12 Years of Ghost  ## Metadata - Author: [[YouTube]] - Full Title: 12 Years of Ghost - Category: #articles - Document Tags: [[video-notes]] [[youtube]] - Summary: 12 Years of Ghost - URL: https://www.youtube.com/watch?v=jmGVS5apLps John O'Nolan: Hello, hello. It's uh, it's lovely to be here. Um, first time back in the UK for a while. Uh, First time at a WordPress conference, in fact, since the last Loop Comp, uh, in 2015 in Las Vegas, which was very exciting. Uh, It was unfortunately my second time in Las Vegas at a tech event of some description where at the end of the event I said, right, everyone, ladies and gentlemen, we're young, we're free, we're single. We're in Las Vegas. What should we do? And they all said, we're gonna go home and work on a bug. So maybe third time lucky, I don't know. still haven't had that kind of exciting Las Vegas experience. But here we are. So my name is John O'Nolan. I'm the co-founder and CEO of Ghost. in fact, the other co-founder and CTO is on her way here. She's on a train, I believe right now, but maybe at some point the door will open before I'm done and she'll step through. how many people have heard of, used Ghost? familiarity. Okay, cool. That's like a solid 70%. if you're in the group that haven't, don't worry. We're gonna, we're gonna cover some of that today. if you're in the group that have you might be wondering, what's this guy doing here? Which is a good question. I'm wondering the same thing myself. Fortunately, Brian kind of alluded to some of the reasons why they asked me to come here today, and I'm very happy to be here. In 2015, there were in fact two open source conferences I spoke at. One was that LoopConf and another was strangely enough, a DrupalCon that took place in Vienna. And at the time, Dries was running an initiative, which he called, Getting Off the Island, and he had observed how the Drupal community had become relatively insular in its thinking. and was evolving, but was not necessarily paying attention to the way in which the rest of the industry was moving, the way in which web application development was changing. And so he encouraged all core contributors to Drupal to get off the island, go and explore other open source projects and ideas and bring back what they learned to the Drupal projects so that it might benefit from them. And so I was invited as a part of that to come and talk about, in fact, WordPress and the idea of Ghost, which was just getting going and it went very well. So I think there'll be some parallels here. So there's this wonderful post, some of you might remember on a site called, Wait, but Why, which sort of maps out the journeys of life that you go through. And when I say life as a fork, it's sort of representative of all these different paths that we take in life and a fork. I think in this room we would think of as software development. And in some ways this looks a bit like a set of GIT branches, but it's really any decision that you might take. life is just this meme over and over again until you die, right? It's choices that you make about which direction to go, and sometimes the good versus the bad path, it's clear, and sometimes you can't quite see over that hill, but you make a choice about a direction to go. And so every decision that we make, all of us collectively in life has taken us up to this moment, this point in time, this shared space that we're in right now. And we all know the path that we've taken. But there are all these other paths that are no longer open to us, right? That the decisions we could have made differently, a different direction we could have gone in. And so, well, all of us here sort of know the WordPress path. a fun exercise that you don't often get to do. Is to explore what one of the other paths could have been. you know, it's all very theoretical. If I'd studied differently or harder in school, could I have ended up in a different career? Could we all be doctors perhaps if we'd made one choice differently? It's all theoretical, but you don't often get to see what the reality of another path might be. and so that's what I'm going to try and sort of do over the next half an hour or so is talk you through where Ghost has come from and ended up starting from a similar origin point from within WordPress, in fact, but then splitting off and forking into a different direction, not technically, but philosophically. So my background I was on the WordPress core team. I was part of the UI group within the sort of circle of people who were doing those things at the time. I was in the, the outer circle, if you will. And when I first touched WordPress it was early days. It was really early days. It was, I was getting into web development. A new bit of PHP and I got FTPs, my HTML onto a server. But I didn't really understand what a CMS was, but this was the first time I saw WordPress, which was pretty early after the original B2 fork. And when I really got into it it looked like this 2.3. How many of you remember back 2.3? Yeah. Old timers in the room. That's my, that's my boys and girls. and then of course we had the Happy Cog redesign with Jason Santa Maria and things changed a bit. And then as we got to around the sort of 2.7 era it was when I started getting involved in, in contributing, this was called the Crazy Horse redesign. And then it took us through to 3.0 and 3.1, which I was a part of. that helps refine the UI from there. And. I loved it. It was wonderful. I got into contributing to WordPress 'cause I was a freelance web developer and I thought this would be good for business. Like if I'm a contributor to WordPress, I could probably increase my freelance rates. but along the way, what I didn't expect to happen was that I became completely enamored by the story of open source and why it exists. And what happens. There's this formative essay, 1997 by Eric Raymond about the cathedral and the bazaar, comparing the meticulous centralized waterfall development of Microsoft and windows to the chaotic, decentralized, iterative collaboration process of Linux. And how is it that you can take billions of dollars and tens of thousands of developers collaborating in presumably a very organized way and have them be competed with by a group of people all over the world who've never met each other successfully? And this philosophy of collaboration over competition and building infrastructure where we would have shared software rather than endlessly rebuilding the same things over and over again was something that spoke deeply to my hearts and the idea of doing it to democratize publishing even more so. And so I went from being an open source pragmatist to an open source evangelist. I spoke at a lot of WordPress events. I contributed a lot of my time to core, and a lot of it was great. It really instilled a lot of motivation and passion within me. and at the same time some other things weren't that great and they kind of bothered me and didn't sit quite right. Something something about it just felt weird. It had like a, an off kind of taste and I couldn't figure out what that was or why that was happening. But there were certain dynamics that meant decisions that were getting made were often not about what was best for the user. Not necessarily about what was best for the product, but perhaps what was best for business interests. There were conflicts of interest around how we did things within core that over time really grated on me. They really detracted from the idea that I had kind of bought into when I signed up for contributing to democratized publishing. And so that was, that was difficult and. It was a nagging feeling in the back of my head. And so what's kind of funny we'll show some things I've never shown before. So this is exciting, is I've found, I've tracked down the exact moment when that little pink branch of Ghost forked off as an idea from that original green branch of WordPress. and it was in a DM with Chelsea Otan, who was my kind of co collaborator, co-conspirator on the WordPress UI team. And I said to her, one fine morning in August 14th, 2012, which I can read easily. I'm millimeters away from starting a new project, forking WordPress. She said, step forward presumably by more than one millimeter. And I said, open source, subtle, kind of, or WordPress lite, publishing focused, not CMS focused. The premise would be that WordPress is now a CMS and I want a blogging slash publishing platform, all the great parts of WP for blogging. It's much easier to read here, minus all the extras. And she said, lots of people talk about it. No one has done it yet. Or perhaps the subtext, no one ever does it because that's really what was the case and to some extent is still the case at the time. And that little, like, that little jab of like, yeah, lots of people say it and no one does it. Was a bit of a motivator for me, and it kind of got the wheels turning. And an hour later I said, you know, the more notes I make, the more I want this to exist. Oh God, this is a slippery slope. And let me tell you, 12 years later. It was a very slippery slope. and what's funny is I actually took a picture of these notes, which is not something I usually do, and I don't think I even sent this picture to her. But this is the notes from that day in the notebook on my lap scribbled under the heading WP Lite of the things I liked, the things I thought were important, the things that were bothering me. the genesis, the beginning, the seed of an idea. And so the question really then is why for any sort of fork or any significant decision in life or in business or in open source, why, why do this thing? And as I said, the, the mission that I'd really bought into that was exciting to me was democratizing publishing. That was the thing that I had a clear idea of, of what it meant and what it should mean. and that was different to what I had kind of discovered the, the reality to be. And there were some parts of WordPress that I loved and wanted to carry forward and thought should be exactly the same. And there were other parts that I didn't love. I loved there was open source, decentralized, customizable, that it enabled people to do all these things without kind of taking credit. but I didn't love the new product direction, which was by all accounts first sort of for the first time being described as going in the direction of becoming a web operating system. Not say there's anything wrong with that, that's just not what excited me. I was more excited about the publishing side. the conflicts of interest I touched on bothered me a lot. There was a lot of kind of cloak and dagger stuff around that. There's this confusing business model of webpress.org, which is opensource projects and webpress.com, which like social network be can't have plugins, you can't have themes. So I, I'm not sure why you'd use that. And then there were these hosting companies which are third parties, but they didn't sustain the open source project. So like how does. Every time there was a tech crunch story, they'd sort of say WordPress, when they meant Automattic. It was all very confusing and conflicted. And then there was you know, core UX bloat. Things were getting a little messy. Despite my best efforts in the core UI team stack stasis you know, PHP 5.1 is the minimum today still. MySQL five as the minimum. that was sort of holding things back, subversion still today while the world was moving to GitHub and PHP was actually evolving quite rapidly, but WordPress wasn't and the GPL and look I'm not gonna start a licensing debate here. There's reasons for and against every license, but the GPL. In my experience, I had witnessed more as a tool to take people's freedom away than a tool to grant people more freedom, and it has never been something that has felt like a very powerful idea to me to say, you can have this code and do anything you want with it. With conditions feels like a less powerful idea to me than some alternative models. So what could be better about this? That was kind of what kept me up at night and I had a handful of answers. I wanted to do something that was focused on publishing. A single use case rather than trying to compete with Squarespace and Shopify and Medium and Tumblr simultaneously focus on one core demographic, have a transparent, straightforward business model with sustainable revenue. Simplify the design. Adopt a modern tech stack and have a just a dead simple. No need to explain it. License. Here's the code. Do what you want with it. You don't have to do the same thing as me if you don't want to, but here's the reasons that I did. So that was the principles behind the idea. And I encoded this into a blog post in 2012 that sort of laid this out as a product pitch and almost for no other reason than to get the idea out of my head. I didn't really know if this would be something that resonated or would catch on. It was like, well, I've just been thinking about this for such a long time. I'm just gonna at least blog about it, and if no one cares, great, I've put the idea out into the world. And alongside the product pitch of what a new sort of lite publishing focused version of WordPress might look like. I had a clear idea for a business model. We'd have a single nonprofit foundation that would employ developers to build great open source software. If the software was good enough, it would attract users and those users would need hosting. And so we would offer that hosting for that software, and all of the revenue from the hosting would fund the nonprofit foundation, which would hire more developers, which would make more great software. And so the cycle would continue. Very straightforward. Nothing revolutionary. In fact, I found an old slide deck from 2014 or so where I described the business model as being software like WordPress, hosting like WP Engine, and at some point a marketplace like the App Store, goes back a long way. But the blog post did take off, it hit the front page of Hacker News. It immediately went viral to some extent. And following that kind of engagement and interest that turned into a Kickstarter campaign about six months later. This is what I looked like then. So young, so naive. Oh, the things you'll see. Do you ever think about, like, when you see an old picture of yourself, I was 23. Like what would you, what would you whisper in your, what would you tell yourself as like a bit of light, like, stay away from redheads. You know, like what would really help you just overcome some hardship. and the Kickstarter campaign did incredibly well. Over the course of a 29 day campaign we raised $350,000 in pre-orders for the hosting service. And so that was a way to sort of kickstart that flywheel to get going, to get that cycle going. Initial injection of revenue, pre-orders for the hosting to fund the foundation, to fund the initial build of the open source software that would then create more demand for the hosting. And at the time, it was the third most funded project on Kickstarter in the open source category. Out of curiosity, anyone here fund the Kickstarter campaign? Oh wow. Thank you. You too wouldn't be here without you quite literally. and about three months after that, we released our first version, which was, Hmm. It was something, it was very, very minimal. you know, the old adage about how you should be ashamed of the first version you shipped, and if you're not, you waited too long. Mm. Check that box. It was it was basically just this very simple markdown editor. a list of posts, a setting screen where you could change your title and description and you could invite staff, users, people to edit. And that was about it. It didn't do very much, but it took us from being an idea in theory, vaporware for want of a better word, to something real, something that exists in the world, something that people were using. And so that was the initial. That was the initial branch, right? That was the initial fork, the initial separation, philosophically into something new, putting something new into the world and along the way, there have been other twists and turns that have taken us in a different direction, quite intentionally. we could have forked off and made all the same decisions, but that would've taken us to roughly the same place, and we were hoping to go somewhere new. So the other twists and turns of what I kind of refer to as the unusual forks, I think along the way of making decisions about any products or perhaps in life in general, there's the expected decisions that are made, the usual things the things where you tell people. And so then we did this, and they go, eh, that makes sense. And there's the unusual forks. There's the things we tell people. And so then we did this. And they go, well, what, why the fuck would you do that? And we got a handful of those and they're the more interesting parts of what has kind of contributed to how we think about open source, how we think about software development and this is decisions we've made along the way. So the first obviously is the non-profit structure. and it's funny, perhaps less so in this room, but in most like startup or business communities, whenever I say I run a non-profit it's an overwhelming reaction of, why, like, why would you do that? Don't you like money, kind of thing. which says something about the world perhaps that we live in. But it's a confusing term. Nonprofit is a confusing term because over Ghost's 12 year history, we've been profitable every single year. We have lots of revenue, so what the hell is a nonprofit company? Well, nonprofits doesn't mean that the company doesn't make a profit. It means the company doesn't have owners and therefore the owners don't make a profit. A nonprofit organization keeps all of its revenue within the company and reinvests it towards the mission. There are no shareholders in place of shareholders there are trustees. So trustees steer the company, but they don't have anything to sell. They don't own any piece of the company. The company is guaranteed to be independent forever, and all of its revenue is reinvested towards its mission. So why would you want one of those? Why would you not want a company that you can just have some upside, some benefit from? Well. Certain companies you would hope put purpose above profit because the purpose is actually more important than the wealth of the owners. And there are some really obvious, straightforward examples where you might hope a company would put purpose before profit, say a hospital, a school, perhaps a news organization. And indeed I think an open source project. These are all places where if profits come first, if profits come before purpose it has consequences or side effects that are often not good for people, for society, or for things that we care deeply about. So the nonprofit structure was really key to me to engender trust and transparency behind a project. It felt like the right way to do things. And to make it a transparent and sustainable business where it has a very straightforward revenue stream, we share our revenue. And that's perhaps not so unusual today, but back in 2013, it was, I think it was just Buffer and maybe a couple of other small startups that had had this idea of just saying like, here's how much money we make, here's how it's going and giving a window into the operations of business again to engender some type of trust. And we decided to build a JavaScript. very unusual, particularly in 2013, Node JS was at version 0.10 and we sort of started building this editor, and that was in JavaScripts and then we wanted to put it on a server and store some posts. So we reached for Node JS and as a quick prototype and and we kind of went, what if we just did the whole thing in JavaScript? It could be interesting and there were significant pros and cons to that decision, particularly then. I think the obvious pros were it was new and exciting and trendy, so it attracted a lot of interest to the project that it might not have had we gone with a more traditional stack. There have also been an enormous number of drawbacks for the last 10 years or so people haven't really built web applications with node. They've built microservices, performant APIs, all kinds of stuff, but not things that are necessarily the shape of what we might now do with a Laravel app or a Rails app where you can reach for a dependency off the shelf and solve a whole host of problems that someone else has already solved for you. so upsides and downsides, but ultimately probably a good decision and then finally, and these last two are kind of come together. We've got a headcount cap of 50 ish for the foundation. We're never gonna hire more than roughly 50 people. And there's no concept of certainly no desire to have a benevolent dictator for life of any description. And there's there's a really clear reason for that. Companies sort of always start out the same way. This, when you're starting a company, it's like this cute. Cute little thing and you're just getting it going and you're like, oh, I just wanna grow this thing and see it become successful. And you know, so you hire a few people and they grow a little bit. but the thing about companies and organizations in general is they're hungry little fucks. And by and large, as they get bigger, suddenly they start eating their brothers and sisters. And we just go, well, it's working. And think about how much more we could do if we were bigger. And before you know, it. They kind of consume everything around them until at some point everything is owned by the same company and we're all just trying to figure out who the guy is that's actually in charge of all of this stuff. And my firm belief is that there is such thing as too much growth. But when you are hopped up on venture capital steroids, you end up with this sort of form of economic dysmorphia where you go just a little more growth, bro. Just a little more growth. We could just be a little bigger. So what if, what if companies didn't look like that? What, what if you tried not to grow as big as possible? What if infinite growth and becoming huge wasn't the goal? And that's what we've tried to do with our structure overall. So we have this core foundation with the idea of a limit to its growth will never be more than about 50 ish people. And so, rather than trying to grow a giant company that we control. We're trying to grow a giant ecosystem that we support. So within that, there will be hosts, developers, agencies, partners, publishers people who are building on top of shared infrastructure. And by limiting the sort of core size of the foundation at the center, we're trying to create opportunities within the ecosystem for lots of companies to thrive, for lots of companies, to be economic dependence of the project, to be invested in its future. And our role is helping the collective thrive rather than trying to capture all of the value within the ecosystem itself. And so in the physical world, you or the way I kind of think about it is, you know, which version of the world is more appealing to live in? Is it the one enormous everything store where you get all things under the sun from a single brand? And is that desirable? There's trade offs. I think where I would, the sort of view of the world I prefer is one where there's lots of smaller independent stores. There's lots of different vendors within a community. And so thinking about the ecosystem more as stewarding a community or perhaps a city or something that is not all controlled by one person but is a collaboration between lots of people is the view that we kind of take of it. It sounds nice, right? Sounds like kind of a bit of a utopian ideal. So how, how's it actually going? Is this working or is this all theoretical? Well over 12, 13 ish years now, we've come a long way from that first 0.3 version with the side by side editor and not much else. Today, Ghost is a relatively full fledged publishing platform and we've leaned in hard to the publishing and journalism use cases. we have native analytics built in. You don't need third party services, and that's deeply integrated into the products. We have email newsletters, so when you publish a blog posts, it goes onto the web, it goes out to RSS. You can also deliver it to people's inboxes by email and that's integrated with a membership system so that publishers can have people sign up, log in, indeed, pay to access content, which has brought the business model of SaaS and subscription software that we've enjoyed, and that's worked so well to tech, to journalists, creators, and publishers in a way that is breathing new life into news and into journalism, and giving them a sustainable business model that has been missing for a very long time. And perhaps most excitingly of all, and kind of most recently of all. Ghost is now connected to the Activity Pub Network, so decentralized federated content, which means for the first time we're kind of getting this 2008 era blogosphere back where you have these independent publishers all over the world that can now follow each other communicate with one another. And it's not limited to one platform. It's not just Ghost to Ghost. it's Ghost to WordPress as well. This is a WordPress post streaming into Ghost. And in fact, we've done more work with the WordPress team in the last year than ever before. And Mathias and his team are doing incredible work in the activity pub plugin, which is moving quickly. Like if you tried it a while ago and you were like it's cool, but try it again. They're doing tons of work and it's very, very cool to feel that kind of nostalgia of the old Open web that I think, I assume a lot of us here will remember fondly. so it's come a long way and there's a bunch of people using it. The ones we're most proud of and what has where the products has really resonated the most is the people who care about the use case that we've leaned into, right? So we have lots of media and news organizations a lot of local news, which is some of the use cases we're most excited about in particular in this part of the world. we've got the Manchester Mill. the Glasgow Bell, the Liverpool Post, the Sheffield Tribune, the Birmingham Dispatch, and in fact, in the capital itself, the Londoner, all running on Ghost, all making sustainable revenue, all operating on open source software that is helping them stay alive, and in fact, in many cases, helping them thrive. Outside of that, many of the startup blogs that you might come across or read in your day to day including Y Combinator itself, that's probably the biggest one. And OpenAI as well. In fact OpenAI started as a research, an AI research publication. and up until about GPT-4 We hosted, unfortunately it ran on Ghost and we also hosted openai.com which was interesting. we got to the point where their traffic per hour I think was greater than all other Ghost sites combined per week and then thank goodness they were like we think we might have outgrown this publishing platform. We're gonna move to something else. And we're like, please. Yes, that sounds good. Um. But way back machine has some, they had a like beautiful site. They had the. Oh, his name's escaping me. The original, like founding Stripe designer was the founding designer at OpenAI. So there's beautiful like AI research publication that was incredible. and a small but growing number of the creator economy, right? YouTubers, podcasters, people who are creating content on the internet who don't necessarily identify as journalists or perhaps media but nevertheless are professional publishers in some capacity. And so today, starting from Humble Roots is just me and Hannah getting going and getting things up and running. We're now a team of 35 almost $9 million a year in annual recurring revenue and over the lifespan of Ghost we've managed to earn, what does this say, $37 million in total through the nonprofit org. No outside funding. no investors, no additional stakeholders and more impressively in my mind at least $107 million for other people. $107 million has been earned by independent publishers who use Ghost. And I dunno if there's many open source projects entirely bootstraps, that have managed to generate something like $140 million in revenue. So that's the one I feel the most proud of, in particular, that we've made three times more for others than for ourselves, which was kind of the point. That's what democratizing publishing means to me, is giving more powers, more power to others than you do to yourself. That's the sort of original Bill Gates definition of a platform. And the economic value of the people who use it exceeds that of the platform itself. And so this winding path has taken us in lots of interesting directions based on some of these strange forks we've made along the way. but those of you who've seen this post will know that this is only one half of the diagram. And the second half of the diagram is looking forwards rather than backwards. And while the black paths that are the road not taken are unavailable to you, there is of course an infant number of future forks that you can take as you consider which direction it is that you want to go in next. And so every decision you make about an open source projects a business, um. Something you work on is a sort of a game of, of choose your own adventure. it's a game of looking at the world and deciding which way to go in, which might be not the same way as everyone else has been going in for a while. And so the thought, or at least the question that I want to leave you with is, what could be better? It's very easy to some extent to pick out, um. What's wrong with any system, whether it's WordPress, whether it's a friendship, whether it's a marriage any sort of dynamic that you find yourself within, it's easy to find the faults. It's easy to point to the things that are wrong, but what would be better? one thing I see now nowadays from an outside perspective in the WordPress community is a lot of dissatisfaction with the status quo to some extent and in plenty of different areas and rightfully so. There's a lot to not be happy about as well as all the things that there are to be happy about, but it is all too easy to criticize and it's very difficult sometimes to come up with something better. And the act of coming up with something better requires you to move into the new head space of creativity. And creativity, of course, is the hardest thing to do of all because it opens you up to criticism, but it is how all forks begin. Whether it is philosophical, whether it is technical. They begin with asking What could be better? If not this, then what? And that is the question I would encourage you to ask of yourselves. Ask of each other. Ask within the context of your life. Ask especially within the context of WordPress. What could be better? There are tons of options out there, and whether you look to something like Ghosts, something like Drupal, some of the many other incredible open source projects in the world, there are ideas out there. What ideas can you find and bring back to WordPress, I think is the question that I would love you to take away from here and perhaps through the next rest of the day and the rest of our speakers. now it turns out that I've spoken incredibly quickly. if you wanna find my details there on this QR code. You can get in touch with me, but it looks like we do in fact have time for perhaps two or three questions. So if anyone wants to have a chat. Brian Richards: Ooh. bold, move,. John O'Nolan: Throw, throw a few things at me. we can, we can do that for a minute. Yeah, sure. Brian Richards: Yes. First, thank you, John. I saw this hand first, so I'm gonna give you the microphone first. Question: Alright. I guess, how did you settle on 50? What sort of guided that decision and why is it the magic number? John O'Nolan: Yeah. totally arbitrary, honestly. To some extent there's sort of like bits and pieces. we've got a bunch of people who work at Ghost who've come from larger companies. I've got lots of friends who've worked at larger companies or been at a company as it's gone from small to large. And the story that I've heard the most often is that things were great and then we got to around and it's usually like 60 is what I hear. 60, 70. And then like a new layer of management came in. It started being hard to know everyone's name. things changed, right? It went from this like ragtag group of people working together on something like a tight knit community to now being a job, now being just like a thing. and the vibe of that small community working hard together went away. So it's sort of like just under that is the idea and, um. Aside from the kind of ecosystem reasons to do it, there's also like a, just a very deeply selfish reason, which is, I don't want to run a big company. It doesn't seem like fun. I know people who do it. I talk to them about what it's like, the things they say, don't make me think I wish I was doing that. You know it sounds very stressful. 35 people is quite stressful. 50 feels like the upper limit. but it is arbitrary. I think you could pick any number. and you know what, if I was going to do it again and start another company, I would pick an even lower number. I think I really like it as a, as a forcing factor for sorts of things, and a lot of these things can be really interesting, forcing factors. If you only have 50 people. Then how do you approach this problem? I think those constraints often breed creativity in a way that you might not otherwise. if the answer to all your problems is we'll just hire more people sometimes that means you don't think deeply about the problems. does that answer your question? Kind of a roundabout. Brian Richards: Yeah. I, I love creative constraints. They breed all kinds of interesting solutions. Okay. I saw a couple more hands. We have time for two more questions if anybody still has them. Yes. Perfect. Here's a microphone for you. Question: I'm curious. You you use we a lot and company, and you showed us a lot of the, the sites. How many actual sites are self-hosted versus running on Ghost hosting? 'cause I don't think I've ever come across a Ghost site in the wild that hasn't been basically on Ghost's own hosting platform. John O'Nolan: Yeah. So we host, I think it's about 27,000 sites. It might be slightly more Hannah, 29,000 maybe. I think there's there's at least the same amount within the kind of Alexa top million that are self-hosted. And there's, I think, the same amount again in the sort of longer tail. And then there's obviously an unknown amount beyond that. Ones that are private within internal networks or in some way, like not accessible to built with or w three techs or any of these other things. and it fluctuates, but somewhere, anywhere between like a hundred to 150,000 sites in the world. So we, we host the minority. There's a lot of self hosters. It's not as easy to self-host Ghost as it is WordPress, which is really just a way of saying it's not as easy to self-host Node as it is PHP. and so that's a constant challenge is how do we make it easier to self-host. But it's a fun one because to come back to the last question, if you have a 50 headcount cap at a certain point, you can have too many customers. At a certain point you can go, look, we can't hire more support staff. We can't hire any more infrastructure staff. We have to, we are incentivized to make it easier to self-host. In fact, we need hosting companies to support Ghost and customers to be able to self-host Ghost easily. Otherwise, the project's not going to succeed because our hosting is limited in the number of customers we can support. So I expect that ratio to actually get bigger. I expect to host a vast minority compared to a relative minority in the next 10 years. Brian Richards: Amazing. Well, if anyone else has a question for John, keep it to yourself. It's too late. No John's gonna be hanging around for a little while. You can find him here and you can ask him more questions. but for now, we need to start changing over. So give John another round of applause. This is amazing., 12 Years of Ghost, YouTube, 01k9avqkpf2yx7464x69ttk4v9.