RE: Quick Update Regarding User Retention
We don't know the number of fake accounts but it may very well be >90%. Please reread Dan's post and consider all the different reasons that accounts may be created that don't correspond to an actual person (including something he didn't mention -- actual people who do sign up but do so multiple times).
Furthermore some portion of users dropping off is normal and expected, especially in the time immediately after signup. It isn't ever going to be for everyone; some will try it and not find that it is their thing. It is impossible to ever achieve success with every single user.
The problem is not enough real users signing up. If we use your number of 90% fake, that gives us about 150-200 "real" users per day signing up. That is nowhere near enough to get us anywhere at all. And by "problem" here, I mean just low growth right now. If in the future enhancements to the platform and greater exposure increases that 150-200 (or whatever the number), then it won't be a problem that the growth rate was low early.
What I'm witnessing through my friends and family who are reading my blog and my parents blog and trying to sign up, is that the very first steps to registration are not user friendly.
A few weeks ago I wrote about this problem and proposed a solution to this but there another issue not mentioned in that post which is the Facebook sign up. While most people I know couldn't get passed the password to register, some simply don't like logging in through Facebook.
All the optimistic interpretations require a decrease of real users, which is not desirable.
@clayop
I completely agree. I do not find the current level of growth encouraging, though apparently Dan has some data that paints a different picture. I asked him for more detail, and it will be interesting to see what that is if he provides it.
My only disagreement (or perhaps questioning is a better word) with your interpretation is whether the issue is less growth in real users (bad) or a decline in user participation (also bad). In reality it is probably some of both, but I do feel you have been painting too harsh a picture on the latter, especially when you presented the chart from Dan's post without accounting to any meaningful degree for the issues both he and I have identified with it (fake accounts, antispam system changes, etc.)