Although it is easy to download and install free or “premium” themes, WordPress does come with a default theme that is updated yearly. Last year I did performance tests of Twenty Twelve and there were some performance-issues, which led me to create a cleary faster child theme (2012.FFWD). But how does Twenty Thirteen fare, you might wonder? And how did the WordPress themes do before? And what about Twenty Fourteen? Just the questions I was asking myself as well, so here goes!
I decided to compare the raw performance of the default WordPress themes from 2010 to 2014, creating new blogs in my multi-site test-environment. All 5 test-blogs’ homepages were tested 9 times with webpagetest.org, using the Amsterdam node, IE9 as browser and with a DSL traffic shaping profile and the median test result was used. No performance enhancing substances (such as WP Super Cache or Autoptimize) were used in this test, a few bunnies might have been slightly injured though … The result? Well, Just look at this graph;
As you can see performance (until document loaded) got worse with every release. Download time skyrocketed from 1.7s to 4.2s (and even 5.3s the upcoming 2014), mostly because download size went from 60Kb for Twenty Ten to a whopping 489Kb for Twenty Thirteen (and 659Kb for the non-final Twenty Fourteen). All details of the tests can be found in this Google Docs spreadsheet.
So what is the reason for this important performance degradation? Web Fat, that’s what. In 2013 jQuery & co were added (there was already one smallish JS-file in 2012), but even more damaging is the explosion in font-file downloads; 4 in 2012, 9 in 2013 and 15 in 2014! I’ve already expressed my dislike for webfonts and -although I think it can be immensely useful and I use it for the admin pages of my plugins- I think one should also try to avoid including jQuery where possible, especially for content-oriented sites such as the ones powered by WordPress? Unless you don’t care about performance off course.