This makes sense to me if you consider the type of person who is likely to leave reddit is also less likely to be just a passive consumer of content. I imagine in ten years time we'll have two kinds of "social" media: decentralized activitypub discussion-based networks, and commercial entertainment platforms that might have comments but little else in the way of connecting.

Another stat I like to highlight is the moderator-to-user ratio on Lemmy (and the rest of the Fediverse) is similarly around 10x more improved.

With Lemmy/Piefed being much smaller, I comment a lot more than I would on reddit since I know my voice will likely be heard. Outside of niche subreddits, posting a comment was like yelling into the void. It sort of all came down to how early you made it into a thread on whether anyone would actually see what you said

I think thread/forum-like platforms do a lot better at a small to medium sized community size

Bots? Chances are a lot of the traffic on Reddit that is of lurkers is just bots being used to gin the number of "unique" visitors each day. Actually, now that I think on it and reflect how Reddit went decidedly sideways when the IPO dropped, it's got to be bots..

People go to Lemmy because they got banned on Reddit or because their communities fell apart because of reedits bullshit or because they noticed that the feed is completely jacked with bots and disinformation.

In other words, the people going to Lemmy are generally highly active users. Which would probably worry Reddit if they were concerned about actually getting content from their members instead of AI and bots.

One of the weirder information I found. The website seems to have quoted semrush here, so idk, but monthly unique users from America is more than America's population

Not surprising. People are saying the same thing about x vs. Bluesky

Lemmy has like 40k MAU. Where did you get the 1.2M?

I'll admit that I'm not that updated on recent changes but reddit used to be open about the vote counts being intentionally inaccurate. I believe this at a point started to include scaling the vote count so that the average popular post would have 10-20k or so up votes, even if the actual vote count would be much higher.

So this would make it almost impossible to make any meaningful conclusions from vote count / active users alone.