Given that I first wrote my plug-in WP to Twitter almost 8 years ago, and it’s been in continuous development since then, you might expect that I know all about it. But that would only be half the truth.

What I know about WP to Twitter is the code base. I know how it works, I know why I made various decisions, and I know what I really want to change in future development.

But I know very little about the user base and what they might need from the plug-in. That’s the primary reason that I recently integrated the Freemius opt-in product into WP to Twitter. It’s only been a week since that release, but I’ve already learned some really interesting information that I’ll hopefully be able to put into immediate use.

Things I’ve learned

The first thing I learned is about internationalization. While the most common language group using this plug-in is English (not surprising), the second most common appears to be Japanese – by a significant margin. I wasn’t expecting that, and it’s very interesting. It’ll definitely be relevant in focusing my efforts on reaching out to translators to improve the internationalization of WP to Twitter.

The second thing I’ve learned, after only a very short time with the data, is about the most common reasons the plug-in is uninstalled. Obviously, it’s an important goal for me to deliver a plug-in that does what people expect it to and is easy to use. On the whole, I think I’m succeeding at that: it’s not a huge percentage of users uninstalling the plug-in. However, the most common reason is a run-away with “Didn’t understand how it works.” This makes it clear to me that making the usage of the plug-in clearer is a good goal to tackle.

Supporting the past?

I also learned some interesting statistical information that is less actionable. I learned that there’s at least one person using WP to Twitter on WordPress 3.5, which is well out of the versions of WordPress I’m still officially supporting or testing on. But it’s nice to know that it apparently still works!

In terms of PHP (Hypertext PreProcessing) versions I need to support – well, there are considerably more people still using version 5.2 than there are on version 7. There’s strong evidence that moving forward from version 5.2 isn’t something I can realistically do right now. It’s not an enormous percentage, but it’s enough people that I’d feel pretty irresponsible making that kind of change today.

Altogether, gathering statistics has been tremendously valuable in terms of understanding the long-term future of the plug-in. I should have done it years ago. (Which will be a future post: “mistakes I’ve made with WordPress plug-ins”.)