Site speed is one of the most important elements in providing a good online user experience as nobody wants to sit around waiting for pages to load.

In fact research has shown that a one second delay in load time can cause 7% loss in customer conversions.

But a new report from Radware suggests that despite the effort that businesses put into ensuring a fast load time, average site speeds are actually getting slower.

The study shows that in spring 2012 the median page took 6.79 seconds to load, but this has now dropped to 7.72 seconds, a slowdown of 13.7%.

Among the top 500 leading retail websites, as ranked by, Ikea was found to have the fastest full load time at just 1.85 seconds followed by AbeBooks (2.06 seconds) and Pixmania (2.20 seconds).

10 fastest sites (full load time)

However the report also measures sites based on their ‘time to interact’ (TTI), which is defined as the point at which a page displays its primary interactive content (e.g., feature banners with functional call-to-action buttons).

Using this metric, Adpost actually clocks the fastest time at 0.7 seconds followed closely by Groupon and AbeBooks (both 0.8 seconds).

To give an idea of the distinction between the two metrics, Netflix has a TTI of 1.4 seconds but a full load time of 4.37 seconds, while LivingSocial has a TTI of two seconds and a full load time of 8.58 seconds.


According to the report, TTI is actually seen as a more relevant metric than the full load time as it’s the point at which visitors can begin to use the site.

Among the top 100 sites tested the median TTI was 4.9 seconds. Only 8% of the top 100 sites had a TTI below two seconds, while 9% had a TTI of eight or more seconds.

In order to reduce the TTI, Radware suggests taking the following action:

1. Defer rendering ‘below the fold’ content

Page speed can be improved by delaying loading and rendering any content that appears below the initial visibility area.

To eliminate the need to reflow content after the remainder of the page is loaded, replace images initially with placeholder <img> tags that specify the correct height and width.

2. Ensure that interactive features are optimised to load early and quickly

The report suggests making sure that homepage elements such as carousels are optimised to be among the first elements to load, as on the carousel didn’t begin loading until the two second mark then took a further three seconds to load fully.

However a better idea might be to remove slow elements that also do little to improve the user experience. We’ve previously investigated the use of carousels on ecommerce sites, and found there’s some doubt over whether they’re worth the effort.

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. 

For example, scripts that support interactive user behaviour, such as drag and drop, can’t possibly be called before the user has even seen the page. The same logic also applies to script execution. 

Therefore it’s a good idea to defer as much as possible until after onload instead of needlessly holding up the initial rendering of the important visible content on the page.

David Moth

Published 30 July, 2013 by David Moth

David Moth is Editor and Head of Social at Econsultancy. You can follow him on Twitter or connect via LinkedIn

1719 more posts from this author

You might be interested in

Comments (4)

Stuart McMillan

Stuart McMillan, Deputy Head of Ecommerce at Schuh

Thanks David, always keen on a page speed post... Great to hear people talking about "time to interact", as it really is what matters most to users. However, it can be difficult to measure in a consistent way.

Our newest site is our tablet site (, it was very much built to minimise time to interact, it takes about 1.6s for it to be ready for the user (including first carousel image), there are then further carousel images loaded in the background. The basket indicator and other dynamic elements are then service based and loaded in after page-load. One of the reasons for doing this was to allow the page to be cached without any "user-specific" elements. The site is perfectly usable without them.

I'd encourage people to start off with a basic site that loads as fast as possible, then progressively enhance it through web-services post page load, although you must be careful to stop the entire page redrawing and interrupting the user experience.

almost 5 years ago



Thanks David for such a good finding, page speed is always important. It is not only imp for seo but also Imp for use engagement to website.
and key things like logo, headline etc must be optimized for quick loading.

almost 5 years ago


Tammy Everts

Thanks for the great write-up, David. I directed this research at Radware and helped author the report. If any of your readers have any questions about our methodology or findings, I'm happy to field them.

I also wrote a blog post in which I offered a bit more background into why we undertake these benchmark studies and what led us to measure "time to interact" as a more meaningful user experience metric:

almost 5 years ago


Adrian Swinbourne

Very useful article indeed. In Google Analytics it looks possible to measure TTI - either using Average Document Interactive Time, or Average Document Content Loaded Time.

Are these metrics reliable enough to use in reporting and which one is more relevant in the case of TTI as it seems Average Document Interactive Time is the best one to use?

almost 5 years ago

Save or Cancel

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Digital Pulse newsletter. You will receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.