Page Size Matters
Written by Janet Swift   
Sunday, 08 January 2012

Recent research results have been interpreted as web page size suddenly ramping up, which is a cause for concern. Is this the real deal?

The data that led to the headline "Web pages show end of year bulge" on BBC News, a site not known for sensationalism comes from HTTP Archive a site that monitors web content and performance. The BBC reported:

average webpage sizes were trending steadily upward throughout 2011 and jumped sharply in October

Another report of the trends reported by HTPP Archive  concluded:

the average size of a single web page is now 965 kilobytes up 30% from last year's average of 720 KB

pointing out that this presents a problems for mobile surfers who are subject to a data cap.

However, while there is some truth in both these reports they miss the fact that in November 2011 HTTP Archive changed the basis on which it collects its data. Prior to this date it monitored 18,026 URLs from a variety of sources including Alexa but then switched to data from Alexa's top 1,000,000 sites.



(Click in chart to enlarge)


From this pair of charts, the jump in the total transfer size at the end of 2011 appears to be an artifact of the fact that a much larger number of sites were included in the analysis.

To investigate this look at the trend when we select only the smaller subset of 12706 sites for which data was collected in every time period - while it shows a steady increase there is no sharp jump.



(Click in chart to enlarge)


So what accounts for the increase, which works out at around 20% over the year? 

Looking at the data for JavaScript Transfer Size and Image Transfer Size for the most popular web sites they both increase and at the same  rate as Total Transfer Size and the increase in transfer size is primarily to larger graphics files.





What is also very noticeable by comparing these two charts is that the Top 100 sites manage to keep total transfer size down well below the average page size.

So why is there a jump when more websites are included. Well partly its chicken and egg. Faster websites are more popular so when you relax the  restriction and widen the net, slower websites are included.

Top websites can afford to optimize their performance (or perhaps it is that they can't afford to lose popularity by becoming slower) and so employ teams of web developers to maintain speed and efficiency.

Put simply if you increase the range of your survey to include the less special websites you have to expect that the technical quality will fall.

More information:

HTTP Archive






or email your comment to:


To be informed about new articles on I Programmer, subscribe to the RSS feed, follow us on Google+, Twitter, Linkedin or Facebook or sign up for our weekly newsletter.


Hour Of Code Adds Lessons In Unconscious Bias

Microsoft has announced details a new Hour of Code lesson. In "A Minecraft Tale of Two Villages!" the task will be to bring two villages together with the power of code, learning and practicing inclus [ ... ]

Nvidia's AI Supercomputer For Medical Research And Drug Discovery

Last month Nvidia unveiled plans to build a supercomputer intended for AI research in health care. This prompts us to look at AI's potential role in health care and how it is already being used.

 [ ... ]

More News

Last Updated ( Friday, 03 May 2013 )