Seems like tons of Mastodon advocates are the ones always pitching it as an alternative to Twitter. Every time that happens, I'll just get the numbers out since you brought up the total number or registered users which is irrelevant to active users:
Twitter: 220M+ DAU over 17 years.
Mastodon: ~50,000 DAU (1.4M MAU) over 7 years.
What matters in social networks is scale and it still means more users and there is already evidence that Mastodon not only was and is still struggling to scale, but eventually needed to re-centralize to handle their scalability issues, ie. using Cloudflare. Most of the instances there, are hosted by enthusiasts and hobbyists which intermittently struggle with more users.
If an instance is too small, it will easily be at risk of getting attacked and overloaded with users, with the instance falling over and having to close signups. If its too large, they become more centralized and to incentivise federation they have to close signups. Either way, centralization is inevitable.
It's already bad enough for normal users to 'choose an instance' if their can't register on a closed one or if a rouge admin bans you or an entire instance over something silly. Thus, It is quite clear to me, that for normal users who are not techies or geeks, Mastodon is not a viable alternative or even remotely to replace Twitter or even a recommendation.
>If an instance is too small, it will easily be at risk of getting attacked and overloaded with users, with the instance falling over and having to close signups.
Co-admin of https://functional.cafe here. Mastodon lists us as having 213 active users currently, so I guess this counts as a small instance. Haven't seen our instance fall over for more than a day or two except for hitting hardware limits, mostly hard disk space.
Also, I have no idea why you perceive closing sign-ups as a sign of failure. Closing sign-ups or going invite-only protects the current users and state of the instance (and, indirectly, the rest of the Fediverse) both from the dangers of attacks from new users and from explosive growth. Is it some sort of measure of success, to be able to always invite new people even at the expense of already existing ones?
What value does scale have, if you get only lazy retweets and likes and low-quality responses? I don't need 'eyeballs-on-tweet' kind of exposure, but interesting discussions. I get these on the Fediverse, and yes I managed to pass the 15 minutes selection of an instance. Which isn't all-too-important either, as instances federate with each other and you can migrate later, if you wish.
> What value does scale have, if you get only lazy retweets and likes and low-quality responses? I don't need 'eyeballs-on-tweet' kind of exposure, but interesting discussions.
Normal users use the block button, protected tweets and whitelisted or follower-only replies which those already exist. Problem solved. This is a non-issue.
Even by looking at the so-called 'migration', it has failed to convince existing Twitter users to stick around on Mastodon which by comparing with the figures: 220M vs 50,000 daily active users, it is not even close to 1%.
It's clear the Mastodon has decided to immediately disqualify itself and has decided to add more barriers to entry and migrations within migrations as features as you have just admitted:
> ...yes I managed to pass the 15 minutes selection of an instance. Which isn't all-too-important either, as instances federate with each other and you can migrate later, if you wish.
My point(s) still stand. An island for techies and geeks isn't a great pitch to normal people who dislike Twitter and are looking for a viable alternative.
Scale isn't the whole story. The most noticeable thing for me on fedi was that people actually saw my posts and interacted with them. All the DAU on Twitter doesn't matter when you need to pass the Great Filter of the algorithm to get seen. Every post I made on Twitter got almost no engagement, and my friends who followed me would tell me they never saw it. Instead my timeline was full of influencers, bots, and that week's outrage.
On fedi I feel like I have a small community who engages with my posts. I think that's what people want, even a few hundred followers is enough to feel heard. I can control my own timeline and find content I want. It really doesn't matter if it's only 1% the size of Twitter because my timeline is leagues better.
Engagement on Mastodon is simply higher quality IME as well.
Whether it's an "Eternal September" thing or not that Mastodon hasn't hit yet remains to be seen. It really reminds me of modern social media meets older school internet (IRC & web forums days).
I have about 100x as many followers on my biggest twitter account as on my Mastodon, and yet typically get more engagement on Mastodon... Much of it down to the algo, but also, I suspect simply churn: A whole lot of followers who are now longer active. My Mastodon account hasn't had time to accrue many of those so far.
Hold on, are you... getting DAU by dividing MAU by 28? That's very much not how that works; perhaps the easiest way to illustrate this is by considering Facebook, which has ~2bn DAU, but obviously does not have 48bn people using it every month.
In general for social media stuff, DAU and MAU are pretty close to each other; Facebook is 2bn vs 3bn or thereabouts, and AIUI Twitter is even closer.
All the important accounts I followed on twitter switched to Mastodon when Elon got his hands on the platform. I don't want Mastodon to be a Twitter alternative, really, even if it is only for some circles.
The endless Twitter drama and hot takes being pushed in your face by Twitter's engagement algorithm is tiring. I'd be happy if that dredge would be kept away from the Fediverse. It sadly isn't, but at least the situation hasn't gotten nearly as bad as Twitter's thanks to smaller communities that aren't afraid to practice moderation. I know the idea that someone might kick you off your server for being toxic is a problem for some free speech absolutionists, but personally I'm glad to see them return to Twitter/Truth Social/Gap/that weird web3 Twitter without moderation.
If you're trying to build a brand, Twitter is a lot better. Their fake engagement statistics below every post help a lot when it comes to inflating popularity statistics and you can't use ads on Mastodon to force your brand down other people's time lines. Only people who genuinely want to see what you're saying are seeing your content and that's why Mastodon will never replace Twitter:you can't buy your way to some twisted sense of popularity.
I doubt complete centralisation will ever happen. We'll probably end up with a sort of email light, where it's quite easy to set up a smaller instance but the majority of users flock to a few popular servers the same way Gmail and Outlook have replaced ISP mail accounts. However, because Mastodon and other Fediverse services don't really have a way to do AI moderation in a way that pretends to work well enough, I don't think scaling up to servers of more than a few thousand people is really viable.
Because the more people you have to read or talk to, the better? Because you want all your friends and people you want to hear from to be already there?
It's like those threads about Signal, talking about how much better it is compared to whatsapp and the rest - okay, maybe, but if my friends aren't there already, it's worthless.
No, for some maybe, for others not. Also quantity vs quality.
And also, what about those monopolies, I understand motivation, but the outcome is problematic.. If we want the everybody encompassing user platform it shouldn't be commercialized and owned like that, with a lot of idealism ;)
> but if my friends aren't there already, it's worthless
See that is another level problem there also.. if everybody jumps out of the window, I need too?How did we survive just prior to mega social networks.. huh.
So again, I know for some what you mention is the important point. But don't tell evryone, and also every playform, that if they are not aiming for super growth they are doing it wrong, because (and likely even intentionally) they aren't!
What matters is if the right people are there. Same reason why we're on HN even though it's tony compared to Reddit. And many of us are both places.
For me, Mastodon now have more of the right people than Twitter, but I still hold on to my Twitter accounts too because there are others I still want to talk to there.
What's this thing with the absolute wish to centralise everything?
The Internet is decentralisation, it's birth is interconnecting networks, having one massive network being just not scaleable.
The web is a decentralise content system, allowing anybody to add content easily is the reason it scaled. Same goes for the domain name system.
Companies like Twitter HAVE to decentralise their system in some ways to make it scale, because centralisation does not scale (using CDNs, caches, replication, etc...).
Centralisation do makes seems much simpler to build, but it's not a strength, it's a big weakness and create single points of failure.
> These stats seem to suggest that there are well over a million DAU
Please. That is the figures for the MAUs [0] and you're clearly using that to attempt to pass that as a source for the non-existent daily active users metric and you know it.
That does NOT say "Daily". Surely you can read that.
> Moreover, you're still attempting to pass off 50,000 as something you just didn't make up out of thin air.
Even it is wrong, There are no sources out there that is verifiable and you haven't provided one after I debunked your figures (which is actually the MAUs).
So as long as the exact DAUs is unknown, the figures can be estimated to be as low as 100,000 DAUs and even worst case 50,000 DAUs.
> That does NOT say "Daily". Surely you can read that.
Yes, I discussed the issue in the link I gave. Surely you can read that.
> So as long as the exact DAUs is unknown, the figures can be estimated to be as low as 100,000 DAUs and even worst case 50,000 DAUs.
This doesn't follow. I can estimate it as 2 people. Or 2 billion people. Where does 100,000 or 50,000 come from? You repeatedly refuse to give any evidence that you haven't made those numbers up out of thin air.
Anyway, it's redundant to argue the exact same thing in 2 different threads, so let's stick to the other thread please, since it has more info and explanation.
Nope. After 7 years that is the MAUs as clearly shown to everyone. There are no issues with that figure.
> This doesn't follow. I can estimate it as 2 people. Or 2 billion people. Where does 100,000 or 50,000 come from? You repeatedly refuse to give any evidence that you haven't made those numbers up out of thin air.
NO source exists for Mastodon that verifies the DAUs, since it is not given; meaning that one can only estimate. The problem with your 'link' is that you continuously use it as the wrong metric when it is for the MAUs.
So until you provide a verifiable source for the 'DAU' figures it can only be estimated. I gave mine and you can disagree with it. But without such a definitive verifiable source confirming the DAUs, I can just dismiss it like you can dismiss my estimate.
Furthermore, it is no good blaming a 'label' that has been there for 7 years and now having a problem with it given you still don't have a source for the DAUs.
I'm not sure why you continued in this thread rather than in the other thread with more info and explanation as I suggested, but anyway...
Even if we assume for the sake of argument that "active_user_count" in https://api.joinmastodon.org/statistics is monthly active users, there's still a major problem with your daily estimate, because the monthly active user count is remarkably stable from day to day.
We have 32 days of data, and on most days the change in monthly active users is only around 5000, sometimes less. There was one day where the active user count dropped 13K, and one day where it grew 27K, but those are outliers.
How is it possible that ~92-96% of Mastodon users are not daily active users, yet the monthly active user count only changes from day to day by an average of less than 10K? Your numbers just don't add up. They don't even make sense, as far as human behavior is concerned, and it certainly doesn't align with what we see on every other social network.
They certainly seem to be very loyal, consistent monthly active users. I'm not sure what Twitter means exactly by a "daily" active user. I think I would have been considered one, but I also definitely missed days, so what qualifies or disqualifies an account as a daily active user? How many days can you miss?
This seems erroneous "~50,000 DAU (1.4M MAU)". It can't be true that the average monthly user only interacts once per month (1.4M/50k). How do you explain the figure? Dividing MAU by 30 would not be a good approach.
At least one 'Toot' from a registered user a day = 1 daily active user (DAU), no matter how many times they toot, boost, reply or whatever.
> Dividing MAU by 30 would not be a good approach.
It is. By doing so with the top 20 Mastodon instances [0] we get even less than 50,000 users using it daily, hence the generous approximation. Even if we don't, we can only imply that the DAU is ultimately far less than 100,000 users a day, posting at least once on the platform.
This is why many Mastodon supporters would rather not tell me the daily active users and would immediately avoid mentioning it, which this is the real reason why they cannot claim that millions are using it 'daily'.
> At least one 'Toot' from a registered user a day = 1 daily active user (DAU), no matter how many times they toot, boost, reply or whatever.
This is not how Twitter measures DAU. In fact, most Twitter users are quiet lurkers who rarely if ever tweet. Twitter makes money (or used to make money) from their eyeballs, not from their tweets.
We are talking about Mastodon's DAU which that is still a mystery. Twitter's DAU is already known to be over in the hundreds of millions approximately.
The whole point is:
>> Even if we don't, we can only imply that the DAU is ultimately far less than 100,000 users a day, posting at least once on the platform.
Either way, it is safe to assume and with the data from Fedlist and matching it with the instances of 'active users' it is still less than 1% of the DAUs on Twitter, hence why Mastodon fans (like yourself) cannot claim and proudly show that there are 'millions of users' using it daily.
Furthermore, debunking the so-called 'Twitter migration' that wasn't.
You don't even know yourself. That statistic is "MONTHLY" active users, not DAILY active users. You gave a figure that is collected directly from the 'servers' page that literally says "Monthly Active Users". [0]
I think it's actually the "Monthly" label there that seems somewhat misleading, unless perhaps it means they update the page once a month.
In any case, as another commenter mentioned, DAU and MAU tend to be relatively close. You don't go from 1.2M monthly users all the way down to 50K daily active users, that's nonsense. https://news.ycombinator.com/item?id=35600846
> I think it stands for, you made it up. You still haven't given any citation whatever for that number.
We both know there are NO sources. You gave a source for the MAUs. Thus, we can only give approximations and estimations of its DAUs.
> I think it's actually the "Monthly" label there that seems somewhat misleading, unless perhaps it means they update the page once a month.
Now you think it's the label's fault, after 7 years? Oh no! "It's misleading"!
Something that has been representative of Mastodon's monthly usage for 7 years is now not believable and is 'misleading'! /s Oh dear.
> In any case, as another commenter mentioned, DAU and MAU tend to be relatively close. You don't go from 1.2M monthly users all the way down to 50K daily active users, that's nonsense. https://news.ycombinator.com/item?id=35600846
Not for federated networks, yet here we are comparing centralized platforms like Facebook, Twitter and counting the numbers as if they are the same which we both know they aren't.
Henceforth, give a proper source that is verifiable that shows Mastodons DAUs and NOT its MAUs next time since you're still struggling to do so.
You're being deliberately obtuse. I said the raw data for 2023-04-15 on https://api.joinmastodon.org/statistics matches exactly the numbers on https://joinmastodon.org/servers for "Data collected by crawling all accessible Mastodon servers on Apr 15, 2023." Everyone can see that, including you if you choose. I didn't say the numbers were "not believable".
The question is what "Monthly" is supposed to mean exactly, given that the raw numbers change every day, as everyone can see from the above link. And as I said, one possible interpretation is "unless perhaps it means they update the page once a month".
> Not for federated networks
Says you, with no evidence, again. What exactly was your methodology for producing the estimate "50,000"?
> You're being deliberately obtuse. I said the raw data for 2023-04-15 on https://api.joinmastodon.org/statistics matches exactly the numbers on https://joinmastodon.org/servers for "Data collected by crawling all accessible Mastodon servers on Apr 15, 2023." Everyone can see that, including you if you choose. I didn't say the numbers were "not believable".
And that represents the 'monthly' usage as I repeatedly said. That only shows MAUs, not daily. Never was a problem for 7 years.
> The question is what "Monthly" is supposed to mean exactly, given that the raw numbers change every day, as everyone can see from the above link. And as I said, one possible interpreation is "unless perhaps it means they update the page once a month".
It is clear that it is representative of the MAU of every "active" Mastodon instance and corroborates with that fedi-list link I gave which also clearly shows "MAU" and matches with that.
There is no room for excuses or blaming labels after 7 years of Mastodon's MAUs count.
> Says you, with no evidence, again. What exactly was your methodology for producing the estimate "50,000"?
Yet it seems that you haven't even tried refuting it with any verifiable source and you are continuously passing the MAU metric as the DAU which is clearly incorrect, which I can dismiss your so-called 'DAU' source for Mastodon. Until then:
"Give a proper source that is verifiable that shows Mastodons DAUs and NOT its MAUs next time since you're still struggling to do so."
Me: What exactly was your methodology for producing the estimate "50,000"?
You: Yadda yadda yadda... [no answer]
I'm actually fine with taking 1.2 million as the assumption for the monthly active user count, but it's still absurd and inexplicable to go from that to a 50K daily active user estimate.
Do you have a verifiable source for the DAUs? Not MAUs?
You: Errr......
So once again, you come back and have zero verifiable source(s) for the DAUs count for Mastodon.
> I'm actually fine with taking 1.2 million as the assumption for the monthly active user count, but it's still absurd and inexplicable to go from that to a 50K daily active user estimate.
It has always been that source for Mastodon's 'Monthly Active Users' for years and you knowingly used it to prove the numbers for the 'Daily Active Users', then when questioned you blamed the 'label' as 'misleading'.
The "Monthly Active Users" label could not have been more clearer and has been for years. You are free to dismiss my estimate, but once again you have admitted that you don't have any sources to even show me the DAUs.
So again:
"Give a proper source that is verifiable that shows Mastodons DAUs and NOT its MAUs"
I never had a strong opinion about how exactly Mastodon measures "active_user_count". I did find it curious from https://api.joinmastodon.org/statistics that they seem to be scraping the servers every day, and this suggested to me that they might be counting daily active users. I could be wrong though. I don't know the technical details of how they calculate it. And I already said earlier: "In any case, as another commenter mentioned, DAU and MAU tend to be relatively close. You don't go from 1.2M monthly users all the way down to 50K daily active users, that's nonsense." https://news.ycombinator.com/item?id=35602391 So I was already granting that 1.2M could be the monthly user account, even while considering the possibility that it might be the daily count. I don't really care either way. I don't think it helps your argument if it's the monthly user account. So let 1.2 million be the monthly user count. Now will you explain your estimate methodology?
At this point I doubt that you will. You seem to be very committed to avoiding the question, even though I've granted the assumption that you wanted.
Twitter: 220M+ DAU over 17 years.
Mastodon: ~50,000 DAU (1.4M MAU) over 7 years.
What matters in social networks is scale and it still means more users and there is already evidence that Mastodon not only was and is still struggling to scale, but eventually needed to re-centralize to handle their scalability issues, ie. using Cloudflare. Most of the instances there, are hosted by enthusiasts and hobbyists which intermittently struggle with more users.
If an instance is too small, it will easily be at risk of getting attacked and overloaded with users, with the instance falling over and having to close signups. If its too large, they become more centralized and to incentivise federation they have to close signups. Either way, centralization is inevitable.
It's already bad enough for normal users to 'choose an instance' if their can't register on a closed one or if a rouge admin bans you or an entire instance over something silly. Thus, It is quite clear to me, that for normal users who are not techies or geeks, Mastodon is not a viable alternative or even remotely to replace Twitter or even a recommendation.