Comparing retention rates between Fakebook and Steem illustrates the difference between normal and abysmal performance. Your figures would provide Fakebook an overall retention rate of 50%. To keep math simple I'm going to estimate Steem retention, and the contribution of socks, which are multiple accounts of one user. We know that some highly staked users have thousands of accounts, and this dramatically impacts estimates of user retention on Steem.

Fakebook user retention is 50% overall, and over 1M accounts on Steem with ~10k active users leaves 1% overall user retention on Steem. That's not normal. It's abysmal. It's 50 times worse on Steem which claims users receive nominal financial incentive for participation than on Fakebook, which makes no such claim.

Given the motivation financial incentive provides, it's highly revelatory of just how poor Steem management is at supporting society. Is it possible that paying people discourages them from participation? No! Not at all. What is factually provable is that Steem is paying the wrong people in such a way as to drive everyone else away.

There's no other explanation for the monstrous dichotomy in user retention, and why I have repeatedly and often pointed out the fundamental difference between profiteering and investment. Steem is so codified as to reward profiteering, and society is based on investment and building civilization - the opposite of profiteering and the destruction of civilization.

It's completely bass ackwards from what it should be doing to support society.