A new report by Radware shows that 9% of the top 100 leading retail web pages took ten or more seconds to become interactive, which is down considerably from 22% of sites last quarter.

Studies have shown that online shoppers will abandon a web page after waiting just three seconds to load. Although this improvement is promising, 10 second load times are still far off of the three second target that most users expect.

Today’s web users are now likely to use mobile devices to browse dynamic pages and expect a high degree of responsiveness. Unfortunately, congested networks and unoptimized web pages lead to a frustrating lag time for users and many sites face an overall increase – not a decrease – in sub-optimal user experiences.

Radware’s report entitled State of the Union: Ecommerce Page Speed & Web Performance, Spring 2015 also found that only 14% of the top 100 retail sites rendered feature content within the acceptable threshold.

“There is no doubt that web pages have been increasing in complexity as well in payload size. Although this trend is focused on enhancing the user experience, it can unfortunately correlate to slower load times if a page is not properly optimized,” says Kent Alstad, VP of Acceleration for Radware. “Our latest report has found that the median page size is 1354 KB in size. As images comprise over 50% of the average page’s total weight, almost half of the top 100 sites have failed to implement core optimization techniques such as image compression. This alone, can help deliver pages quicker to the viewer.”

Radware also lists the fastest ecommerce sites which were the quickest to display actionable content. From a user experience perspective, time to interact (TTI) is a more meaningful performance metric than load time, as it indicates when a page begins to be usable. Among those listed is a well-known Internet based retailer which took 16.3 seconds to load, but boasted a TTI of 1.4 seconds.

“When we discovered what made sites load fast, we found that the median page was 932 KB in size and actually deferred resources that were not part of the page’s critical rendering path. These non-essential resources were mainly ‘invisible’ such as third-party scripts that aren’t needed until a page completes its rendering. Deferral is a fundamental performance technique and should be employed to optimize the critical rendering path of websites,” added Alstad.

Other findings in the Spring 2015 report include:

■ Despite the fact that images comprise 50-60% of the average page’s total weight, 43% of the top 100 sites failed to implement image compression, a core optimization technique.

■ Page complexity, which is a greater performance challenge than page size, has grown by 26% in the past two years. The more complex a page, the greater risk for page failure.

■ Among the top 100 pages, the median time for pages to interact is 5.2 seconds. Although down from the previous quarter of 6.5 seconds, this is considerably slower than users’ wait-time threshold of 3 seconds.

Also outlined in the report is the “performance comeback” of two large eRetailers that show significant changes in their TTI compared to Radware’s report of Fall 2014. Time to interact for the online retailers were 2.4 and 2.9 seconds down from 5.2 and 7.2 seconds respectively, demonstrating the value of implementing optimization techniques to decrease load times of web pages.

Methodology: The tests in this study were conducted using an online tool called WebPagetest – an open-source project primarily developed and supported by Google – which simulates page load times from a real user’s perspective using real browsers.
Radware tested the home page of every site in the Alexa Retail 500 nine consecutive times. The system automatically clears the cache between tests. The median test result for each home page was recorded and used in the calculations. The tests were conducted on February 16, 2015, via the WebPagetest.org server in Dulles, VA, using Chrome 40 on a DSL connection. In very few cases, WebPagetest rendered a blank page or an error in which none of the page rendered. These instances were represented as null in the test appendix. Also, in very few cases WebPagetest.org rendered a page in more than 60 seconds (the default timeout for WebPagetest.org). In these cases, 60 seconds was used for the result instead of null. To identify the Time to Interact (TTI) for each page, Radware generated a timed filmstrip view of the median page load for each site in the Alexa Retail 100. Time to Interact is defined as the moment that the featured page content and primary call-to-action button or menu is rendered in the frame.