Realistically how does the fediverse make money and remain sustainable? only donations?
I started using mastodon when musk bought twitter and kbin with all the reddit API shitshow. I like the idea of the fediverse but how does it have longevity?
I'm new to the fediverse but have a long history with FOSS and P2P. Realistically, the fediverse is not a money-making venture. It's a passion project to fill a gap in existing software offerings. The source code is hosted online (just confirmed by visiting kbin and Lemmy on GitHub). But the reason they can run without massive influxes of cash is because they are P2P in the resources they consume (fediverse style).
So donations are key. @ernest won't get rich on donations, but they can keep him supported while he develops and maintains kbin, and his experience with kbin could underpin and propel a very prosperous software engineering career, where his reputation precedes him. The passion project continues and the software remains in positive development. I've done the exact same thing in my career and I'm probably 5-10 years ahead of where I'd otherwise be because of my FOSS passion project.
Since the fediverse is a peer-to-peer structure with freely available foundational software, anybody can pitch in resources. Anybody with a computer, an Internet connection, and some sysadmin skills can run a fediverse server. Mileage will vary based on connection speed, hard drive space, etc, but it's completely doable. Lemmy/kbin servers are probably going to be the next most popular service that homelabbers start hosting, and peak-demand can be offset with scalable solutions like virtual servers (VPS offerings like AWS, but these cost more immediate $).
So knowing that, hopefully your question is answered. Short story is the fediverse follows the peer-to-peer (P2P) backbone that torrenting and other piracy methods used (e.g. Napster or Soulseek, but without a central server!) and similar to Tor as well. I specifically name those networks because P2P networking has been incredibly resilient for decades (except Napster, whose central server was shut down).
Decentralization is not only the future, I fear it’s our only hope of maintaining an internet that is truly free as opposed to one built solely for data collection and corporate profit.
To that end. Passion project can end of life, if the dev quits or have other priorities in life. Also something to keep in mind that comes with decentralized structure - the lack of invested continuity.
Perhaps some form of replication could come to the fediverse. If instances can choose to work together to canonically replicate content, that would increase resilience.
The other side of that coin, though, is that money turns something into an obligation, and obligations are soul crushing.
People can always offer to accept a transfer of ownership from a retiring admin. That's happened multiple times on the Fediverse in the past year, and it used to be relatively common back in the days of community forums.
I pay for a seedbox that has 4TB of storage and ungodly speed. Technically I could load my own instance of Kbin yeah? But then what? To be useful to the fediverse at large wouldn't I have to open it up to the public to use? What would attract users to some randos kbin server over any other?
The fediverse is a step towards p2p but there is further to go before "anybody" can pitch in resources. Running a server is not for everyone, I've done it and won't do it again (I've been doing this stuff for a long time BTW).
This creates an imbalance been those running and controlling servers and those not, which has a number of downsides which in the extreme lead to capture and centralisation. The fediverse is a fantastic attempt to reduce those downsides (FOSS, portable accounts etc) but the server/service model remains at the core of the downsides including as the OP asks, how this can be funded and sustained.
IMO the fediverse will struggle to do this at scale. It is here to stay but will find it difficult to reach so many of the population because while small it remains separate, and where it scales sufficiently it becomes attractive for capture and centralisation.
So IMO we need something that breaks from the server model as well as solving the problem that not everyone can contribute resources, and I'm hoping p2p projects can do that.
If there was a p2p network where anyone can contribute resources just by running an app, and anyone can use by paying those running those apps, I think we would be able to have the nice things we want - at increasing scale - without fear of capture, and the real prospect that everyone can participate without being made subjects and exploited.
Is there any reason an instance couldn't just copy Reddit's model? (Advertising that's hidden if you subscribe, paid awards, etc.)
The beauty of federation isn't that we have a million instances. It's that if there's a problem with one instance (such as greed), we can just go to another. In effect it adds competition instead of walled gardens.
Even if there's only a couple big instances that make up the majority of users it's not really a problem as long as the federation option holds. It's not the leaving that makes federation great, it's the ability to leave.
An instance can do whatever it wants, within the terms of the license of the software it runs.
But so can every other one. And ad drive social media has demonstrated itself to be toxic social media. So, you'll end up with a lot of other sites defederating from it.
Being ad driven is what has led to Reddit doing what it's doing, after all. The people who have come here from there should be much more concerned about emulating that kind of model all over again. Otherwise, why not just stay there?
It needs to be a subscription model, an ad model, or a donation model. It's not free to host a website especially one with millions of users. I think a Wikipedia style model could work.
True, I always throw a few euro at Wikipedia when it needs some, and I regularly fund the open source software I use (like Krita).
I'd be curious to know if there are any data on how many people on average would donate to support something they deem valuable vs people who don't.
I mean, this is not a project started for profit, but if a significant chunk of reddit migrates here, it could get a pretty penny just out of donations, I think.
While I'm not discouraging you from donating to the Wikimedia Foundation, it's worth knowing that it's not really a situation of Wikipedia needing your money to survive. Last year's fundraising drive provoked something of a revolt among contributors (they seem to be going around). While it's spread across several pages, the English Wikipedia Request for Comment discussion is probably a good place to start.
In short, the Wikimedia Foundation runs those fundraising drives, previously presented as if they were coming from Wikipedia itself, asking for donations to "keep the site running". In reality, these donations go to the Foundation and are used to fund its "sister projects", leading to a situation where "Wikipedia" is asking for donations that are neither necessary for Wikipedia to survive, nor in most part, going toward Wikipedia at all.
I'm not suggesting the Wikimedia Foundation is off committing acts of evil with your donations, but the Foundation has significant wealth, to the point where they themselves actively donate large amounts to other charities like the human rights org Tides Foundation, with which it shares some current or former staff.
tl;dr: Wikipedia isn't actually struggling financially, isn't asking for donations, and your donations to Wikimedia are in reality going largely to other causes.
If you're a former Redditor, I hope this multi-paragraph "Akshually" screed makes you feel at home.
The idea that any given instance here is going to host "millions of users" is, uh... Frankly, kind of counter to the point of the whole space, really.
Millions of users will be spread out over thousands of instances, and those instances cost from a few dollars to a couple hundred dollars per month to run. Most of these will be primarily funded by the admins' day jobs.
There will be a couple of very, very large servers, which will be a bigger money sinks, but those have also successfully been funded via donations, and are usually run by the underlying software maintainers.
At present, and throughout Rexxit, most need users are going to rock to the largest instances. Yes, this is contrary to the spirit of the fediverse, but it is consistent with the perception of Rexxitors. Perhaps eventually, things will spread out . I think the initial de facto centralization will persist for some time.
It depends on the scale, and on who the advertisers are. The whole idea that ads-based social media is toxic isn't due to annoyance, but due to the fact that it incentivizes the hosts to extend user sessions and stimulate engagement. That means promoting more controversial content, and refusing to moderate toxic people.
The hardware is getting smaller and cheaper, it’s coming pre built with Docker and decentralized apps. It’s catching up. I see a future where you type a command and you have a running instance on your network. On a usb drive. Wherever.
I honestly think they should add rewards for replies like Reddit had - simply because it makes money for the website, and then it's less reliant on absolute goodwill.
All we've known for decades is ad revenue and user data mining with us the users as the product. Reddit and twitter becoming greedy hopefully kick starts a new side of internet. I'd donate monthly to a developer to protect a site where we can learn, talk to each other and shit on each other as well once in a while.
#believe or not, some people do things for the love of it and not for profit. it's very sad that even the very idea of this, seems be lost on on the internet today.
the earliest public online communities were dial in BBS. People would buy 2nd PC's (when PCs cost a couple of grand), get second phone lines (some time more than 1!) modems etc. and set up a dial in BBS because it was their hobby. For the fun of it. For the pleasure of it. For the thrill of it. Some people rock sneakers, some people soup up cars, some people carve wood but they do so because it is a hobby., It makes them happy, they meet new people, they learn new things. The idea of making money from it ? well it never even enters the heads of most of them.
##Thoughts of revenue stream, and clicks, and engagement are all the bastard children of capitalism , which infects all hobbies like a virus. Skating, fashion, Bmx, gaming, fitness, - nothing is immune from the foul contagion.
#the immaculate beauty of a federated social media platform like KBIN is that you can host your own instance, a for a sizeable number of users, for like £20 a month on a VPS. It plugs into all the other instances using internet magic, and you can be the boss. If instead you wnant to kick in a couple of quid to your favourite host , have at it. If you're broke, well, you're still welcome.
community is a thing and it will outlive capitalism.
The OP didn't say anything about profit. As kbin and other instances get bigger, the cost involved to keep it running increases. I don't think anyone expects the developers to shoulder the entire cost, so they must generate revenue. The big question is how they will go about this.
Although I get your point, I would come at this from another angle. For example, I’ve never heard anyone say that a website “only” runs on advertising. We somehow take this as a given as if there’s nothing more normal on the planet than a business sustaining itself solely through our various corporate overlords.
I think we need to reframe this approach and instead make user-funding the norm, exactly as it is with every non-internet service. No one goes to a car wash or a house painter or a whatever and expects the service to be free.
Google fucked our minds so much with their approach to revenue generation that we somehow remain blind to what a strange approach to service provision that this is, but it’s end result is clear—an internet built not for us, but by the rich and for the rich. Fuck that.
Most fediverse instances are small enough for a hobbyist to run on their entertainment budget or small donations. Very few of them are the size of, say, Mastodon.social. Even some of the next largest ones like mstdn.social cost about $400-something a month which is a lot for a hobbyist but is covered by donations by virtue of being so big. But your average fedi instance? In a recent reddit thread where someone asking the same question, the reply they got was "it runs off a Mac mini I keep in a drawer". This is why fedi is sustainable. (This is also why people trying to give Lemmy and Kbin instance admins advice on how to grow indefinitely miss the point, fedi works by not having instances be the size of reddit)
Yes, exactly. This growth for the sake of growth mindset is a global sickness we need to overcome. If my Kbin instance has enough people I find interesting and they in turn don’t hate my guts, that’s enough for it to be considered a community by me.
I don’t need infinite growth and more than the admins fear it, and I sure as hell don’t need advertising here to further encourage yet another endless cycle of growth.
As a developer myself, thinking about the cost factor can also lead to some positive outcomes, too!
Trying to lower costs will incentivize minimalist design and implementation, leading to something more akin to old reddit. Not surprised that Kbin, Mastodon and Lemmy's interfaces are extremely simple compared to modern reddit and other social platforms.
All of the crazy tracking that people using Reddit alternatives don't like is also very 'heavy' from a web performance perspective. Cutting that out immediately improves load times and reduces data usage.
Relying on donations removes the main incentive to add pointless features to categorize and identify users for advertising.
I suspect a lot of Reddit's profitability issues are due to growth-imperative things that go beyond just running the site. In their attempt a blitzscaling, they are developing recommendation algorithms, SEO, and other ways to lure users. They have departments to do ad sales and develop user targeting systems for ads. There's also R&D the development of the site itself, including stupid features like RPAN and Chat to try to get the normies to join.
A kbin or lemmy instance, while not cheap to run by any means, doesn't need most of those functions. The R&D costs can mostly be solved by open-source software development, and frankly old.reddit.com-level tech was good enough anyway. There's no SEO, no algorithms, no ad sales, no profiling and targeting, etc. Most of the cost that's left is probably hosting, maintenance, and trust + safety and moderation.
I think hosting and bandwidth could be cheaper than people expect. I believe all of HackerNews was/is run off a single machine to serve 6M requests per day. If this stuff eventually takes the place of early reddit, the scale will be bigger than HN, but I believe probably maintainable via donations drives. You might have the big annoying banners like Wiki does, but I think it would be possible to cover the hosting. (Reddit also has extra-expensive hosting costs due to the use of AWS, which hopefully Kbin would be able to avoid.)
Trust and safety and moderation are probably the most likely to be the biggest time sinks. On the plus side, Reddit showed that people are willing to to subreddit-level moderation on a volunteer basis. Frankly, Reddit was getting a lot of free labor that way, but it proves people are willing to do it. A server that fails to adequately moderate itself can end up being de-federated and if illegal content is posted, shut down by law enforcement. And short of that, failures in moderation can result in a poor experience.
Thus, at the server-level, I think some kind of hierarchical governance structure will eventually be needed for the Kbin and others to avoid that. Say that every magazine can vote for its moderators. Then form a random group of 100 mags, and those mods can vote for a super-mod. And then every random group of 100 super-mods can vote for a super-super-mod. Add as many levels as needed to cover the site.
I'm not saying it will be easy, but I think a community-owned, non-profit reddit can be a thing. Yes, at a certain point, it will need staff and handle non-trivial amounts of money. Probably some kind of small non-profit corporation or cooperative can be set up at some point to manage it all. I think it will end up being the case that a few large instances will come to dominate, because that's kind of the nature of things, but I think Wikipedia has shown that this doesn't have to be such a bad thing. As long as there's at least a few big instances, no single one can go rogue.
@anthropomorphist Donate if you can. When I joined mastodon it took me a little to get my bearings, then I found the instance I wanted to stay on, liked the group running it, so I setup a monthly $2 patreon donation. It's not much, but if enough people do it, then it's more than enough to keep the server up and running.
Eventually as lemmy/kbin settle down and I find my home instance, I'll add another monthly donation to that to.
Yes? I don’t have anything to support this speculation, but I think donations from people who want a quality, ad-free service that they can have direct input into (at the instance level), will sustain itself.
And the developers can also do ok on donations (I hope)
And the amount of money needed is hopefully a lot less because the goal isn’t to repay venture capitalists and mint new billionaires.
Realistically, the problem from the get go with Reddit is that it's a company. They have staff, they have managers/directors/c-levels for their staff, they have marketing, they have HR, they have a board, etc etc etc.
These costs quickly dwarf even hefty infrastructure.
Any "company" that builds a product will inevitably follow this trend. You either find a way to milk your customers for all their worth, or you fold. Venture Capitol and lenders eventually lose their patience.
Fediverse is different because it's a passion project by a handful of programmers and sysadmins that are volunteering their time and in many cases money. The overhead is basically nil, the only real costs are time (which people are voluntary donating) and hardware.
Still, it seems like user numbers are exploding right now and servers don’t run on love and good wishes. I’m sure we’ll see some sort of monetisation at some point because free time of qualified people is limited
The structure of the fediverse won't allow for pervasive and consistent monetization. If a server starts running ads, people can move to another server. If a service (e.g. Lemmy, kbin) start running ads, people can switch services or fork a new service from an existing code base. The mobility it allows for let's people move freely but still connect with content across the fediverse.
The structure is easier for the masses to support it through computing resources (e.g. start your own fediverse server), crowdsourced efforts (software development and server admin/moderation), and crowdsourced funding.
Unless people support the advertising, which is absolutely possible.
I've gotten into several products primarily because of their sponsorship of League of Legends. I started going to Jersey Mike's subs because of their sponsorship of the Team Liquid League of Legends team (and their subs for dubs deals). Now I'm a regular customer there. I bought an XPS laptop in part to support Alienware sponsorships.
I would say sponsorships and on-website advertistments are very different from each other. Sponsorships are agreed on by both sides and made together, website advertisements are just random banners that show up everywhere.
exactly my thinking. I'm happy to donate something monthly and i do that on patreon for some creators but i don't know how reliable this option is for long term esp for a massive cost like development and servers
The fediverse has already been running for years now with a TON of users. I don't think you need to worry that hard. As the number of users grows, the number of instances grows, and the load is spread out.
The cost, as others have said, is staffing. Most instances are run by one or two people.
Most people (in the developed world at least) have very high speed connections now. If you couple that with a solution like Yunohost which you can easily install on a Pi or any old (or new) PC, you can then run a huge variety of solutions. I got a 4 year old PC base unit with 2TB HDD and 16GB RAM off ebay for less than £80, put Yunohost on it and now have the option to run a massive variety of Fediverse software including Mastodon, Lemmy etc etc. The cost to me is literally the PC. I'm already paying my ISP and all the software is FLOSS.
I think this is an option more people can take up.
Then the instances will have to become far more specific. At this point every single instance tries to replace the whole of Reddit. Otherwise the fragmentation will kill it.
I've not dug too far into it, but I'd like to see more ability to aggregate and replicate existing instances but I'm not sure that gels too well with the model in general.
In my head I'd like a thing where if I do choose I could spin up an instance, federate with lemmy.world and kbin, add a gaming community and have that be aggregation of the lemmy.world, kbin and my own gaming communities.
Fwiw I'd also like to have a single user entity that propagates/can migrate seamlessly across all instances. I guess the model would be personal federation?
Maybe some kind mutli associated with my account that I can migrate across instances would solve one half of the problem.
The other side is replication, would allow local caching in my instance, or, if lemmy.world were to go rogue or shut down then that history would be preserved and on defederation I could keep the history.
I know. However, I do not believe that reliance on the good will of the users will keep a fast growing platform running for long without there being large, uncovered costs for the operators.
Add comment