Akoha experiments, open-sourcing ‘django-lean’
In my first post to this blog, I’m excited to announce that Akoha has just released django-lean, our open-source experimentation framework. Read on to find out more about why we created this tool and how it might affect you.
In order to best serve our Akoha community, we have made it our mission to learn more about who our community members are and what they are seeking from Akoha. One of our recent initiatives is the use of in-site experiments (also known as a/b or split tests) to gauge the effectiveness of our changes to Akoha.
Using experimentation, we are now able to implement a new idea but expose it to only a portion of our community members. Over a few days, we watch the aggregate activity of those users who saw the new feature (the test group) and compare it to those who did not (the control group). If the test group demonstrates more engagement with Akoha (more missions played, more comments, …) we consider the experiment a success and roll it out to all of our users. If the test group engagement is unchanged or worse as compared to that of the control group, we roll back the change and reconsider our assumptions.
For the most part, you will probably not notice that any of this is going on. But if you find yourself spending more time on Akoha over the next few weeks and months, it might just mean that our experiments are bearing fruit!
When we set out to begin our experimentation journey we found a lack of tools to help us along the way. We created django-lean to fill that gap, and decided to open-source it so that other developers might benefit from experimentation. For the technically minded amongst you, you can check it out at bitbucket.org today or come hear our flash presentation at Montreal-Python 9 on September 30th!
Posted in Akoha, AkohaBlog, Web Development having no comments »
