Fediverse

It's a normal situation when you have big migrations like the reddit migration. Same thing happened on mastodon with the various twitter migrations.

A lot of people migrate thinking they're reaching the promised land, realize they aren't getting what the want from the new platform and go back. It's the nature of bandwagon jumping.

It's just fine. The process of growth is dynamic, and the people who remain are the ones who like the platform.

I think the general consensus seems to be bot and/or spam accounts being removed but I haven't seen evidence of it. Anecdotally I have seen many people say it feels more busy around Lemmy lately but I haven't been here long enough to know myself. It is a curious one though.

If it is genuine then I suppose it could be a typical cycle for platforms that get raided by another, in this case an exodus from reddit to Lemmy - then after the dust settles some people decide they don't like it after all and go back or elsewhere.

I don't think that is a bad sign though that is a natural cycle and as long as some people remain then it shows overall growth.

A plausable explanation is that a ton of people have been instance hopping (making different accounts on different instances) before finally deciding on their primary instance. I have an account on lemmy.ml and lemmy.world before sticking with lemm.ee, so that's 2 inactive accounts.

Pretty sure there have been a few spam bot/latent bot purges over the last month on multiple instances. I remember people discussing them a little while back. And it was a ton of bots.

So part of what you're seeing is normalization, as the user numbers were always artificially inflated with bots.

When looking at graphs, it's extremely important to consider all possible aspects that the data isn't capturing. A lot of the time, it's easy to take the "easy" interpretation of the data and get the wrong conclusion (for instance, think about survivor bias and how it almost led the British military to the wrong conclusion about where to reinforce their warplanes)

Here, it's important to remember several things: what exactly is the data counting? And what happened that might change our interpretation of the data?

For the first, it's unclear what the statistic is, but I think the general interpretation is that "active users" only counts people who have posted or made comments. For the second, ofc the Reddit Migration just happened a month ago. The fact that it happened almost exactly a month ago likely isn't a coincidence.

Here's my interpretation: people from Reddit jumped on board to Lemmy during the Reddit Migration. They posted or commented a lot to test out the waters on Lemmy. Then, once they settled in, they started lurking (after all, the vast majority of people lurk). As the month continued, these new users are no longer considered active users, since they're only lurking. So the "active last month" count is dipping almost exactly 1 month after the Reddit Migration. Of course, part of the dip can be explained by people moving back to Reddit. But based on my understanding of how "active users" is counted, I think this is the leading explanation, especially since Lemmy feels more active now than during the Migration.

Now, what can we conclude about the dip? Honestly, if my interpretation is correct, this seems pretty normal. I wouldn't think too deeply about it. As Reddit enshittification continues, we might expect more waves of migrants, and I generally expect that we'll see this pattern every time (a sudden increase in userbase, followed by a shallow dip after 1 month, and then the number starts to stabilize)

There was a big influx of new users from Reddit in June. Many of those probably made an account to check out Lemmy, but somehow lost interest. Lemmy user count will most likely stabilize in a month or two.

I'm lurking, this is my first comment since signing up!

Bots that don’t identify as such count towards active users. There have been a number of bot purges.

Regression to the mean?

Off-topic question, but where can you see the stats?

Yep, it’s slowed down and it’s noticeable