368k.

10 Slowest Sites (anonymized)

Looking at the slowest sites, they generally (though not in every case) had both more requests and larger footprints.

Another general area of woe? Third-party JavaScripts and trackers/analytics tools.

While it’s understandable that the industry would want as much customer data as possible in order to be responsive in its marketing efforts, data trackers increase a site’s weight and up the number of data fetches tasks, leading to slowdown.

Consider this section from the waterfall of one of the slowest sites:


Source: All waterfall charts from WebPagetest.org

The items in yellow are highlighted as “warnings.” They are related to ad networks and analytics with undesired redirects. The sum total of all the JavaScript elements (with 110 requests), is 60.2 percent of the mobile page’s weight.


(Note: "MIME types" are used to denote the type of information that a file contains, such as .html meaning an HTML page) Source: WebPagetest.org

Conclusion and Takeaways

For mobile travel and hospitality site owners, the performance hit taken by third-party trackers and analytics tools needs to be weighed against the benefit of the insights gained. Is it worth it?

The abandonment rate for the industry as a whole is around 41 percent across all websites, desktop and mobile included. Anything that can be done to smooth out the experience for users can help combat that.

We recommend you:

1. Consolidate JavaScript and CSS

Consolidating JavaScript code and CSS styles into common files that can be shared across multiple pages should be a common practice. This technique simplifies code maintenance and improves the efficiency of client-side caching. In JavaScript files, be sure that the same script isn’t downloaded multiple times for one page. Redundant script downloads are especially likely when large teams or multiple teams collaborate on page development.

2. Minify Code

Minification, which is usually applied to scripts and style sheets, eliminates non-essential characters such as spaces, newline characters, and comments. A correctly minified resource is used on the client without any special processing, and file-size reductions average about 20%. Script and style blocks within HTML pages can also be minified. There are many good libraries available to perform minification, often along with services to combine multiple files into one, which additionally reduces requests.

3. Defer Loading and Executing Non-Essential Scripts

Many script libraries aren’t needed until after a page has finished rendering. Downloading and parsing these scripts can safely be deferred until after the onload event. Defer as much as possible until after onload instead of needlessly holding up the initial rendering of the important visible content on the page.

The script to defer could be your own or, often more importantly, scripts from third parties. Poorly optimized scripts for advertisements, social media widgets, or analytics support can block a page from rendering, sometimes adding precious seconds to load times.

Following these and other best practices can help increase the user engagement for your site. Remember, even a basic-looking page of entry fields can be too slow for users.

You don’t want them asking, “Are we there yet?

Get the Report: 2015 State of the Union: Mobile Performance of the Top Travel Industry Sites

Kent Alstad

Kent is an entrepreneur, software architect, and technology innovator. Before taking his former role of VP Acceleration at Radware, Kent was CTO at Strangeloop Networks, where he was instrumental in authoring all of Strangeloop’s issued and pending patents. Prior to helping create Strangeloop, he served as CTO at IronPoint Technology. Kent also founded Eclipse Software, a Microsoft Certified Solution Provider, which he sold to Discovery Software in 2001. In more than 25 years of professional development experience, Kent has served as architect and lead developer for successful production solutions with The Active Network, ADP, Lucent, Microsoft, and NCS. ”Port View”, an application Kent architected for the Port of Vancouver, was honoured as ”Best Administrative System” at the 1996 Windows World Open Competition.

Contact Radware Sales

Our experts will answer your questions, assess your needs, and help you understand which products are best for your business.

Already a Customer?

We’re ready to help, whether you need support, additional services, or answers to your questions about our products and solutions.

Locations
Get Answers Now from KnowledgeBase
Get Free Online Product Training
Engage with Radware Technical Support
Join the Radware Customer Program

CyberPedia

An Online Encyclopedia Of Cyberattack and Cybersecurity Terms

CyberPedia
What is WAF?
What is DDoS?
Bot Detection
ARP Spoofing

Get Social

Connect with experts and join the conversation about Radware technologies.

Blog
Security Research Center