WEBVTT 00:00.000 --> 00:10.000 Underful time to introduce our next speaker, Robert Hodges, who will speak to us on building 00:10.000 --> 00:15.200 a great business on open source without selling your soul, keep your hands up. 00:15.200 --> 00:20.480 Robert Hodges has worked on database systems since 1983, spanning pre-relational technologies 00:20.480 --> 00:24.000 all the way to modern, real-time, analytic databases. 00:24.000 --> 00:29.920 His other technical interests include open source software, distributed systems, and Kubernetes. 00:29.920 --> 00:35.200 Robert is the founder of the open source analytics conference, which has a community table 00:35.200 --> 00:36.840 at Phosphum. 00:36.840 --> 00:41.200 In his day job, he is CEO of Al-Tiniti. 00:41.200 --> 00:47.080 I love it when I say things and pronounce things correctly, and this is the part where I 00:47.080 --> 00:48.080 handy the mic. 00:48.080 --> 00:54.000 And still make noise until it is precisely 1135, so we are please properly to make the video 00:54.000 --> 00:56.000 operations that are like easier. 00:56.000 --> 01:01.000 To the video operation center team, thank you for all that you do for us to bring wood to 01:01.000 --> 01:02.000 no problem. 01:02.000 --> 01:10.000 Robert, thank you very much for the trouble, Robert. 01:10.000 --> 01:11.000 Thank you. 01:11.000 --> 01:15.000 Can you guys hear me okay? 01:15.000 --> 01:16.000 Great. 01:16.000 --> 01:18.000 That means people in the live stream can hear me too. 01:18.000 --> 01:20.000 Yeah, it's wonderful to be here. 01:20.000 --> 01:23.000 I want to start with a huge shout out to Phosphum. 01:23.000 --> 01:26.000 This is just an absolutely awesome conference. 01:26.000 --> 01:29.000 I was here for the first time last year, and it's totally energizing. 01:29.000 --> 01:32.000 If you're into open source, it's a place to be. 01:32.000 --> 01:37.200 I also want to start this talk with a shout out to two people without whom it would not have 01:37.200 --> 01:38.200 been possible. 01:38.200 --> 01:40.200 That's Tatiana Cropenia. 01:40.200 --> 01:41.200 She runs DBVR. 01:41.200 --> 01:44.000 That's one of the companies I'll be talking about. 01:44.000 --> 01:48.000 Also, Peter Zitesf, who's right here, is a friend. 01:48.000 --> 01:54.000 And somebody that I looked up to as, you know, sort of an icon of how to build a successful 01:54.000 --> 01:57.000 open source business will also be talking about his company. 01:57.000 --> 02:02.000 And in fact, what we're going to be talking about is how you build open source businesses 02:02.000 --> 02:05.000 that also give back to their communities. 02:05.000 --> 02:10.000 And I'm going to give you three examples, and then talk about some general business issues. 02:10.000 --> 02:15.000 These are not the only ways to build successful businesses, but they're three that worked. 02:15.000 --> 02:19.000 So I bear that in mind, and let's jump in. 02:19.000 --> 02:21.000 So I've been talking about me and my friends. 02:21.000 --> 02:23.000 Let's talk about you. 02:23.000 --> 02:28.000 So you have a wonderful open source project out on GitHub. 02:28.000 --> 02:31.000 You've got hundreds of contributors. 02:31.000 --> 02:37.000 You have thousands of stars, and the stars are growing by leaps and bounds every day. 02:37.000 --> 02:40.000 You have community members who love you. 02:40.000 --> 02:45.000 Moreover, you have a nice Apache two license where you know about corporate lawyers, right? 02:45.000 --> 02:48.000 They don't feel love the same way that we do. 02:48.000 --> 02:53.000 But when they see the Apache two license, their hearts swell with joy. 02:53.000 --> 03:01.000 So what you have here is a project that is successful and poised to begin maybe building a business. 03:01.000 --> 03:04.000 And you start to think, hey, I can do this for a living. 03:04.000 --> 03:07.000 It's not just a side project, it's not just something fun. 03:07.000 --> 03:13.000 So the first thing you might want to do is go look at other people that had successful projects 03:13.000 --> 03:15.000 and turn them into businesses. 03:15.000 --> 03:17.000 Let's look at a few of them. 03:17.000 --> 03:21.000 So here's some businesses that are probably familiar to all of you. 03:21.000 --> 03:25.000 Some of you may know all of them, certainly many people know Hashty Corp. 03:25.000 --> 03:27.000 I've used their stuff for years. 03:27.000 --> 03:33.000 These are big companies founded by very smart people backed by a lot of money. 03:33.000 --> 03:41.000 And what they wanted to do was build businesses on products that were basically open source. 03:41.000 --> 03:43.000 So how did it go? 03:43.000 --> 03:46.000 Well, not well, actually. 03:46.000 --> 03:49.000 I picked those companies because they're the best of my knowledge. 03:49.000 --> 03:52.000 Not one of them has ever been profitable. 03:52.000 --> 03:59.000 So, and along the way, in their search for profit, they all released. 03:59.000 --> 04:02.000 And I'm not making a moral judgment here. 04:02.000 --> 04:07.000 I think that, as we'll see in this talk, part of the reason why they ended up releasing things, 04:07.000 --> 04:09.000 they're just facing some kind of bad math. 04:09.000 --> 04:14.000 But building a successful project or taking a successful project 04:14.000 --> 04:19.000 and turning it into a successful business is a very hard step and not easy to do. 04:19.000 --> 04:23.000 So if you're going to do it for your project, you actually have to take a step back 04:23.000 --> 04:27.000 and kind of think about how businesses are built in order to do this. 04:27.000 --> 04:32.000 And what I'm going to do for the rest of this talk is answer two questions. 04:32.000 --> 04:38.000 How we can build an open source business that one actually makes money so that it's stable. 04:38.000 --> 04:41.000 It can persist. 04:41.000 --> 04:44.000 And so that you actually make a decent living from it. 04:44.000 --> 04:49.000 And it also gives back to the community that helped build the project in the first place. 04:49.000 --> 04:51.000 So that's what we're going to focus on. 04:51.000 --> 04:55.000 So to begin with, I'm going to talk about some business theory. 04:55.000 --> 04:57.000 And it's really not that bad. 04:57.000 --> 05:03.000 This is about all you need to know to start an open source business. 05:03.000 --> 05:05.000 It fits on one slide. 05:05.000 --> 05:07.000 So here's your project and your project. 05:07.000 --> 05:11.000 You're going to offer some kind of product, some kind of services related to it, 05:11.000 --> 05:14.000 maybe a combination of both, and you will collect money. 05:14.000 --> 05:15.000 That's called revenue. 05:15.000 --> 05:22.000 And what we have here, this little bar graph thing is a profit and loss statement, 05:22.000 --> 05:27.000 which is the way businesses track how what happens to the money it comes in. 05:27.000 --> 05:31.000 So the first thing that happens is say you collect 10 euros, it comes in. 05:31.000 --> 05:36.000 And whatever good or service you're selling, you have to pay to deliver that. 05:36.000 --> 05:40.000 So let's say that it's support and it costs two euros to deliver. 05:40.000 --> 05:42.000 So that's gone away. 05:42.000 --> 05:47.000 The next thing you do is, and you're going to have to pay that every time you sell something. 05:47.000 --> 05:50.000 The next thing you have is just operating expenses. 05:50.000 --> 05:52.000 It's just the overhead of running a business. 05:52.000 --> 05:53.000 You have marketing. 05:53.000 --> 05:54.000 You have sales. 05:54.000 --> 05:56.000 You have development. 05:56.000 --> 05:58.000 You have administrative drones. 05:58.000 --> 05:59.000 People like me. 05:59.000 --> 06:03.000 So that's usually the big in large in medium sized businesses. 06:03.000 --> 06:06.000 That tends to be where most of the money goes. 06:06.000 --> 06:10.000 And then you have this special category, which is something. 06:10.000 --> 06:14.000 Is things like taxes, like interest. 06:14.000 --> 06:16.000 They aren't related to operating the company, 06:16.000 --> 06:21.000 but they are things that they are expenses that you incur particularly as you make money. 06:21.000 --> 06:24.000 One important one is actually stock-based compensation. 06:24.000 --> 06:27.000 If your business hands out stock, that's actually a cost. 06:27.000 --> 06:34.000 And this is the real reason why a company like MongoDB is not truly profitable because they're actually giving away a lot of stock. 06:34.000 --> 06:40.000 And they don't want to count that when they're saying how much money they make. 06:40.000 --> 06:44.000 But you take this stuff out and then whatever's left is profit. 06:44.000 --> 06:48.000 That's what you can actually stick in the bank and you know that you actually made it. 06:48.000 --> 06:49.000 You don't owe it. 06:49.000 --> 06:51.000 You don't have to perform work to get it. 06:51.000 --> 06:52.000 That's that's real money. 06:52.000 --> 06:56.000 You can stick in a bank either spend it or save it for retirement. 06:56.000 --> 06:59.000 So there is one really important number in here. 06:59.000 --> 07:00.000 Our concept. 07:00.000 --> 07:03.000 And that's what we call gross margin. 07:03.000 --> 07:10.000 And this is the percent of your revenue that is left over after you get done paying for the service you sold. 07:10.000 --> 07:11.000 Why is that important? 07:11.000 --> 07:19.000 Well, because it's a way of judging how efficient you are at delivering a product or service. 07:19.000 --> 07:25.000 And we'll use that to compare different ways that businesses can make money. 07:25.000 --> 07:28.000 So let's start with model number one. 07:28.000 --> 07:34.000 So what we're going to do is we have a pure open source project or set of projects. 07:34.000 --> 07:37.000 And we're just going to offer services on top of them. 07:37.000 --> 07:39.000 So the software is completely open source. 07:39.000 --> 07:43.000 You don't try to have any proprietary extensions or anything like that. 07:43.000 --> 07:45.000 And what kind of things can you do? 07:45.000 --> 07:49.000 Well, as there's all kinds of services that you can do around open source. 07:49.000 --> 07:51.000 You could educate people about how to use it. 07:51.000 --> 07:55.000 You know open source projects don't always have the best documentation. 07:55.000 --> 07:58.000 You could develop features for it. 07:58.000 --> 08:00.000 People want to have stuff added to it. 08:00.000 --> 08:04.000 You could do consulting about how to set it up and use it in real applications. 08:04.000 --> 08:11.000 The one that's actually the most interesting for many purposes is what we call enterprise support. 08:11.000 --> 08:14.000 And that is for something like a database. 08:14.000 --> 08:18.000 What you're doing is selling a service that supports the database. 08:18.000 --> 08:20.000 So somebody has a contract with you. 08:20.000 --> 08:23.000 If something breaks, you help them fix it. 08:23.000 --> 08:26.000 If they're deploying a new application, you tell them how to do it. 08:26.000 --> 08:28.000 If they're upgrading someone and so forth. 08:28.000 --> 08:29.000 Why is this good? 08:29.000 --> 08:32.000 Well, because it has very high gross margins. 08:32.000 --> 08:34.000 We're doing it right. 08:34.000 --> 08:41.000 So typically on support, I know in our business, we look for about 80% or more gross margin. 08:41.000 --> 08:44.000 So that means 10 euros comes in. 08:44.000 --> 08:49.000 Two, we take two euros to deliver the service, but the other eight euros go to our company. 08:49.000 --> 08:54.000 That's really good because that means we can actually pay for the other expenses and build a company that makes money. 08:54.000 --> 09:00.000 So one of the best examples of a company in open source that does this really well is Procona. 09:00.000 --> 09:04.000 And Peter, I have a terrible picture of your business. 09:04.000 --> 09:07.000 It's really leaving a lot of detail out. 09:07.000 --> 09:12.000 But there's, in Procona, there's a number of major services that Procona offers. 09:12.000 --> 09:15.000 The 24 by 7 support, that's the enterprise support. 09:15.000 --> 09:19.000 Manage services where they have sort of white-blove operation of systems, 09:19.000 --> 09:22.000 where they'll actually log in and help you run things. 09:22.000 --> 09:25.000 They'll do consulting on specific projects. 09:25.000 --> 09:26.000 They'll do monitoring. 09:26.000 --> 09:29.000 They have a product that is a monitoring service through databases. 09:29.000 --> 09:31.000 They do other stuff as well. 09:31.000 --> 09:36.000 One thing, and they do it for, they started out doing my SQL Peter and Vadim who founded it 09:36.000 --> 09:38.000 or performance experts in my SQL. 09:38.000 --> 09:39.000 That's where they got started. 09:39.000 --> 09:44.000 They since added support for Postgres MongoDB, other databases on the way. 09:44.000 --> 09:46.000 These are all completely open source. 09:46.000 --> 09:52.000 One really interesting thing about the 24 by 7 support and the reason you can charge for it is 09:52.000 --> 09:54.000 it's basically insurance. 09:54.000 --> 09:58.000 So for the most part, as I said, it doesn't cost very much to deliver it. 09:58.000 --> 10:03.000 On average, but when something really breaks, like you're running an e-commerce site 10:03.000 --> 10:08.000 on my SQL, better still, you're running a stock trading system on an open source database. 10:08.000 --> 10:11.000 When that database goes down, the market is down. 10:11.000 --> 10:16.000 That's when people want you to, that's when that becomes a good deal for the customer 10:16.000 --> 10:20.000 because, of course, everybody who's giving support wakes up, helps you get it back online 10:20.000 --> 10:23.000 and then everybody's happy. 10:23.000 --> 10:26.000 So that's why this is a really great model. 10:26.000 --> 10:30.000 And this has actually worked out pretty well for Percon over the years. 10:30.000 --> 10:33.000 They've built a strong business. 10:33.000 --> 10:37.000 And also, they've done, you know, spent quite a bit of time giving back to the community 10:37.000 --> 10:41.000 in so many ways that this is sort of an eye chart. 10:41.000 --> 10:44.000 I'll just point out a couple that have been particularly significant. 10:44.000 --> 10:47.000 How many people read the Percon a blog? 10:47.000 --> 10:49.000 Okay, this is, okay. 10:49.000 --> 10:52.000 So you're my, my SQL users have been around. 10:52.000 --> 10:55.000 If you were my SQL users or 10 years ago, this would be your daily reading. 10:55.000 --> 10:57.000 You'd go see what's on the Percon a blog. 10:57.000 --> 11:02.000 They had just wonderful articles about running the databases that they support. 11:02.000 --> 11:05.000 Another really important thing is, you know, I said that they, 11:05.000 --> 11:07.000 the software is 100% open source. 11:07.000 --> 11:09.000 They have bills of the software. 11:09.000 --> 11:14.000 So they had basically a fork of my SQL that they've maintained for many years. 11:14.000 --> 11:18.000 It has fixes for production systems. 11:18.000 --> 11:24.000 A really important feature that is product that they built was what's called extra backup. 11:24.000 --> 11:27.000 It is backup in the store for my SQL databases. 11:27.000 --> 11:33.000 This is a boring, difficult problem that everybody needs to have solved. 11:33.000 --> 11:38.000 And so by solving this problem and providing it as open source software, 11:38.000 --> 11:44.000 Percon a, you know, delivered a major, you know, sort of a major benefit to the my SQL community. 11:44.000 --> 11:49.000 They also run conferences and have done many other things as well. 11:49.000 --> 11:52.000 So that's model number one. 11:53.000 --> 11:58.000 By the way, I should mention that there are many other companies that have done this in a way, 11:58.000 --> 12:01.000 the most, the best example is Red Hat. 12:01.000 --> 12:05.000 Red Hat, of course, did builds, which they then sold you. 12:05.000 --> 12:08.000 They, you know, basically the ability to receive them in a channel. 12:08.000 --> 12:13.000 But it's a fundamentally a services model on open source software. 12:13.000 --> 12:16.000 So the next model is what we call open core. 12:16.000 --> 12:18.000 And how many people have heard of this model? 12:19.000 --> 12:22.000 Yeah, okay. And you probably have kind of, thank you. 12:22.000 --> 12:24.000 You probably have kind of mixed feelings about it. 12:24.000 --> 12:29.000 But basically the way it works is you have an open source product. 12:29.000 --> 12:33.000 And the way we're going to be looking at it is it's an open source product that's fully usable. 12:33.000 --> 12:35.000 It might not do everything. 12:35.000 --> 12:39.000 But it's something that you can use in actually use on production systems. 12:39.000 --> 12:44.000 And then what you have, and that might have a build, you know, sort of a community addition or something like that. 12:44.000 --> 12:46.000 And then you have proprietary extensions. 12:46.000 --> 12:48.000 They might be in a separate build. 12:48.000 --> 12:50.000 What would those extensions be? 12:50.000 --> 12:56.000 Well, they would be things that are useful to companies that want more than just basic features. 12:56.000 --> 13:01.000 So for example, when you're building a project yourself with open source, 13:01.000 --> 13:04.000 username and password might be sufficient to log into a database. 13:04.000 --> 13:09.000 But if you're actually in a large company, like a bank or something like that, 13:09.000 --> 13:12.000 they may have hundreds of systems that people operate with, 13:12.000 --> 13:14.000 and they need to have single sign up. 13:14.000 --> 13:20.000 So adding that to your product is something that isn't really an open source. 13:20.000 --> 13:22.000 necessarily using open source. 13:22.000 --> 13:25.000 It may actually be built on proprietary software. 13:25.000 --> 13:29.000 Things like integration with Octa or all-thuro. 13:29.000 --> 13:31.000 So that's something you can add. 13:31.000 --> 13:35.000 And in fact, integrations with proprietary systems, 13:35.000 --> 13:38.000 performance monitoring, analytics, HADR. 13:38.000 --> 13:44.000 These are things that typically go into the close part of the product. 13:44.000 --> 13:49.000 If you do this well, it's, again, another kind of offering has very high-growth margin. 13:49.000 --> 13:52.000 So 80% that's kind of typical software. 13:52.000 --> 13:56.000 It can be even higher sometimes if you do it right. 13:56.000 --> 13:58.000 And who's done this well? 13:58.000 --> 14:01.000 Well, one of the best examples I can think of is D-beaver. 14:01.000 --> 14:04.000 Is there anybody in the room who's used D-beaver? 14:04.000 --> 14:05.000 Oh, great. 14:05.000 --> 14:06.000 Okay, that's it. 14:06.000 --> 14:09.000 If you have not, I urge you to go out and try it out. 14:09.000 --> 14:16.000 If you use databases at all, it is a DBA tools, you know, sort of visual database. 14:16.000 --> 14:20.000 Tool that allows you to connect to any open source database. 14:20.000 --> 14:22.000 If you use a community version, which I use, 14:22.000 --> 14:24.000 you can edit data. 14:24.000 --> 14:25.000 You can submit queries. 14:25.000 --> 14:27.000 You can do simple automation. 14:27.000 --> 14:28.000 Edit your schema. 14:28.000 --> 14:30.000 It's really, really handy. 14:30.000 --> 14:33.000 It works great with, with click-outs, which is the database I use. 14:33.000 --> 14:36.000 And then what they've done is, so that is completely open. 14:36.000 --> 14:38.000 I'll show you some stats on the project. 14:38.000 --> 14:40.000 They're, they're actually quite impressive. 14:40.000 --> 14:44.000 But then what they've done is created a D-beaver Pro, which is the closed source version. 14:44.000 --> 14:46.000 It's an enterprise build. 14:46.000 --> 14:47.000 You pay money for that. 14:47.000 --> 14:48.000 It's pretty reasonable. 14:48.000 --> 14:53.000 And they have been incredibly creative at thinking of things that are useful for large businesses. 14:53.000 --> 14:56.000 They range from single sign-on, which I already discussed, 14:56.000 --> 15:00.000 to things like talking to Oracle, Oracle, not an open source database. 15:01.000 --> 15:02.000 No SQL databases. 15:02.000 --> 15:04.000 Dance design tools. 15:04.000 --> 15:06.000 You know, task scheduling, things like that. 15:06.000 --> 15:09.000 So they have a bunch of features that large enterprises 15:09.000 --> 15:12.000 or enterprises that run many databases find useful. 15:12.000 --> 15:17.000 So this is enable them to build a profitable business. 15:17.000 --> 15:20.000 Tatiana mentioned the last time I talked to her, 15:20.000 --> 15:23.000 that they had something like 10 million people using their software. 15:23.000 --> 15:24.000 That's big. 15:24.000 --> 15:29.000 And of whom about half of them, I think, are actually enterprise enterprise users. 15:30.000 --> 15:33.000 So this enables them to give back. 15:33.000 --> 15:37.000 Fundamentally, the way that the D-beaver gives back is, 15:37.000 --> 15:39.000 they just offer this community edition. 15:39.000 --> 15:40.000 That's the biggest thing. 15:40.000 --> 15:42.000 But this is a real, this is a real gift. 15:42.000 --> 15:44.000 And in a couple of interesting ways. 15:44.000 --> 15:47.000 One, it's a very popular project with many, many contributors. 15:47.000 --> 15:51.000 So it has 41,000 stars, hundreds of contributors. 15:51.000 --> 15:54.000 One interesting thing that it does is if you, 15:54.000 --> 15:58.000 and that are out in the community, making your own open source database, 15:58.000 --> 16:02.000 because D-beaver is open source, you can, 16:02.000 --> 16:06.000 you know, do an open source implementation as a PR, 16:06.000 --> 16:09.000 get it checked into D-beaver and D-beaver will then support your database. 16:09.000 --> 16:14.000 So this is something that is actually giving back to other databases. 16:14.000 --> 16:16.000 And then, of course, there's other ways, you know, 16:16.000 --> 16:19.000 all the way to contributing to upstream projects. 16:19.000 --> 16:21.000 So they've built a good business. 16:21.000 --> 16:23.000 Again, profitable looks pretty stable, 16:23.000 --> 16:26.000 and if you're not using it, I definitely recommend it. 16:27.000 --> 16:30.000 So the third way that you can, oh, and by the way, 16:30.000 --> 16:34.000 at large scale, there's another company that has done this 16:34.000 --> 16:36.000 that is worth looking into. 16:36.000 --> 16:38.000 How many people use tail scale? 16:38.000 --> 16:41.000 Yeah, tail scale does something kind of similar. 16:41.000 --> 16:44.000 I use the free version, which is basically, 16:44.000 --> 16:45.000 it's an open source project. 16:45.000 --> 16:47.000 I mean, it's MIT license. 16:47.000 --> 16:49.000 It's the thing that allows you to be here, 16:49.000 --> 16:51.000 and like talk your service back in the closet, 16:51.000 --> 16:54.000 but there's an enterprise version, which is very popular. 16:54.000 --> 16:56.000 So this is, it's not just D-beaver, 16:56.000 --> 16:59.000 it's some very large companies, you guys use this. 16:59.000 --> 17:01.000 So the final model, and this is one, 17:01.000 --> 17:05.000 my company does, is to build cloud services. 17:05.000 --> 17:10.000 And what you do here is you have some open source software 17:10.000 --> 17:13.000 that runs in the cloud that people want to just have run 17:13.000 --> 17:15.000 as a service that they can connect to. 17:15.000 --> 17:18.000 So that's 100% service, 17:18.000 --> 17:23.000 and what you're going to do is you're going to offer support. 17:23.000 --> 17:28.000 And I'll give you an offer cloud management sort of the sort of layer 17:28.000 --> 17:31.000 that takes care of automating it, 17:31.000 --> 17:35.000 having upgrades, fixing performance, security, 17:35.000 --> 17:38.000 delivering support, so on and so forth. 17:38.000 --> 17:40.000 And that, of course, is completely proprietary, 17:40.000 --> 17:43.000 but the thing that it's managing is open source. 17:43.000 --> 17:45.000 A couple interesting things about this model. 17:45.000 --> 17:48.000 One is, again, it has really good gross margins. 17:48.000 --> 17:52.000 The fact that it goes down to 50% is deceptive. 17:52.000 --> 17:56.000 That's mostly because if you are managing it and running it 17:56.000 --> 17:58.000 in your own account, you also have to pay the cloud providers. 17:58.000 --> 18:00.000 So that lowers your margins. 18:00.000 --> 18:04.000 But if you run it in the user account, which is a popular model, 18:04.000 --> 18:06.000 then you gross margins are very good. 18:06.000 --> 18:10.000 So again, a good basis for building a solid business. 18:10.000 --> 18:12.000 It's also very sticky. 18:12.000 --> 18:14.000 When you're running something for somebody, 18:14.000 --> 18:16.000 they're not going unless you screw up, 18:16.000 --> 18:19.000 or they run into financial problems, they're not going to ever give you up, 18:19.000 --> 18:21.000 and they have a tendency to use more of it, 18:21.000 --> 18:23.000 if they like what you're doing. 18:23.000 --> 18:25.000 There's another interesting thing about this model, 18:25.000 --> 18:28.000 which kind of circles back to what Procona did with 18:28.000 --> 18:30.000 and others have done with services. 18:30.000 --> 18:34.000 These managed services are also good vehicles 18:34.000 --> 18:37.000 for delivering very, very high quality support. 18:37.000 --> 18:39.000 Because you're operating the system, 18:39.000 --> 18:41.000 and if you also have a great deal of expertise 18:41.000 --> 18:43.000 about the software and the applications, 18:43.000 --> 18:45.000 you can have very proactive support 18:45.000 --> 18:47.000 that, for example, detects problems 18:47.000 --> 18:50.000 and helps solve them long before users notice them. 18:50.000 --> 18:52.000 Who does this? 18:52.000 --> 18:53.000 We do. 18:53.000 --> 18:56.000 So, as it's how it works, 18:56.000 --> 18:58.000 we have a product called Tenidiot Cloud. 18:58.000 --> 19:00.000 It manages Clickhouse, 19:00.000 --> 19:03.000 and at this point, four public clouds, 19:03.000 --> 19:05.000 of which one is Hatsker, 19:05.000 --> 19:08.000 and then the other three are the usual suspects. 19:08.000 --> 19:10.000 How it works is, 19:10.000 --> 19:12.000 we basically stand up Kubernetes, 19:12.000 --> 19:16.000 whenever people want to manage databases, 19:16.000 --> 19:18.000 and then what the management software does, 19:18.000 --> 19:21.000 is it basically just pushes open source components 19:21.000 --> 19:23.000 into that Kubernetes cluster, 19:23.000 --> 19:27.000 so that, and then takes care of remotely managing. 19:27.000 --> 19:30.000 It can run, you know, in our accounts 19:30.000 --> 19:32.000 or it can run in the user accounts. 19:32.000 --> 19:35.000 The interesting thing here is because this is fully open source, 19:35.000 --> 19:38.000 the customers can just disconnect whenever they want. 19:38.000 --> 19:39.000 They usually don't, 19:39.000 --> 19:41.000 but, you know, because if something's working, 19:41.000 --> 19:43.000 you don't want to mess with it. 19:43.000 --> 19:47.000 But that's, that's basically a big value for the customers. 19:47.000 --> 19:50.000 How do we give back in a bunch of different ways? 19:50.000 --> 19:53.000 We contribute PRs to upstream Clickhouse. 19:53.000 --> 19:55.000 We've done hundreds of them. 19:55.000 --> 19:56.000 You know, there's a, 19:56.000 --> 19:57.000 it's a big list. 19:57.000 --> 20:00.000 Some of it overlaps the kind of stuff that Procona does. 20:00.000 --> 20:03.000 Probably the most single successful thing we've done 20:03.000 --> 20:05.000 is to build our cloud. 20:05.000 --> 20:07.000 We built an operator, a Kubernetes operator, 20:07.000 --> 20:08.000 for Clickhouse. 20:08.000 --> 20:11.000 That is the most popular way to run Clickhouse on Kubernetes. 20:11.000 --> 20:14.000 So it's been the most successful piece of open source software 20:14.000 --> 20:15.000 we've developed. 20:15.000 --> 20:19.000 And then, of course, there are many other things as well. 20:19.000 --> 20:21.000 Licensing. 20:21.000 --> 20:24.000 Everybody stresses about open source licenses. 20:24.000 --> 20:26.000 And especially when, you know, 20:26.000 --> 20:28.000 you're building a business and you're thinking, man, 20:28.000 --> 20:30.000 you know, if I don't pick the right license, 20:30.000 --> 20:32.000 I'm going to get ripped off. 20:32.000 --> 20:34.000 So the question is, 20:34.000 --> 20:36.000 how much should you stress about this? 20:36.000 --> 20:38.000 Well, as a practical matter, 20:38.000 --> 20:40.000 at least in my experience, 20:40.000 --> 20:42.000 the answer is not much. 20:42.000 --> 20:45.000 It really doesn't make that much difference 20:45.000 --> 20:47.000 in your ability to make money. 20:47.000 --> 20:51.000 So my advice on this is two things. 20:51.000 --> 20:54.000 Pick a license that supports the four freedoms. 20:54.000 --> 20:56.000 I mean, that's why we're into open source 20:56.000 --> 20:58.000 as we believe in sharing the software. 20:58.000 --> 21:00.000 We believe in granting rights to people 21:00.000 --> 21:02.000 to continue to use it to understand how 21:02.000 --> 21:04.000 or it's to make changes to distribute it. 21:04.000 --> 21:06.000 So pick one that supports it. 21:06.000 --> 21:09.000 There's lots of choices and stick with it. 21:09.000 --> 21:10.000 Don't change it. 21:10.000 --> 21:12.000 Don't mess with people. 21:12.000 --> 21:16.000 So, and everything and, you know, 21:16.000 --> 21:18.000 like, yes, there are some differences 21:18.000 --> 21:21.000 in the license like patchy is more popular for companies. 21:21.000 --> 21:23.000 You know, 21:23.000 --> 21:27.000 GPL V3, a lot of corporations don't like it, 21:27.000 --> 21:28.000 but it has other advantages. 21:28.000 --> 21:31.000 They don't make that much difference in the end. 21:31.000 --> 21:34.000 What should you stress about? 21:34.000 --> 21:36.000 Well, here's three things. 21:36.000 --> 21:38.000 And these are not related to code. 21:38.000 --> 21:41.000 And this is where it gets hard because we're all geeks, right? 21:41.000 --> 21:42.000 Well, okay, I am. 21:42.000 --> 21:44.000 Some of you guys are too. 21:44.000 --> 21:46.000 So, what should you stress about? 21:46.000 --> 21:48.000 First thing is marketing. 21:48.000 --> 21:51.000 Actually, in a way you don't have to stress 21:51.000 --> 21:53.000 when you're starting out, 21:53.000 --> 21:54.000 because you can give back to the community. 21:54.000 --> 21:56.000 That already is getting your story out 21:56.000 --> 21:58.000 about how great you are, 21:58.000 --> 22:00.000 how you can be trusted, 22:00.000 --> 22:02.000 how you can produce good software. 22:02.000 --> 22:03.000 But the problem is, 22:03.000 --> 22:05.000 as your company grows, it gets harder. 22:05.000 --> 22:07.000 We work in very competitive markets. 22:07.000 --> 22:10.000 It is expensive to tell people that we have a product. 22:10.000 --> 22:13.000 It's expensive to find the people who are interested in it, 22:13.000 --> 22:15.000 and it takes special skills. 22:15.000 --> 22:18.000 So, you generally have to hire people who have those. 22:18.000 --> 22:20.000 Unless you're really deeply interested in it, 22:20.000 --> 22:22.000 you'll have to hire people who do this. 22:22.000 --> 22:23.000 Sales. 22:23.000 --> 22:27.000 Asking for money is a real art. 22:27.000 --> 22:29.000 In fact, quite a deep one, 22:29.000 --> 22:32.000 knowing how much to ask, knowing when to ask, 22:32.000 --> 22:35.000 those are also, those are also real skills. 22:35.000 --> 22:38.000 So, you will, unless you are very good at doing it yourself, 22:38.000 --> 22:39.000 and even if you are, you will need to handle. 22:39.000 --> 22:41.000 You'll need to hire people who do this. 22:41.000 --> 22:43.000 Final thing is financial management. 22:43.000 --> 22:46.000 So, managing money is critical. 22:46.000 --> 22:47.000 When you're building, 22:47.000 --> 22:50.000 we do everything off Google cheats, 22:50.000 --> 22:52.000 but we're very, very careful about money. 22:52.000 --> 22:54.000 And our company has been blessed 22:54.000 --> 22:56.000 by two people who are very, very gifted at this. 22:56.000 --> 22:58.000 They were there when I arrived, 22:58.000 --> 22:59.000 so I got totally lucky. 22:59.000 --> 23:01.000 This does bring up, 23:01.000 --> 23:03.000 you know, so you have to be really interested in things 23:03.000 --> 23:04.000 that cash flow. 23:04.000 --> 23:06.000 Like, you know, how much money do I have in the bank? 23:06.000 --> 23:07.000 What is my projections? 23:07.000 --> 23:09.000 Revenue, otherwise you'll end up laying people off 23:09.000 --> 23:11.000 or just blowing it. 23:11.000 --> 23:14.000 And so, you've got to do this right. 23:14.000 --> 23:15.000 This brings up a question, 23:15.000 --> 23:17.000 well, what if my company is getting big 23:17.000 --> 23:19.000 and I do need money? 23:19.000 --> 23:21.000 Should you seek that, your funding? 23:21.000 --> 23:23.000 Well, I'm not going to, 23:23.000 --> 23:25.000 this talk is running long, 23:25.000 --> 23:27.000 but in a nutshell, 23:27.000 --> 23:30.000 if you go and take seed or stage, 23:31.000 --> 23:33.000 or stage a funding, 23:33.000 --> 23:35.000 or a series funding, 23:35.000 --> 23:37.000 the people that are giving you the money 23:37.000 --> 23:40.000 are looking for a very, very big payoff, 23:40.000 --> 23:42.000 like on the order of 100 to 1. 23:42.000 --> 23:43.000 And the reason for that 23:43.000 --> 23:46.000 is early stage funding is very uncertain. 23:46.000 --> 23:49.000 VCs have no idea whether you're going to be successful. 23:49.000 --> 23:51.000 They just know that if they fund enough people 23:51.000 --> 23:53.000 and they all have a potential 100 to 1 shot, 23:53.000 --> 23:55.000 a couple will come through 23:55.000 --> 23:57.000 and they'll return their fund 23:57.000 --> 23:59.000 and pay off all that people. 23:59.000 --> 24:01.000 That's something that you need to keep in mind, 24:01.000 --> 24:04.000 because they are looking for huge returns. 24:04.000 --> 24:06.000 What's the effect on you? 24:06.000 --> 24:09.000 Well, you're running an open source business. 24:09.000 --> 24:12.000 Open source business is to not make a huge amount of money. 24:12.000 --> 24:15.000 Here's a quote from Larry Ellison from 2006 24:15.000 --> 24:18.000 at that, you know, they're looking at my sequel. 24:18.000 --> 24:21.000 That year, Oracle was going to make $15 billion. 24:21.000 --> 24:24.000 My sequel is around 30 to 40 million. 24:24.000 --> 24:27.000 To get the kind of payoff that VCs are looking for, 24:27.000 --> 24:29.000 you essentially have to have proprietary pricing. 24:29.000 --> 24:32.000 You have to, so this is going to cause you 24:32.000 --> 24:35.000 to make some potentially bad choices 24:35.000 --> 24:38.000 because you're going to have to do things like 24:38.000 --> 24:41.000 re-licensing that, you know, in an effort 24:41.000 --> 24:46.000 to get that, what amounts to proprietary software pricing. 24:46.000 --> 24:48.000 This is a great picture from Tintin, 24:48.000 --> 24:50.000 his dog snowy is constantly being tempted 24:50.000 --> 24:54.000 in this case to either carry a message or eat a bone. 24:54.000 --> 24:56.000 He chose the bone. 24:56.000 --> 24:59.000 So, and believe me, it's, you know, you think, 24:59.000 --> 25:01.000 early on, you think you're not going to make this choice, 25:01.000 --> 25:04.000 but when you're talking about $100 million 25:04.000 --> 25:07.000 or you have people on your board that are thinking 25:07.000 --> 25:08.000 about hundreds of millions of dollars, 25:08.000 --> 25:11.000 it gets very, very difficult to say no. 25:11.000 --> 25:13.000 So, a couple of answers to that, 25:13.000 --> 25:16.000 and then where you're ready to start your business. 25:16.000 --> 25:18.000 Number one, I think a really important thing is 25:18.000 --> 25:20.000 if your software is really taking off, 25:20.000 --> 25:23.000 how to avoid that kind of choice, 25:23.000 --> 25:25.000 we'll don't have it. 25:25.000 --> 25:27.000 Just solve the problem, and one way to do this 25:27.000 --> 25:29.000 is to put the software in a foundation. 25:29.000 --> 25:31.000 This is a really important step 25:31.000 --> 25:34.000 to ensure that your project remains open. 25:34.000 --> 25:35.000 There's some great foundations out here. 25:35.000 --> 25:39.000 I'm just naming three of the most widely known ones. 25:39.000 --> 25:41.000 Answer two is other funding sources. 25:41.000 --> 25:43.000 VCs are not the only way to get money. 25:43.000 --> 25:46.000 The all-time best way to get money 25:46.000 --> 25:48.000 for investments, get your customers to pay for it. 25:48.000 --> 25:51.000 Then you're sure you're doing something they really want. 25:51.000 --> 25:54.000 The trouble is they won't necessarily pay for, 25:54.000 --> 25:57.000 you know, very speculative investment. 25:57.000 --> 25:59.000 There's other things you can do creative financing. 25:59.000 --> 26:02.000 We've used our revenue, what's called revenue-based financing. 26:02.000 --> 26:06.000 It's based on the amount of revenue coming in at high interest rates. 26:06.000 --> 26:09.000 Or you can just skip to late-stage funding, 26:09.000 --> 26:11.000 which has better options. 26:11.000 --> 26:13.000 And that's it. 26:13.000 --> 26:15.000 So, where to go next? 26:15.000 --> 26:17.000 If you're going to build a business, 26:17.000 --> 26:19.000 you could look at our talk that we did for a 26:19.000 --> 26:20.000 second time. 26:20.000 --> 26:22.000 If you're going to look at companies that have done this 26:22.000 --> 26:23.000 and then dive in and do it. 26:23.000 --> 26:27.000 And one thing I want to say is if you believe in open source 26:27.000 --> 26:30.000 and you believe that the foundations of computing 26:30.000 --> 26:33.000 should be open source services that are available to everyone 26:33.000 --> 26:35.000 and they should stay free. 26:35.000 --> 26:37.000 Building businesses is a support that is a good way 26:37.000 --> 26:39.000 to pitch in and help. 26:39.000 --> 26:42.000 So, please go ahead and try it. 26:42.000 --> 26:44.000 And if you want advice, 26:44.000 --> 26:46.000 these are ways you can find me. 26:46.000 --> 26:50.000 All of us, we did this by, you know, 26:50.000 --> 26:52.000 through the help of people who are paying it forward. 26:52.000 --> 26:53.000 We are the same thing to you. 26:53.000 --> 26:56.000 So, if you want to do this, come talk to us. 26:56.000 --> 26:57.000 I'd be happy to help. 26:57.000 --> 26:58.000 Thank you. 26:59.000 --> 27:00.000 Thank you. 27:00.000 --> 27:01.000 Thank you. 27:01.000 --> 27:02.000 Thank you. 27:02.000 --> 27:03.000 Thank you. 27:03.000 --> 27:04.000 Thank you. 27:04.000 --> 27:05.000 Thank you. 27:05.000 --> 27:06.000 Thank you. 27:06.000 --> 27:07.000 Thank you. 27:07.000 --> 27:08.000 Thank you. 27:08.000 --> 27:09.000 Thank you. 27:09.000 --> 27:10.000 Thank you. 27:10.000 --> 27:11.000 Thank you. 27:11.000 --> 27:12.000 Thank you. 27:12.000 --> 27:13.000 Thank you. 27:13.000 --> 27:14.000 Thank you. 27:14.000 --> 27:15.000 Thank you. 27:15.000 --> 27:16.000 Thank you. 27:16.000 --> 27:17.000 Thank you. 27:17.000 --> 27:18.000 Thank you. 27:18.000 --> 27:19.000 Thank you. 27:19.000 --> 27:20.000 Thank you. 27:20.000 --> 27:21.000 Thank you. 27:21.000 --> 27:22.000 Thank you. 27:22.000 --> 27:23.000 Thank you. 27:23.000 --> 27:24.000 Thank you. 27:24.000 --> 27:25.000 Thank you.