Replies: 6 comments 14 replies
-
The good news is that with our server requirements, I think we might be able to handle thousands of users, even on a tiny $5/month server. The free plan has a hard-cap on the amount of data stored per user ( if we put reasonable link and bio size limits ), too, so while we'll still have to monitor the impact of free accounts on the server, and how much bandwidth we're using, I think we'll be able to support many free accounts without having to worry too much about costs. |
Beta Was this translation helpful? Give feedback.
-
|
This is a very interesting business model. I love the idea, but I wonder if you'll be able to support enough free users in the long run. I think typical ratios for freemium services are much higher than 30/70 or even 10/90. Definitely worth trying though. |
Beta Was this translation helpful? Give feedback.
-
|
When free is the default, the ratio is indeed more like 1% to 0.1% of users. We’re not doing the same as e.g. Discord Nitro though, in that the only thing we’re giving away for free is the ID+website service that costs practically nothing for us to host. That’s more like if basic Discord was read-only. To do just about anything, you’ll need Weird+. But because we don’t want to make the interactive services exclusively pay-to-play, we are going for a 30/70 ratio in the paid plan. That ratio won’t change unexpectedly, because the amount of free Weird+ accounts available is literally hard capped:
|
Beta Was this translation helpful? Give feedback.
-
Alternative payment schemeThe first scheme I came up with looks like this:
What I liked about the tiered unlocks:
But there are two major complications:
Maybe we need to consider an alternative designed to work with annual subscriptions: Weird Nerd 🤓$50/year ($25/yr pre-alpha)
Extras
Rev-share for bundled apps.For as long as we are bundling 3rd party open source apps in our core offering we will pay forward $5 of each yearly sub to those upstream services. That money will be divided based on usage metrics mapped to some flat-fee tiers, meaning a very active Reader app might get $1.50, while infrequently used ones get $0.25 each (which is the minimum payout) and any practically unused services get $0. Rev-share for extrasEvery Extra service add-on simply costs a flat $10/yr, and it all goes directly to the service maintainer. We’ll partner with the maintainers to let them be primarily responsible for that whole service offering. Free sponseesSponsored Writer accounts would initially be reserved for partners and contributors of the Weird project. Even later on they may only be made available to sponsees of higher trust levels. Gotta earn the privilege to speak by first doing your reading & re-sharing. Sponsees do not get any 🆙-stream service apps. (Maybe they could be gifted by sponsors). It's not like anyone would be barred from participating, since there are already plenty of free options to choose between among individual hosts for ActivityPub, Bluesky etc., which can all be added as external integrations to ones Weird profile. What I like about this alternative structure is that it emphasizes Weird's primary value-proposition, which is an interweaving of multiple protocol-services into a cohesive whole. By not being forced to pick between these services individually, we relieve the user of a significant cognitive load at the onboarding stage. Instead they're free to try out different add-ons as they see fit, organically arriving at whichever app composition best suits them. |
Beta Was this translation helpful? Give feedback.
-
|
I think we can have a smaller amount of core apps in each tier, and offer additional services (Weird-hosted apub account etc.) at annual-only prices, like $10 a pop. You’d get a one month long free trial though. (Get two more months for free if you sign up for the subscription ;) ) Rev-share would change with this. Instead, it’d be something like ‘upstream service-apps get 3$ of each $10’. Might make more sense to grant them the full amount 🤔 added a (option B) accordingly. |
Beta Was this translation helpful? Give feedback.
-
Licensing
While the general-purpose utilities of Weird (e.g. our graph database and any prospective protocol layer built on top of it) will be licensed as permissive open source, the Weird app will be Fair Source licensed. https://openpath.chadwhitacre.com/2024/the-historical-case-for-fair-source/ Our version of that will entail the Polyform license plus a countdown addendum: |
Beta Was this translation helpful? Give feedback.
-
That would be a great start. And alongside of that gather feedback of non-core team community members and lay the puzzle of how people may fit in. I'd also like to bring up another point to consider wrt fostering emergence of an ecosystem. The potential of Weird / Leaf depends in part also on the successful introduction of new computing paradigms. First of all in the realm of "local-first apps" and secondly relating to the "open social web". There are risks if the technology base doesn't mature. And then there are big opportunities for the pioneers. Major innovation awaits and there's advantage if you can move fast in a collective with many players, take position, be ahead of the game. And an ecosystem can bring major boost to the attractiveness
Exactly. I see "agentic fediverse" more as a technology vision. What is Muni Town even? Would it be an idea to make a clearer separation between the business entity you represent, and the collective itself including all community members (anyone who is active in the ecosystem, no member registration needed)? This way you could launch a portal focused on the collective adding their Lego blocks, with Leaf at its core. Your business being one of independent initiatives that is part of the collective, having its own web presence for e.g. Weird. The "Leverage the ecosystem" strategy aims to reap maximum mutual benefits from each others efforts, on a continual basis. It might be literally named "Collective Development", having the Muni Town collective as its scope. And it could encompass the idea of Partnering you have. The traditional way is a "Contact Us" and then further dealings in closed partner channels. I feel that is not a good match to foster a vibrant ecosystem, though you might use this specifically for product partners relating directly to Weird. Partnering can be made explicitly transparent. All collaboration & contributions to the ecosystem by members of a collective stir the cauldron of opportunities, and may see new partnership relations emerge. Here again it is emergence that can - and in my opinion should - be fostered via Collective Development. PS. Note that imho thinking of collective members as a "community" is a fallacy, or rather what constitutes community must be well defined. The collective members operate within grassroots movements with unique social dynamics, where they rather individualistically pursue own dreams. Active participation is based on concrete incentives to collaborate and intrinsic motivation. |
Beta Was this translation helpful? Give feedback.
-
|
Great conversation so far, and I’ll be adding to it shortly. I just need my busted-laptop replacement to arrive! |
Beta Was this translation helpful? Give feedback.
-
I think I like this idea.
I agree that agentic fediverse is a technology vision. I'm not sure whether Muni Town is the business or the collective, if we want to draw a line there. Personally I feel like Muni Town is closer to the collective. I feel like "Weird" is the business. Muni Town is like everybody that has kind of come together to respectfully share thoughts, collaborate, or just just hang out and watch what's happening. Muni Town is all of the people that have kind of centered around our overlapping visions of some sort.
I have heard others who have more specific ideas of what "community" means than I tend to have when I say it. I'm not sure what is most common, but when I say "community" I usually mean anybody who ends up visibly involved I can't remember who said it, but it made sense to me when somebody said that you can see somebody anywhere related to your project, whether because they joined the chat room or commented on a GitHub issue or whatever, then they're part of the community. Anyway, definitely appreciating your perspective @aschrijver! |
Beta Was this translation helpful? Give feedback.
-
|
Yes, the Muni Town you describe is how I thought of it too. As for community. It is probably best to always think of "community-in-context". Sometimes you can be handwavy (e.g. loosely refer to "the FOSS community") and sometimes you must define very explicitly. For instance in communal software, where Community is a core business domain. You might use your definition of community for Muni Town collective. And set your expectations accordingly. The definition is very loose. From such definition results that you can't expect anything from the community. It includes any passers-by as they come and leave and there are no commitments. I think that corresponds effectively to a community of interest. I would simply refer to this as the "collective" and then reserve "community" for more focused areas. For instance for Leaf you might want to have a dedicated community of action to evolve the protocol within the ecosystem. And you want a self-sustained community where tasks are properly delegated and people also do the chores. The actual terminology to refer this community may be different, though (might be "working group" within the collective, idk). |
Beta Was this translation helpful? Give feedback.
-
|
I found Anytype as a local-first p2p product with a similar licensing approach to Weird/Leaf. The company looks to be a cooperative. See this HN comment and the mentioned page Our Open Philosophy. |
Beta Was this translation helpful? Give feedback.
-
|
From @erlend-sh on Discord regarding charging for top-level names, i.e.
|
Beta Was this translation helpful? Give feedback.
-
|
What do y’all think of this offer? Buy now ($25) 💵 get two subsequent years for free 🆓 🆓 It’s like a ‘lifetime subscription’ middle-ground, available only to earliest adopters. We need to be signing up 100 additional yearly subscribers per month anyway for baseline sustainability, so it doesn’t make much of a difference to our long term financials if the first few batches (months) of paying subscribers get a much sweetened deal. These early adopters are so early that we will also make it clear for them: ‘our existence past the next 6 months isn’t yet guaranteed though, fyi 🤷♂️ ‘ |
Beta Was this translation helpful? Give feedback.
-
|
That sounds reasonable to me. 👍 |
Beta Was this translation helpful? Give feedback.
-
Latest draft here.
There will be two tiers of Weird service.
plain Weird
..will always be free. It consists of two things:
However, one can only attain membership by invitation from another member. This is to ensure a baseline of user authenticity. If someone invites a member that turns out to be a spambot, the inviter will also be heavily scrutinized. Paying for an account is the easiest way to authenticate the member as well-intentioned, so this invitation system serves as another such vetting system.
Weird plus
Costs $5/month; first 100 customers can pay $25 for a year.
Enables custom domains and includes a core service add-on:
unlock any additional service for $2 per add-on. $1 each when surpassing $11/month.
Revenue share
Every open source service included in the Weird bundle will be offered $0.50 for each paying user of that service in our platform. The mechanics of this split are likely to change as the platform matures and we learn more about our unit economics and margins. We’ll hold ourselves responsible to pay a fair share forward by keeping our revenue metrics transparent.
Subsidizing freemium memberships
Free plans are commercially dangerous. Oftentimes they only make sense for a VC-funded business that can afford to break into a new market at a loss. We're not planning to do any of that.
At the same time, a free service is a more accessible service, so we want to figure out a way to sustainably provide 0$ premium accounts to anyone who'd need one.
The basic plan is to do something like a 30/70 split between paid/free users. Basically the number of paid users we have corresponds with the number of freemium slots we have available; if there aren't enough paying users, registrants for freemium will be waiting in queue.
To begin with it'll probably be more like 10/90 paid/free. It all comes down to sustainable operating margins, and we will be transparent with our community about what those look like for us.
Our model is inspired by the likes of the Waking Up meditation app, which is a highly priced premium app, but they have a simple policy of ‘anyone who emails us directly and asks for it will get a one year subscription, and you can ask anew forever’.
Prior art:
Beta Was this translation helpful? Give feedback.
All reactions