25 March 2014
Curious how website structure has changed during the past few years? What are the most popular technologies used, and what is their current rate of growth?
Since the HTTP Archive gathers technical data on the 300K most popular sites, it's possible to analyze such changes. These are not surveys, but instead hundreds of Gigabytes of crawled, raw technical data.
At the end of 2013 average page size hits 1700 kB. This is a 30% raise for second year in a row. The biggest growth was recorder for non-image binaries like custom fonts. However typical website size s about 1020 kB and 25% of them are smaller than 475 kB. But still quarter of all websites hits more then 2000 kB what seems to be the average size in the near future.
Trends in web tech 2012 - 2013
Digging in HTTP Archive Data there is visible change in web tech usage in the last two years. Significant increase in custom fonts usage is a fact. It has increased from 7% to 36% in two years.
Flash declines slowly from being used by 50% in the end of 2011 to 30% in the end of 2013. However it is still an industry standard for reliable online advertisements.
JavaScript landscape
Todays websites are dominated by jQuery framework that is present in 70% of them. Second league frameworks like mootools, prototype or yahoo’s yui are declining slightly in favor of newcomers: backbone and angular.
| framework | change 2012 - 2013 | usage |
|---|---|---|
| prototype | -40% | (3,14%) |
| scriptaculous | -27% | (1,73%) |
| dojo | -24% | (0,36%) |
| yui | -19% | (3,63%) |
| framework | change 2012 - 2013 | usage |
|---|---|---|
| backbone | +577% | (0,29%) |
| angular | +361% | (0,20%) |
BigQuery on HTTP Archive, credits and how tos
The Google Spreadsheet App Script used to generate the stats above you can find at https://gist.github.com/mstepien/8180832#file-drive-appscript-js together with the BigQuery sql queries to play with at https://gist.github.com/mstepien/8180832
All that is based on to lya Grigorik's article HTTP Archive + BigQuery = Web Performance Answers where you can find introduction to use BigQuery on HTTP Archive data plus some Google AppScript usage example.
Header image by snowfleikun.deviantart.com.