In November 2015 I published an analysis of the WordPress plugins repository. I was predicting that 2015 was going to be a record year in terms of plugins submitted to the repo. While working on the data I also discovered that 4 out of 5 plugins in the repository were not updated anymore after 2-3 years from their addition to the repo.
The post was relayed by WPTavern and there were quite some discussions in the comments, especially about the findings on the longevity of plugins.
Recently there was a renewed interest around data in the WordPress community with quite a few posts on the topic. All this activity prompted me to update the analysis I did back then. Without further ado here are the results.
WordPress Plugins Additions declined in 2016
A year ago I looked at the results of the first four months of 2016 concerning the WordPress plugins repository. Everything seemed to head towards another record year. Like the majority of predictions, things turned out a little different.
That’s why I decided to approach this post like real consultants do, i.e by predicting the past…
As the title of this post suggest, 2016 was the first year that saw a decline with regards to the number of WordPress plugins that were added to the repository.
In absolute values, the WordPress plugins repository saw in 2016 about 400 plugins less than 2015. In relative terms this represents a 5% drop. Before anybody asks, I am not spelling doom and gloom here, just reporting the numbers.
Can we draw any conclusion from this graph? Hard to tell, we know that there were not operational issues when it comes to the WordPress Plugins team. We can simply assume that there were less additions.
My reaction when I saw the figures
I am good at forecasting the past
As you can see, during the first part of 2016, the amount of submissions was always at the same level or slightly higher than 2015. Things changed as of July 2016, since then plugins additions were consistently lower than the previous year at the same period.
By the way, for those of you interested in Information Visualization, here’s a better representation of the above graph to highlight differences. Through this graph you can immediately see and quantify the difference between each period. A further improvement could be the color coding each bar whether it’s a positive or negative difference, but I am digressing. As you can see, as of July 2016 plugins additions were consistently below 2015.
The major finding of my 2015 post was that a very high percentage of free WordPress plugins do not get updates after 2-3 years from their initial submission.
Let’s have a look at the Last Update date for plugins (an indicator of the plugin maintenance). We can see that only 25% of the plugins introduced in 2013 were updated in 2016 or later. It confirms that within 2-3 years there’s an 80% chance that plugins won’t be updated anymore.
The reasons why this happens are multiple, but there’s no way to get a clear reason. Guts feeling tells me that 80% of the reasons probably fall under these two categories. Reason number 2 is the main one (just a personal feeling, no hard data to back this up):
- Plugin is still functioning as expected, and there’s no need to update
- Author is not actively maintaining the plugin anymore
If you look at the above graph you can see that a few years after being added to the repository plugins tend not to be updated. For example, plugins added in 2013, about 65% of them haven’t been updated in over 2 years. The good news here is that based on my previous findings I was expecting to have the percentage more towards 75-80%.
How can a one-year old plugin have two-year old update date?
If you have a sharp eagle eye, you might notice in the graph two small red segments for the year 2015 and 2016. The red segment in 2015 can be attributed to plugins that were added at the beginning of 2015, but haven’t been updated since. The red segment for 2016 is more disturbing. By definition, no plugin added in 2016 can have an update date from more 2 years ago.
I had a look at the details and there are about 70 plugins that have an Added date in 2016 while their late update date is way in the past (some even in 2008). As this is a really small amount ( about 1% of the yearly amount of plugins) I will just leave it. I am guessing this is a result of some maintenance work on the repo.
Those 70 plugins though are clearly not from 2016. If we deduct these plugins from the 2016 results, the decline for 2016 is getting close to 6%.
2016 saw the first decline in plugins additions to the WordPress repository, with about 450 less plugins or a 6% decrease.
The findings concerning the update cycle for plugins of 2015 still hold true with some good news as the percentage of plugins updated has slightly increased.
Overall I don’t think this small decline in plugins submissions is a real issue. WordPress and its ecosystem are still enjoying a healthy growth. Let’s see how things will evolve in 2017.