[ftfish]

Never said that, though, plus provided sources. Just adding context for what the total number of users means.

[jacobgold]

You still seem to be implying the number of real registered users on Bluesky isn't ~50M, which it is. The PLC identity spam you referenced is not being counted in this number.

[ftfish]

If that's the case, then I stand corrected, but a source for that claim would be helpful.

[jacobgold]

I was rounding up, the actual number is ~45 million: https://bsky.jazco.dev/stats (these stats are based on real activity, not PLC identities)

At current rate it will 50M in 6 months.

[verdverm]

It is confusing to say "users" when it is actually "accounts", humans tend to associate "user" with another human, where as "account" can cover people and bots (many on atproto)

eg. I personally had more than a dozen accounts

[jacobgold]

That's a fair distinction to make but my (educated) guess is that something like 95% of users have a single account on Bluesky.

Most users signed up by downloading the app or visiting the web site and created a single account.

As a point of reference, the official Google Play store independently verifies that the Bluesky app has had 10M+ installs.

https://play.google.com/store/apps/details?id=xyz.blueskyweb...

There are no official stats on the iOS app or web but those are both likely similar or larger sources of users.

[verdverm]

I would counterpoint that a non-insignificant number of the accounts are spam/bots. Jaz's stats overestimate because it does not remove certain accounts which have been takedown/deleted, let alone those that remain. I have shared analysis of this on HN, Bluesky, and Discord to avoid making educated guesses.

The reality is that a significant chunk (>50%) was Blue MAGA, or turned off by them, and I see little prospect that they will reactivate. Outside of the Bluesky echo chamber, there is a deep brand association with Bluesky being primarily political refugees. They see Bluesky as the left-wing Truth Social. I've asked hundreds of people IRL if they have heard of Bluesky, they are far more likely to have this brand association than to have heard of "atproto" (more than half vs 1-2% / can count on one hand).

[jacobgold]

Sure, it may overestimate the exact count, but that doesn't change the fact that tens of millions of real people have downloaded Bluesky and signed up.

Right now, Bluesky has one large community, which is already great for some people but not most people. Once Bluesky adds Communities, new communities can form, making it interesting to the other 90% of people who were excited initially and then turned off by it being "one note".

[verdverm]

I think your association with Blueksy is painting a rosier picture for yourself than reality portends. Do you know how many of those accounts never added a profile picture or even liked a single post? Do you think people are likely to reactivate to something they checked out once and has only shrunk since? Can ActivityPub/Mastodon add some new feature that will reactivate all the people who tried that out and moved on?

If you'd like to crawl the network to get real data, I built this a while ago https://github.com/verdverm/atmunge (it will take a few days to get sufficient data to do analysis, due to rate-limits)