This is a topic of great interest to me. Somewhere around the late eighties or early nineties I began a somewhat morbid hobby of watching online communities di

This is a topic of great interest to me. Somewhere around the late eighties or early nineties I began a somewhat morbid hobby of watching online communities die. I will try to keep my personal conclusions out of it and allow you to form your own, but will at least hint at my methodology. First, you say "healthy" and I say "not dead." Having seen a lot of communities declare themselves healthy and vibrant, doing all the right things, and so forth, only for them to wither and die, statements of health are meaningless. Look for multiple measures. You know it is dead when they pull the plug, but what about the withering before that? What does that look like? Identifying that is going to be key to determine what health looks like, because it looks like "not dying." User turnover. Moderator turnover. Account creation. Account retention. Rate of comments per user. Is the "old guard" still present? (This sound subjective but can be worked on) Less objective are incidents of "purity spirals." These are a great way for a community to die. Find incidents and examine them. Look for commonalities that can be abstracted into numerics.