When I finished working on my book “Practical Community Management: Lessons Learned While Working on Stack Overflow”, I began to look for areas where I could fully realize my skills (I think I am very good at growing online communities, statistics and programming and I spent my last few years at Stack Overflow building health metrics for our public communities). In my research, I’ve noticed that a few people have trouble calculating basic metrics like monthly active users, but almost everyone faces the same problem when they try to interpret the results obtained.
Let’s say you have 100 people registering on your website every day and only two of them participate in the community. Is this okay or does anything need to be improved? How can you understand whether you are building your community well by looking at some absolute numbers?
Actually, there is no way to answer those questions if all you have is your numbers. The only option to answer them is to compare your results with what other communities have. This is the essence of my current endeavor, which I want to tell you about: Online community management industry benchmarks. So, I am working on a registry of metrics that are based on real public communities. The registry will be available to everyone and community managers will be able to evaluate their communities based on real data.
In fact, I have already made the first version of the registry and it is available at benchmarks.practical-cm.com
Gradually I will add more data to the registry. For example, new metrics such as 30-day retention rate, a split of metrics by the platform on which the communities are run, and much more but at this point, it is very important for me to understand how much and what exactly you need! If you see that those benchmarks could help you at your work, please let me know! If you would like to see other benchmarks, please let me know as well! I will be glad to discuss your ideas and implement them if possible!
Let me briefly talk about what the benchmarks are at the moment. To build the first version of the benchmarks I used 15 online communities. In total, this is 466,751 users, 2,797,680 posts, 45,128 days of activity. Here is the list of metrics that I calculated for now:
- Stickiness that shows the percentage of monthly active users that are active on a given day.
- Utilization that shows the percentage of active users among those who could act.
- Time to the first answer that shows the time between starting a discussion and the first reply appearing in it.
- Churn rate that shows the percentage of users left the community among those whom we consider alive.
- Community growth percent that shows how fast the number of alive users is growing.
Notes: I calculate most of the metrics daily and then average them over month. Monthly active users is calculated as a sliding 30-day window. Since I use publicly accessible data, the only actions that I consider in my models are posts. Each metric has three values: overall value, based on all data, value for the periods when the community was growing and another value for the periods when the community was declining.
In the following posts I will share how I calculate these metrics in more detail, but now, I will be glad to hear your feedback!
P.S. I also have a very useful book on how to grow online communities!
P.S.S. Need help with measuring, researching, or predicting performance of an online community? Please write me an email! I will be glad to help =)