Lovely stats, and thank you for taking the time to dig them up and show them here.

Of course we track these as well but I've pulled some others that might be of interest.

This is the gini coefficient for followers. 0 is perfect equality - every user has same number of followers - and 1 is for perfect inequality - 1 user has all the followers.

Curation

User attrition

There's a lot to digest here. Moving forward there is a couple of things Liketu users can look forward to. In the immediate near term, within 2 months, a mobile experience is coming. That has been a strong pain point for us for over a year.

Along with that, there will be a thread like feature we'll be calling something else which will allow users to post even more micro / bitesized content that is not part of their 'main' blog.

We do a lot of stats mining because user growth to us is the most important. We believe optimizing for more users increases network effects, so a crucial step we're taking is to democratize the curation such that our curation efforts will be based on signals from users of liketu themselves.

To achieve this, we've had to revamp the 'Engagement league', work out relative network influence via eigenvector/katz centrality studies and then work out a score for each user. This is kind of like an on-chain rep score. This score is then used to calculate the 'weight' of their votes (reacts) on posts. After a given amount of time, the Liketu curation account will vote for what the users deem the best posts.

There is no world where the opinions of the few outweight the opinions of the many. The challenge was to find a way to see which users carry the most weight based on their activities on chain. More proof of brain, and more optimizing for attention is the way forward.