[00:00.000 --> 00:11.840] Hello, everyone. Hello, hello, hello. It looks like it is time for me to begin. So if you [00:11.840 --> 00:17.760] are here for this topic, you know, you're in the right place. That's awesome. Who am [00:17.760 --> 00:24.480] I? I am Matt Yakovic. I am the head of open source strategy at Scarf or the Haas for short. [00:24.480 --> 00:29.920] You may have heard me or seen talks from me before. Multiple ways to connect with me. [00:29.920 --> 00:34.800] Podcasts, different things that I do. If you don't know what Scarf does, what we do is we help [00:34.800 --> 00:41.840] projects and maintainers grow their open source adoption, help track it, help analyze it, help [00:41.840 --> 00:46.880] you understand who's downloading and using your open source software. We also do a hacking open [00:46.880 --> 00:55.120] source business podcast. And so that is up to 14 episodes, I think are recorded now. So if you're [00:55.120 --> 01:00.320] interested in understanding from others on what is working in the open source business space, [01:01.200 --> 01:06.320] okay, it seems like I need to hold it a little closer, then that would be a good place to check [01:06.320 --> 01:12.480] out. We're also launching a new website that has all of the information from these slides, [01:12.480 --> 01:16.800] as well as a whole bunch of other details on how to measure your open source businesses. [01:17.600 --> 01:22.880] And we have it under opensourcemetrics.org. If you're interested, there's links to the podcast [01:22.880 --> 01:29.920] and various things on running an open source business and measuring it. So how many people [01:29.920 --> 01:35.680] here know open source is different, right? Unlike other businesses, obviously you are competing [01:35.680 --> 01:41.360] with free, which changes the mindset quite a bit. Let me give you an example. It's a very easy [01:41.360 --> 01:46.400] example and people will pick it apart, but let me just get through it real quick. If you're selling [01:46.400 --> 01:52.080] a car nowadays, if you're a car manufacturer, you have the ability to set your price for the car, [01:52.080 --> 01:57.440] people can walk away from the car. If they don't buy a car or a vehicle, then they would have to [01:57.440 --> 02:04.320] choose some other alternative modes of transportation. If you use a car that's in the open source space, [02:05.120 --> 02:09.280] you can still set the price for the car, but you don't actually have to pay for it. So someone can [02:09.280 --> 02:14.000] come get the car for free, drive it around, and do whatever they want with it. So consumers can [02:14.000 --> 02:19.600] choose to buy your car or someone else. And that boggles the mind of many traditional business [02:19.600 --> 02:27.200] folks because it's giving something away for free. And one of the key things that people misunderstand [02:27.200 --> 02:33.840] about open source is it's great if it's free and people can use it, but in order for people not [02:33.840 --> 02:38.480] to choose the free version and pay for something that's in the commercial space, it has to give [02:38.480 --> 02:43.200] you more value than what you're going to get in that free space. You have to figure out some way [02:43.200 --> 02:50.240] to connect with those folks, and you have to make sure that they are finding value in what you have. [02:50.240 --> 02:53.920] Now, some of you, and I'm interested to know, how many people are more on the tech space in here? [02:53.920 --> 02:58.480] Can you raise your right hands? Okay, how many people on the business space? Raise your left foot. [03:00.000 --> 03:04.720] Okay, so a few. Well, so many on the tech space consider this type of a talk when we talk about [03:04.720 --> 03:09.760] revenue and money, evil. So I'm going to go ahead and put on my other hat, which is my super evil [03:09.760 --> 03:15.360] villain hat, you know, for this talk. And that doesn't necessarily mean that this is wrong, [03:15.360 --> 03:20.800] or that it is super evil or villainous, but it's hard for tech folks, especially founders, [03:20.800 --> 03:25.600] to talk about revenue and money. This is outside of the wheelhouse, right? From a technical standpoint, [03:25.600 --> 03:31.280] when we talk about growing, you know, your business, that's not something that is naturally inherent. [03:31.280 --> 03:35.440] So people seek out different advisors, they look for who's done it before, what sort of things [03:35.440 --> 03:40.800] have worked, and then they try to repeat that process. So I want to start by giving some [03:40.800 --> 03:46.400] baselines. There's a lot of common open source business models out there. There's there's several, [03:46.400 --> 03:51.120] but I'm going to talk about the three primary ones. And that's open core, a SAS model, or is, [03:51.120 --> 03:54.320] and as a service model, it could be infrastructure, platform as a service, [03:54.320 --> 04:01.760] just consider all those the same, or services themselves. Now, from an open core perspective, [04:01.760 --> 04:06.160] okay, and excuse all the graphics in this, I'm not a graphics designer, so, you know, [04:06.160 --> 04:11.920] they're they're they're a little hokey, but open core, how many people have heard of open core [04:11.920 --> 04:18.320] before? Oh, the vast majority, but not everyone. Okay, so open core is when you have an open source [04:18.320 --> 04:22.080] version of your project that's freely available for everyone. But then you have an enterprise [04:22.080 --> 04:27.440] version, the enterprise version reserves some set of features. And then if someone wants those [04:27.440 --> 04:32.320] features, typically more enterprise features like maybe higher security integration with partner [04:32.320 --> 04:37.280] products, things like that, then people will then have the ability to purchase that from you, [04:37.280 --> 04:42.880] and then they'll pay a regular ongoing fee. A lot of companies have dual licenses in this case. [04:42.880 --> 04:48.400] So their open source is, you know, let's say GPL, and then there's some enterprise license for their, [04:49.680 --> 04:55.200] you know, enterprise version. Okay, now this is dwindling a little bit in popularity, [04:55.200 --> 05:02.480] as the cloud has taken over much of the tech industry. And that's where a lot of new businesses, [05:02.480 --> 05:07.120] especially in the open source space, are looking. And that's in this SAS or this cloud model. Okay, [05:07.840 --> 05:12.480] you can download the software, but if you really want to run it, then, you know, you should use [05:12.480 --> 05:19.680] our SAS thing will make it easier. This provides a higher level of quote unquote stickiness. Now, [05:19.680 --> 05:26.720] if you are a manager, director, executive at a company that has, you know, venture funding or [05:26.720 --> 05:31.120] that you're, you know, has executives who are, you know, kind of in been in the industry for a while, [05:31.680 --> 05:36.400] I guarantee that they have had at least 10 conversations about stickiness in open source [05:37.440 --> 05:42.400] in a calendar year. And what stickiness is, is the ability for someone to start to use your open [05:42.400 --> 05:47.440] source software, and then continue to use it. So how do we make it sticky? How do we retain them? [05:47.440 --> 05:53.840] In many companies, this is a model that provides a higher level of stickiness. Because once you [05:53.840 --> 05:58.480] get into a cloud service, it's often really, really difficult to migrate away. Because you're [05:58.480 --> 06:04.240] relying on that cloud service to be your sys admins, your DBAs, your, you know, DevOps folks, [06:04.240 --> 06:08.320] they're handling all those functions. And so to replace that, you either have to go to another [06:08.320 --> 06:13.520] cloud service, or you need to find someone who can do those functions for you. And that's often [06:13.520 --> 06:20.080] a really difficult prospect to do. But there are some issues with this model. This model is [06:20.080 --> 06:26.720] disruptible, because other people can run that exact same as a service. What we've been seeing [06:26.720 --> 06:32.240] in the space of the SAS lately, is a lot of open source companies have changed their licenses to [06:32.240 --> 06:38.000] prevent that. And they've gone from open to not open. So you might have heard of SSPL, you might [06:38.000 --> 06:45.600] have heard of the BSL, other things. These licenses are kind of sort of open. But that means that, [06:46.880 --> 06:51.040] you know, as you get into the cloud space, you're restricted from competing with them. So it's more [06:51.040 --> 06:56.480] of a preservation of their revenue stream, if you will. And then there's the services model, [06:56.480 --> 07:03.920] which is really the original kind of OG, the old, you know, model for open source software. [07:03.920 --> 07:10.560] And that model is definitely something that we have seen be very popular and very successful in [07:10.560 --> 07:16.880] a lot of different companies. This is a viable pathway for a lot of companies to start. It's an [07:16.880 --> 07:23.520] easier pathway to revenue. But it is hard to grow beyond a certain limit. So, you know, you might [07:23.520 --> 07:28.560] be able to get a handful of customers working with you from a support perspective, from a [07:28.560 --> 07:34.480] consulting perspective. But you're really bound on the number of people. Because you have to have [07:34.480 --> 07:37.760] people to deliver those services. And if you don't have people to deliver those services, that [07:37.760 --> 07:44.000] becomes quite an issue. And if you compare the three models, I'll, you know, you can compare them. [07:44.000 --> 07:47.920] You know, you've got that open core, which is a proven model, but it is getting disrupted [07:47.920 --> 07:54.000] and eroded by the cloud a little bit. And, you know, there's still some success in that space, [07:54.000 --> 07:58.480] because a lot of companies, especially when you listen to, like, some of the sovereign cloud, the [07:58.480 --> 08:03.200] cloud native, they're looking at enterprise versions or open core versions. They want to run [08:03.200 --> 08:07.040] their operators. They want to build their own private clouds. So, there is still some play in [08:07.040 --> 08:14.400] that space. X is a service or software is a service. You know, that is still a very popular model. [08:14.400 --> 08:20.800] But the question is, if you are releasing features that are only available in a SaaS version, and [08:20.800 --> 08:28.160] only a minimal open source version, is SaaS, the SaaS model, is it still open source? Or is it not? [08:28.160 --> 08:32.160] And that's a question that we really haven't answered yet, especially as companies start to [08:32.160 --> 08:38.400] reserve features and software specifically just for their SaaS customers. And then there's the [08:38.400 --> 08:42.800] service model, which is definitely the easiest to get started if you're a small company. It, you [08:42.800 --> 08:48.080] know, has the most likelihood that you'll get some revenue in the door really quickly without [08:48.080 --> 08:56.800] a huge buy-in cost. Now, part of this model, okay, if you, whether you choose whichever model you [08:56.800 --> 09:04.880] choose, is how do you get people to move from being open source users to actually paying customers? [09:05.840 --> 09:15.680] Okay. And, you know, this is probably the most stressful thing in open source maintainer, founder, [09:15.680 --> 09:22.960] someone in the open source space has. Because ultimately, you're looking at this open source [09:22.960 --> 09:27.920] adoption, and let's say you're tracking 100 million downloads, but nobody wants to pay for [09:27.920 --> 09:33.600] anything. That's actually a use case that happens quite a bit. How do you, how do you, how do you [09:33.600 --> 09:38.000] figure that out? How do you, how do you turn that 100 million downloads? How do you turn that, you [09:38.000 --> 09:46.160] know, 5,000 people who are tweeting about you into something that is actually commercially viable? [09:46.880 --> 09:52.320] And when we go look at the classic sort of marketing funnel, and anybody here in marketing, [09:52.320 --> 09:57.840] I'm just curious, no one will, one person, two people want to admit it. Hey, three people. Okay. [09:57.840 --> 10:03.120] Oh, look at that. All right. So you might be familiar with this if you are in marketing, right? [10:03.120 --> 10:07.760] So this is just this very simple marketing funnel where you've got, you know, people who are hitting [10:07.760 --> 10:15.040] your website, and they turn into a contact or a lead, if you will, where they become a known user, [10:15.040 --> 10:18.720] and then they turn into a marketing qualified lead after they do a certain number of things. [10:18.720 --> 10:22.400] So let's say you're there on your website, they sign up for a webinar, maybe they downloaded [10:22.400 --> 10:25.680] your software, maybe they did a few other things, they get a certain number of points, [10:25.680 --> 10:28.880] and when they reach the point value, then all of a sudden it's like, whoo-hoo, [10:28.880 --> 10:34.800] you are now qualified, so then sales can talk to you. And then you have a, you know, sales [10:34.800 --> 10:39.200] accepted lead. So sales goes, ooh, that looks pretty good. And then you have a sales qualified [10:39.200 --> 10:44.560] lead when they start to say, hmm, I think I could sell something. All right. This is the classic [10:44.560 --> 10:51.760] marketing funnel, but it doesn't work in an open source space. There is a similar way to think [10:51.760 --> 10:58.800] about this. Okay. Because the open source model kind of flips this on its head. There is this model [10:58.800 --> 11:04.160] where we've got the interest phase, which is the people who are discovering what your product [11:04.160 --> 11:11.440] can do, how it can behave, what it can do differently. And you want them to actually try your software, [11:11.440 --> 11:18.240] okay? You want them to install your software or use your containers or use your library in, [11:18.240 --> 11:23.920] you know, their product. And only a subset of those who are actually interested, a subset of [11:23.920 --> 11:27.920] those who will visit your project's web page, visit your web pages, visit your assets, [11:27.920 --> 11:36.640] there, only that small subset is going to actually try. And of those who try, only a subset will [11:36.640 --> 11:44.480] continue to use it in production. And then only a subset will become a customer. So that model [11:46.000 --> 11:52.080] is very important because as we look at this phase, you've got to realize that if you have a million [11:52.080 --> 11:57.120] people on your website and you have 100,000 downloads, that's great. But there might only [11:57.120 --> 12:02.000] be 10,000 who use it in production and that might only have like 1,000 people who would be willing [12:02.000 --> 12:07.120] to pay for it. Now some of you who are familiar with the product space might say, well, we don't [12:07.120 --> 12:12.080] follow the funnel, we follow the, you know, product flywheel, the product growth flywheel. [12:12.080 --> 12:19.200] So anybody familiar with the product growth flywheel? All right, one, okay. So the idea is, [12:20.160 --> 12:23.200] it's funny because they actually stole a lot of this from the open source movement. [12:23.200 --> 12:27.600] When you look at the product flywheel, and this is all the hotness in all the product [12:27.600 --> 12:32.000] companies, right? So when we're talking about a product led company, a product, you know, [12:32.000 --> 12:37.440] product led growth, what they're talking about is people start off in this center, this core, [12:37.440 --> 12:41.760] and it's like, oh, how do we track and engage with people? How do we delight them? How do we make [12:41.760 --> 12:48.080] them really excited about things? Okay. And then we can take those strangers, turn them into prospects, [12:48.080 --> 12:53.120] which then turns them into customers, which then turns them into fans, and then they will start [12:53.120 --> 12:58.880] to become the people who will preach to the other people, to tell other people how awesome [12:58.880 --> 13:04.400] our software is and get them into the community. There's an open source version of this, and again, [13:04.400 --> 13:10.160] excuse my crappy little, you know, diagrams here, if someone wants to contribute these, [13:10.960 --> 13:17.440] better versions, that's awesome. But the idea here is, it's very similar, similar phases, [13:17.440 --> 13:22.560] but if you look at the differences, the differences are we have a download and try, [13:22.560 --> 13:26.960] we have that build and deploy into production, and then we have, oh, they're running it into [13:26.960 --> 13:31.920] prod, and then it kind of goes into, hey, do I need something more? That's when I'd be willing to pay. [13:33.120 --> 13:38.640] And then you'll evaluate a paid option and become that paying customer. It's a very similar cycle, [13:38.640 --> 13:42.640] both are the same, but there's a few differences. I like to clarify this a little bit more in the [13:42.640 --> 13:47.440] open source space because it makes sense, because really what you're talking about here is someone [13:47.440 --> 13:50.640] who's trying your open source software, they're getting used to it, they're like, wow, that's [13:50.640 --> 13:55.200] pretty cool. But if it only did this next thing, and then you're like, oh, well, I have that, [13:55.200 --> 14:01.360] it's in the enterprise version, right? If it only did, you know, you know, integration with this [14:01.360 --> 14:06.560] other product, and so that's where, oh, I need something more, and then you get them to evaluate [14:06.560 --> 14:15.760] it. And that becomes a way to build external evangelists. So there's also this idea of [14:15.760 --> 14:22.800] evangelism or evangelists, a way to foster people who are in your community, who become fans of your [14:22.800 --> 14:28.800] product, who then actually go out and contribute to that flywheel, right? So the goal here is [14:28.800 --> 14:34.800] you want someone who can start off telling people, wow, I was really successful with this application, [14:35.680 --> 14:40.000] and then they'll tell their followers and people who like them and people who are like-minded, [14:40.000 --> 14:47.120] and then those people will watch that content, and then those people will go ahead and try it. And [14:47.120 --> 14:52.080] it's similar when you add DevRel, evangelism, this is all part of that product flywheel, [14:52.080 --> 14:56.720] where you want more people to understand it and then eventually try it, and then eventually you [14:56.720 --> 15:02.160] want those people to create or help foster new people into the community where there are new [15:02.160 --> 15:08.880] evangelists who can go out there and talk about your product. So let's talk specifically about [15:08.880 --> 15:14.880] driving open source adoption, okay? And there's only three things you need to do. Yeah, I'm going [15:14.880 --> 15:19.120] to sound like one of those shill salesmen on the late night, you know, TV commercials where it's [15:19.120 --> 15:23.600] like, you know, these three things, if you buy into my system, you'll be awesome. You know, [15:23.600 --> 15:29.200] this will be the best thing ever for you. And so there's actually three requirements that we'll dig [15:29.200 --> 15:35.920] into a little bit, and as much as we can in the time remaining, for driving open source adoption [15:35.920 --> 15:42.400] and growing your open source adoption. And those secrets to the success are a kiss ass product. [15:42.400 --> 15:47.120] Okay, it all starts with the product. If you don't have a product that people want, it doesn't [15:47.120 --> 15:51.840] matter what else you do, they're not going to, you know, listen to you, they're not going to try [15:51.840 --> 16:00.160] it out. It needs to be rock solid. We need people to know who you are. Okay, a lot of open source [16:00.160 --> 16:05.920] projects that are out there have awesome technology behind them, but nobody knows who they are. Has [16:05.920 --> 16:10.240] anybody run into a product like and said like, how did I not know about that like three years ago? [16:10.800 --> 16:17.600] Right, that happens. And why is that? Because however, you know, whoever's running these projects, [16:17.600 --> 16:23.360] they're not really sure how to bring awareness to everyone else. And then you need to make it [16:23.360 --> 16:30.320] easier than all the alternatives out there in the space. Okay, obvious, right? All these are obvious, [16:30.320 --> 16:37.680] that it's not like it's like anything that's just crazy. All these should be obvious and easy as well. [16:40.000 --> 16:44.800] Now, I can't go into a deep dive for all of these, but briefly, I want to touch on each one, [16:44.800 --> 16:51.520] and we'll go through each one separately. So how do you build a kick ass open source project? [16:51.520 --> 16:56.480] Now, I can tell you, like, you know, all the technical jargon, but I'm going to stick with [16:56.480 --> 17:00.720] from a business perspective, what this looks like. First, I want to talk about the biggest [17:00.720 --> 17:06.800] mistake we all make when we're building a project. At a certain point, we have to realize that, you [17:06.800 --> 17:11.440] know, when we want when we are building a project, we're often starting it, building it for what we [17:11.440 --> 17:17.120] want. And that works for maybe like a little bit of time, and you might even make a successful [17:17.120 --> 17:21.120] company, but you're going to limit your growth if you are building for people who you think [17:21.120 --> 17:28.400] are the exact same people as you, right? It's that, oh, I assume everyone is like me. I know [17:28.400 --> 17:32.320] people who follow this, and they're like, well, I assume that everyone's going to install on the [17:32.320 --> 17:37.520] command line. I assume everyone will install from source. And so that's just the way that they think. [17:38.640 --> 17:43.200] That's not necessarily how everyone else in the community thinks. So make sure that you avoid that. [17:43.200 --> 17:51.040] And that's why the first key, in my opinion, is understanding, right? Understand the product [17:51.040 --> 17:57.120] and its value, right? What can your product do better than anyone else? Okay, are people willing [17:57.120 --> 18:00.080] to even pay for that? You should ask yourself that question. Can you answer that question? [18:00.880 --> 18:06.480] What do you do over the alternative? Right? So is there an alternative to this? Is there some other [18:06.480 --> 18:11.200] piece of software, something else that other people are doing? What do you consider success? Now, [18:11.200 --> 18:15.920] this one is really interesting. I've talked to a lot of founders. I've talked to a lot of CEOs, [18:15.920 --> 18:20.320] executives. And, you know, they'll talk to me about, oh, we want to accelerate our growth, our [18:20.320 --> 18:25.120] adoption. And I'll say, well, what's your goal? And they'll be like, well, we want to, you know, [18:25.120 --> 18:29.600] sell the company, or we want 10x evaluation, or we want, like, you know, fill in the blank. It's [18:29.600 --> 18:34.560] like, no, no, no, what do you really want? What do you want in the next year? What do you really, [18:34.560 --> 18:39.600] really want? And a lot of times it's like, oh, well, really what we want is people to sign up for [18:39.600 --> 18:43.440] our new beta. I've actually had that conversation where it's like, we just want like 1,000 people [18:43.440 --> 18:49.280] to try it and then give us feedback. Oh, it's a very different sort of product that you're building [18:49.280 --> 18:53.920] and sort, and also marketing activities or community activities that you're doing. And so [18:53.920 --> 18:59.280] that's something to consider. Who's going to use it? Are you building this for, you know, [19:00.160 --> 19:04.720] devs? Are you building this for business folks? Who's the end user? Who's going to benefit from [19:04.720 --> 19:13.120] this the most? And who would pay for that product if they're in that space? And where are those [19:13.120 --> 19:18.720] users most likely to be? You know, are these users going to be at Faustin? Are they going to be at, [19:18.720 --> 19:24.320] like, you know, Open Source Summit? Are they going to be at KubeCon? I mean, like, where will you see [19:24.320 --> 19:29.680] these people pop up? And how do you ensure that you know where they are and what sort of things [19:29.680 --> 19:35.040] that they like and what sort of things that they need? And so you want to understand that really [19:35.040 --> 19:41.520] early on, even if you're not trying to monetize or commercialize the Open Source yet, you know, [19:41.520 --> 19:45.840] you want to keep that in the back of your mind and start to funnel things in that right direction. [19:46.960 --> 19:52.240] Now, the next key is to listen and ask questions. Now, remember, I said a big mistake people make [19:52.240 --> 19:57.120] is the founder builds the software specifically for themselves, and they don't necessarily take [19:57.120 --> 20:04.000] into account other people in the ecosystem. And this is where product success is really determined [20:04.000 --> 20:10.320] based on your users and your potential users and listening to their needs. If you're not listening [20:10.320 --> 20:16.160] to their needs, you're going to be missing out on a lot of opportunity for growth. And that's where [20:16.160 --> 20:22.640] that mistake comes in. So avoid that if you can. The other one is you have to set a goal and focus [20:22.640 --> 20:28.480] on it. Be really good at a few things, not really okay at a lot of things, because there's a lot [20:28.480 --> 20:32.640] of other software you might be competing with. There's a lot of other people who are potentially [20:32.640 --> 20:39.200] solving the same issues. And it's really easy to get distracted by shiny object syndrome. Anybody [20:39.200 --> 20:43.680] get distracted by shiny object syndrome? I'm like that all the time. Yeah, I see hands waving all [20:43.680 --> 20:49.680] the time, right? And that is so easy to do. In fact, I was just talking with Avi, our CEO over here, [20:49.680 --> 20:53.360] and we're like talking about, you know, tracking web page stuff. And I'm like, oh, we could do that [20:53.360 --> 20:57.280] in our product. Oh, but we're like, wait a minute, we that's not our core product. Why would we go [20:57.280 --> 21:03.520] down that hole, right? But people do that a lot. And when you do that, it might be cool. And you [21:03.520 --> 21:08.080] might learn that there's a new subset and there might be a time for expanding your product. [21:08.080 --> 21:13.120] But until your core product has that established base, you don't want to do that. [21:13.120 --> 21:22.880] So don't deviate too often. Okay, stick to that original purpose. And if you do that, you will [21:22.880 --> 21:27.680] eventually find that you become better known on the market. People are willing to talk about you, [21:27.680 --> 21:32.080] use you, and then share additional contributions to maybe make your project a little better. [21:33.920 --> 21:40.880] Okay, this one's a big one. All right, has to meet or exceed expectations. Now, this totally [21:40.880 --> 21:46.240] seems obvious, right? But let's assume that I build a cell phone. And my cell phone has the [21:46.240 --> 21:51.200] awesomest speakers, the awesomest music system, and it is the best gaming platform ever, but it [21:51.200 --> 21:58.720] can't take calls. Is it still a cell phone? If I market it as a cell phone, if I sell it as a [21:58.720 --> 22:05.360] cell phone, if I say, oh, this is my cell phone, it doesn't matter. It's not. So if I'm trying to [22:05.360 --> 22:10.560] build a cell phone, but it is great at just music and gaming, but it can't take a call, I have not [22:10.560 --> 22:17.040] met the expectations of that. And that's where you need to figure out what are the table stakes for [22:17.040 --> 22:24.400] your project? What are the key ingredients? Whatever those are, that's where you have to focus. [22:25.360 --> 22:31.440] And that means do the job that it was originally intended to do, and meet some minimum standards. [22:31.440 --> 22:36.240] And there are minimum standards that we all have from an open source perspective. Okay, [22:36.240 --> 22:41.280] what are some of those minimum standards? Hey, you know what? It has to be secure, right? Like, [22:41.280 --> 22:46.080] who wants an insecure piece of software nowadays? Well, we have too many of them. Let's not add [22:46.080 --> 22:52.560] anymore. But it has to protect from data loss. I don't want leaky data. You know how many times [22:52.560 --> 22:57.520] my personal data has been leaked? My daughter's going to inherit free credit monitoring for the [22:57.520 --> 23:03.360] rest of her life, because I've had so many data leaks. I mean, I think there's several Matt Yankov [23:03.360 --> 23:08.960] it's running around right now, maybe even at this conference. So, you know, you know, it also has to [23:08.960 --> 23:14.960] be as bug free as possible. Now, while we strive, I'll strive for bug free software, we know it's [23:14.960 --> 23:20.480] not a viable thing. But shipping with known critical bugs is bad. And I know some companies who do it [23:20.480 --> 23:29.600] just to meet deadlines. It's not good. Next up is UI UX. Who's anybody any UI UX guys, [23:29.600 --> 23:38.000] any people on here? Okay, awesome. Yes, many, many. All right. Well, so it's like this. If your [23:38.000 --> 23:46.240] product is designed like this elevator, that's really confusing. Isn't it? I mean, UI UX is [23:46.240 --> 23:52.160] probably one of the most important things to get right. The user experience can ruin an awesome [23:52.160 --> 24:00.480] technical project. You can have the best technology out there. And if the user interface sucks, people [24:00.480 --> 24:09.360] won't use it. Or they'll use it once, which is even worse. So a bad user experience can ruin any [24:09.360 --> 24:15.280] benefits that you provide. GitHub is littered with projects like this, by the way. So of the [24:16.480 --> 24:21.680] millions and millions of projects that are out there, many of them fall victim to this, [24:21.680 --> 24:29.200] even if they are really solid. And if you want to grow your community and grow adoption, easy. [24:29.200 --> 24:35.840] Remember, easy was my point number three. You know, you need users who can be advanced if they [24:35.840 --> 24:41.440] want to and get into the weeds. But most of the users won't. And again, that gets back to that [24:41.440 --> 24:46.880] assumption that we build things for ourselves. A lot of CTOs, a lot of co founders, a lot of people [24:46.880 --> 24:51.760] in this space, what they think is, I'm super technical. And they are, they're building a company. [24:51.760 --> 24:56.000] Those out there in the space who are founders, you know, at that executive level, maybe they're, [24:56.000 --> 25:01.680] you know, maintainers of projects, you, you are all brilliant. And that means that your [25:01.680 --> 25:07.680] brilliance probably exceeds your average user's brilliance. That's okay. But keep that in mind. [25:07.680 --> 25:15.040] So make it easy for folks. And my product key number six is community. Okay. Open source [25:15.040 --> 25:20.320] projects need community. And we need to bake that in to the beginning. Okay. And that means [25:20.320 --> 25:27.280] including plans for fostering that strong community up front. Now I brought this up. I was previewing [25:27.280 --> 25:33.280] my slides and someone, you know, said, well, wait a minute, what about curl? Does curl really have [25:33.280 --> 25:39.600] a strong community of contributors? Right. And this is what I say is we have to look beyond code. [25:39.600 --> 25:44.480] This isn't just about code. You want people who will use your product. And if you search for curl, [25:44.480 --> 25:50.880] for instance, um, you know, your favorite search engine, you're going to find thousands of examples [25:51.520 --> 25:56.720] of how to use curl, right? You will. You can do it right now if you want tutorials, examples, [25:56.720 --> 26:02.080] cool things, hacks, whatever, that's community, even if it's not code contributions. And that's [26:02.080 --> 26:09.200] what you need to continue that product growth. All right. Moving on. Now we're talking about [26:09.200 --> 26:16.560] how do we super charge the awareness? Alrighty. So let's talk specifically about community. How [26:16.560 --> 26:23.040] anybody here in the community space? Oh, a few people. Oh, yeah. Okay. Great. So, um, this is the [26:23.040 --> 26:29.840] great hope for adoption in open source companies. In fact, a lot of people will prioritize a dev [26:29.840 --> 26:36.400] rel hire over any of the business hires because they want to start to connect directly with that [26:36.400 --> 26:43.760] community. Okay. And it's, it's, this is a balance, but the community can really help accelerate [26:43.760 --> 26:50.480] and drive that awareness if done right. But here's the thing. Okay. Community is starting to turn [26:50.480 --> 26:56.640] into the new marketing. For the last five years, the last 10 years, um, marketing has often been [26:56.640 --> 27:00.800] in control of community. Community has been a subset of marketing, sometimes a subset of product. [27:00.800 --> 27:08.320] But ultimately, this shift in mindset where the community is so important and it's driving so [27:08.320 --> 27:15.280] many conversations, it's supplanting marketing. And there are reasons that you still need marketing, [27:15.280 --> 27:20.720] but in the future, I think that what we'll see is the community team has marketing as a function, [27:20.720 --> 27:25.120] not community as a function of marketing, which is a different way to think of it. [27:25.120 --> 27:30.960] And that's a real mind shift for a lot of people in the business space. But if we do approach this [27:30.960 --> 27:36.480] from a community first approach, we can accelerate and grow the open source space even faster. [27:38.160 --> 27:43.280] And that gets back to this expectation, right? So when we talk about that expectation, I mentioned [27:43.280 --> 27:48.880] earlier that funnel, the flywheel, you know, when we're talking about, Hey, can we have people in [27:48.880 --> 27:54.560] the community who talk about us and then turn into followers of us who then continue to watch [27:54.560 --> 27:59.520] our things and then who continue to try and use our software in different ways, that's the ultimate [27:59.520 --> 28:06.480] outcome that we want. And when we talk about this community team, there's multiple functions in it. [28:06.480 --> 28:10.240] This is a chart that I've used several times before, but, you know, we're talking about [28:10.240 --> 28:14.160] DevRel or evangelism, we're talking about community management, we're talking about these [28:14.160 --> 28:19.840] OSPO roles with contributors, all of these things are part of a strong community team. [28:19.840 --> 28:25.280] All of them need to be satisfied. Now, a lot of us will focus only on evangelism, [28:25.280 --> 28:29.520] and then they won't do the other functions, or they'll focus only on the contributors and not [28:29.520 --> 28:35.200] do the other functions. And so all three of these are needed for a healthy, healthy ecosystem. [28:37.600 --> 28:43.760] Now, what are some popular ways to grow open source adoption? Okay, these are probably obvious [28:43.760 --> 28:50.400] for anybody who's been in the open source space for a while or been in the community, but they're [28:50.400 --> 28:56.720] classic, right? Content, number one, right? Content, everybody loves content. The more relevant [28:56.720 --> 29:01.840] content that you have, the more code examples that you can share, the more conference talks that [29:01.840 --> 29:07.920] are relevant to people, the more blogs you write, you know, those will make a difference, [29:07.920 --> 29:12.400] tutorials, sample applications and whatnot. All of these things are things that are important to [29:12.400 --> 29:17.920] those in the community. There's other activities you can do, especially when your product solves [29:17.920 --> 29:24.080] something that's newsworthy. Has anybody heard of the term news jacking? No? Okay, a couple of [29:24.080 --> 29:28.400] people. It's the idea that, you know, if there's something in the news cycle like, you know, [29:28.400 --> 29:33.920] oh, the latest hack, and, you know, or whatever, if there's an opportunity for you to maybe talk [29:33.920 --> 29:39.840] to people about how to build better products, if your product or your company does something [29:39.840 --> 29:46.240] differently that solves these other issues, this is a way for you to use the stories that are already [29:46.240 --> 29:52.080] out there to reach additional people. You know, if somebody's hacked or some cloud provider goes [29:52.080 --> 29:55.760] down, well, this is how we would prevent that, you know, or here's how our software did this [29:55.760 --> 30:02.400] differently. Those types of things actually can have a really far wide reach because those are [30:02.400 --> 30:09.680] things that are on top of people's minds. Now, you also need to generate friends telling friends, [30:09.680 --> 30:15.040] right? So that peer-to-peer connection, if you will, of people who are willing to share their [30:15.040 --> 30:22.560] experiences and share the different, you know, goodness of your software. And that goes hand [30:22.560 --> 30:29.120] in hand with some other things like good messaging and products and making it easier to start. [30:29.120 --> 30:41.120] Now, how do we make this easier? That's key. The first one is it requires a lot of collaboration. [30:41.120 --> 30:48.960] It's a choreographed dance, okay? So making this easier is you have to have not only easier products [30:48.960 --> 30:56.000] to use, you not only need a great user experience, you also need to make sure that you have enough [30:56.000 --> 31:02.080] information, enough tutorials, enough example code, enough out there in the community activity [31:02.080 --> 31:07.840] to merge those together because you have to have good product as well as strong community, [31:07.840 --> 31:15.760] as long as strong content, good documentation to make this work. And so that easy is important, [31:15.760 --> 31:20.880] but it goes beyond just one team's function, you know, and some of those key things or some of those [31:20.880 --> 31:27.680] things that I see people sometimes stumble with or sometimes fall down on are things like installation, [31:27.680 --> 31:33.520] you know, so how easy it is to install and get started. Same defaults. How many people have [31:33.520 --> 31:39.920] installed software and it's like, how did you not set that up? Like, why isn't that defaulted to on, [31:39.920 --> 31:45.200] you know, like encryption? Oh, encryption is just left off by default. Or my favorite is, [31:45.200 --> 31:49.120] and I don't even know, maybe there's somebody here who's run Mongo recently, but they used to not set [31:49.120 --> 31:53.280] a default password because they're like, oh, we don't want to limit people from getting started. [31:53.280 --> 31:59.520] Well, how can you not set a default password? By the way, that's the number one reason MongoDB [31:59.520 --> 32:04.080] has like data leaks, you know, you read these articles and it's like, oh, because these old, [32:04.080 --> 32:09.200] you know, these old installations didn't set a password. I don't consider that a security issue, [32:09.200 --> 32:14.000] that's more of a user issue, but that's a whole other story. How much automation do you put in [32:14.000 --> 32:22.240] place to help users? That's another critical one. Again, UI UX and how easy it is to debug [32:22.240 --> 32:27.040] problems. You know, if your problems, if the software crashes or has an issue or doesn't work [32:27.040 --> 32:31.680] as expected, do you have the right debugging in place for people to actually find what's wrong [32:31.680 --> 32:37.040] and fix it? Because if not, that's going to cause a really, really, really hard time in [32:37.040 --> 32:46.480] getting people to operationalize it. So we talked about, you know, kick-ass product, [32:46.480 --> 32:53.040] we talked about, you know, the awareness, we talked about, you know, how to make it easy. [32:53.040 --> 32:59.360] Now I got to put on my other hat. I do have another hat. It's my metrics hat. I got to wear my [32:59.360 --> 33:04.400] metrics hat for my metrics section of this. So we're going to talk about how do we track [33:04.400 --> 33:12.640] open-source usage and adoption now. Okay. And there's really kind of four or five major functions [33:12.640 --> 33:19.840] that I want to talk through. Awareness, usage, conversion, customers, and retention. Okay. And [33:19.840 --> 33:25.200] this kind of follows that funnel and that flywheel that we talked about originally. This is one of [33:25.200 --> 33:29.680] those things that, you know, each one of these requires a certain set of things to potentially [33:29.680 --> 33:34.080] look at. You don't have to look at all of these, but these are metrics that you should all consider [33:34.080 --> 33:39.200] and then pick and choose which ones make the most sense for your business. But when we talk about [33:39.200 --> 33:44.240] awareness, okay, a lot of people start with the first one on this list. So how many people [33:44.240 --> 33:51.280] track GitHub stars? Okay. Hey. You know, GitHub stars are a popularity contest, by the way. Did [33:51.280 --> 33:56.960] you know that? Like, it really is. Because it actually doesn't really show much other than [33:56.960 --> 34:01.280] people like your project or maybe they like people on your project. And did you know that for [34:01.280 --> 34:08.560] only $5.99, I can get you 1,000 stars on GitHub? There are services that do that. And so I don't [34:08.560 --> 34:15.920] know how many of the stars are actually real or not on some cases. It's hard to tell. But you [34:15.920 --> 34:20.480] should be looking at that because if it is going up into the right and you're not gaming the system, [34:20.480 --> 34:25.680] that is a good indication that people are finding your project. But keep in mind, from a project [34:25.680 --> 34:31.840] perspective, there are more users than there are users who will go to your GitHub page and get [34:31.840 --> 34:36.960] into your repository. And that's something to consider. Unless you're forcing people to go and [34:36.960 --> 34:43.760] install, you know, via source or your documentation on GitHub, you might have more people who are [34:43.760 --> 34:47.920] downloading some other mechanism. Maybe they're getting from a third party repo. Maybe they're [34:47.920 --> 34:52.560] just going to your website and getting it. So they might never go star or even see your GitHub [34:52.560 --> 34:58.720] pages. But looking at not only GitHub pages, but your traffic to your docs, your website, your [34:58.720 --> 35:04.000] different pieces of content and understanding who's interacting with it, that's another critical [35:04.000 --> 35:09.360] piece to this. And that's a classic marketing thing. A lot of people look at, you know, the website [35:09.360 --> 35:14.400] traffic. But you also want to look at those unique views to each individual page. All of those [35:14.400 --> 35:24.800] are critical to look at. Now, what's interesting when we talk about page views is we've had [35:24.800 --> 35:29.520] conversations with folks who talk about page views and they're like, there are certain pages that [35:29.520 --> 35:35.520] don't matter and then there's certain pages that do. And so understanding and being able to segment [35:35.520 --> 35:40.640] your traffic and look at who's looking at what is a very important thing to understanding that [35:40.640 --> 35:45.920] growth and adoption. So for instance, if someone looks at your pricing page, that's probably a [35:45.920 --> 35:50.160] good indicator that they might be interested in something commercial. If someone's looking at [35:50.160 --> 35:55.120] your documentation, that's a pretty good indicator that they're trying to install or debug an issue. [35:55.120 --> 35:59.120] If they're looking at tutorials on how to get started, that's a very good indication that, [35:59.120 --> 36:05.120] again, they're trying to use your software. And so you can track adoption through looking at those [36:05.120 --> 36:10.960] types of things. Now, when we talk about external evangelists and trying to get additional people [36:10.960 --> 36:17.520] into the ecosystem to talk about you or to figure out what you're doing or maybe get excited about [36:17.520 --> 36:22.160] your product, looking at the referrals to the website is another really good example. So I can [36:22.160 --> 36:29.120] see your blog or your blog pointed back to my website or your social tweet mentioned me. These [36:29.120 --> 36:37.840] referrals are often more critical than the website traffic on itself. Now, there's also two concepts [36:37.840 --> 36:41.360] that are really squishy. I won't get into a terrible amount of details if you're really [36:41.360 --> 36:46.960] interested, check out opensourcemetrics.com or.org. But share a voice. Have you ever heard [36:46.960 --> 36:54.240] of share a voice before? No? Okay. Think of share a voice is how often does the project or the thing [36:54.240 --> 37:00.080] come into people's minds or how often does it show up in search? How much of the market do you have? [37:00.080 --> 37:07.680] So, for instance, if I was to say to you, like, you know, when I say Linux, what pops into your [37:07.680 --> 37:13.200] mind? Like, what distribution pops in your mind? I'm going to guess most of you are probably going [37:13.200 --> 37:19.840] to say one of, like, three or four. But there's lots of other ones. And so those three or four [37:19.840 --> 37:25.840] have a higher share of voice than everyone else. And so if you can measure how much of your share [37:25.840 --> 37:30.560] of voice is, then you know how much of the market you potentially, you know, can reach out. And then [37:30.560 --> 37:34.480] when you have something to say, people are going to potentially listen. And then there's this idea [37:34.480 --> 37:39.520] of social reach. Now, probably everyone here has different social accounts on different platforms. [37:40.560 --> 37:45.200] And that's awesome. And the question is how much of a reach do you have? And this is, you know, [37:45.200 --> 37:50.240] measured by the number of followers, the amount of engagement. And the reason you want a larger [37:50.240 --> 37:54.160] social reach from an awareness standpoint is when you do have an announcement, when you launch a [37:54.160 --> 37:59.040] new product, when you talk about a new release, when you do have something important to say, [37:59.040 --> 38:08.400] this helps amplify that voice. Now, from a usage perspective, usage is where the adoption really [38:08.400 --> 38:16.240] comes into play here. Okay. Usage is how many people are actually using your software? All right. [38:16.240 --> 38:23.680] Now, I know a lot of companies have struggled with usage on the open source side. There's a [38:23.680 --> 38:28.080] lot of times people come back and they ask the question, well, who's using our open source [38:28.080 --> 38:34.000] software? And the answer a lot of times is, I don't know, you know, certain number of people, [38:34.000 --> 38:38.640] this number of people hit the download web page. But there's so many different places to get, [38:38.640 --> 38:43.440] you know, that information. I can't really get it. But some people can get raw download numbers. [38:43.440 --> 38:50.400] Anybody look at download numbers for their software? A few? Okay. It's hard to get. It really [38:50.400 --> 38:54.880] is, especially when you're looking at, you know, I deploy via Docker, some people are doing source, [38:54.880 --> 39:02.720] some people have packages. So that is a challenge. But it's important to understand if there are [39:02.720 --> 39:08.640] more people downloading, that's good. But you have to be careful. Because a lot, oh, I'm going to [39:08.640 --> 39:12.560] tell you a secret. There are a lot of bots on the internet. I don't know if you knew that. Don't [39:12.560 --> 39:20.080] tell anybody. And those bots download stuff. In fact, what we've found is a lot of times we might [39:20.080 --> 39:26.640] have a project. And so one of the things that I mentioned that SCARF does is it tracks, you know, [39:26.640 --> 39:32.160] things like downloads and download metrics. But what we found was stuff like, oh, this project [39:32.160 --> 39:38.560] has, you know, 50,000 downloads a month from two people. It's like, well, wait a minute. There's [39:38.560 --> 39:43.360] only two people actually downloading. So when you see 50,000 downloads, but you see only like [39:43.360 --> 39:47.120] two unique people, you're like, well, what is that? And then you dig into it and it's some bot [39:47.120 --> 39:51.920] that's downloading it and distributing it some other way. So that's important to realize that [39:51.920 --> 39:56.640] that is something that can happen. You want to look at that. But you want to look at that scrubbed [39:56.640 --> 40:02.320] unique download. So how many people are actually downloading versus how many people are, you know, [40:02.320 --> 40:06.640] just downloading and it could be bots or duplications. A lot of companies download the same [40:06.640 --> 40:11.520] software over and over again. A lot of people in their pipelines will download the latest [40:11.520 --> 40:17.920] version of software, et cetera. New users, new companies coming into the ecosystem. That's really [40:17.920 --> 40:23.440] important for usage. Sign up. So if you are a SAS provider, you're offering a free trial. Maybe you [40:23.440 --> 40:28.240] have a webinar. Maybe you have a mailing list. Maybe you have Slack. Sign ups are another thing [40:28.240 --> 40:32.560] that you want to track for usage. But that's not necessarily going to correlate to actual usage [40:32.560 --> 40:36.800] of the open source. But again, that's more of a directionally correct up into the right. [40:38.240 --> 40:46.800] You know, so the one that is really on the bottom here, the last four, this one's sticky. [40:46.800 --> 40:53.280] Okay. This one, this one's tough for me. Call home metrics. So most of the people in the open [40:53.280 --> 40:59.840] source space hate the concept of anything that's going to call home. Right? It's not necessarily [40:59.840 --> 41:06.640] always welcome. But there's a certain amount that might be needed for projects. And the question is [41:06.640 --> 41:11.840] where is that line? Right? And when I talk about call home metrics, is something running, is it [41:11.840 --> 41:17.360] installed? You know, is it available? How many things are running at any one time? These are [41:17.360 --> 41:21.920] critical questions that a lot of projects struggle with. Because just because even if you have a [41:21.920 --> 41:27.280] download doesn't necessarily mean that download actually translated into an actual user. And so [41:27.280 --> 41:31.360] you want to make sure that you can go back and forth between, you know, yeah, they downloaded, [41:31.360 --> 41:35.680] but they stopped using. Because if they downloaded and stopped using or never actually, you know, [41:35.680 --> 41:40.160] use it on a regular basis, that's a very different thing than someone who's just downloaded and you [41:40.160 --> 41:47.680] assume that they're using it. Now, when we, before we get into conversions, I want to say that just [41:47.680 --> 41:53.840] because you have adoption of your open source software doesn't necessarily mean that that's [41:53.840 --> 41:59.200] going to lead to commercial success. Okay? How many people, I mean, there might be some people in [41:59.200 --> 42:03.200] here and maybe you're willing to raise your hand. But I know that several people that I've talked to [42:03.200 --> 42:09.040] struggle with this where they're like, I have a million downloads a month, but I can't get anybody [42:09.040 --> 42:17.760] to buy anything from me. You know, oh, yeah, there we go. We have one brave soul. That happens quite [42:17.760 --> 42:22.320] a bit. And, you know, it gets back to some of those product things. And there's a lot of reasons [42:22.320 --> 42:28.640] or things that you can do to try and change that. But, you know, it gets back to this expectations [42:28.640 --> 42:34.640] on the awareness versus, you know, you know, what the business expects. And from a metric standpoint, [42:34.640 --> 42:38.960] you've got the two sides, right? So when you talk about the community side, the adoption of open [42:38.960 --> 42:45.920] source side, you know, a lot of times our efforts are around fostering that really strong community. [42:45.920 --> 42:49.200] And then there's this big question mark in the middle. And then on the business side, then [42:49.200 --> 42:53.920] there's, oh, now we want, you know, ARR, now we want, you know, to look at expansion numbers and [42:53.920 --> 42:58.640] churn and all this other stuff. So how do you bridge the gap between these two? And this is [42:58.640 --> 43:04.480] where conversion comes into play. And conversion is a bad word for people, right? You know, [43:04.480 --> 43:10.000] yeah, you will be assimilated. You know, we want that from the open source side, right? [43:11.040 --> 43:15.920] Might be a negative connotation, but I think it's funny. But I do have a super evil villain hat. [43:17.760 --> 43:23.040] But you do want those conversion metrics. You want people to say, okay, what does it take to go from [43:23.040 --> 43:30.720] free to paid? And to measure that, okay, there's, you know, different things to look at, right? [43:30.720 --> 43:34.160] That conversion is impacted by a lot of different things. And each one of these are things you [43:34.160 --> 43:39.280] can measure from a product. You know, you want better features, better security, you know, you [43:39.280 --> 43:45.680] want support, things like that. From a policy perspective, you know, companies often will [43:45.680 --> 43:50.080] require support or they might have compliance requirements. How do you meet them? And then [43:50.080 --> 43:54.880] people, right? So all of these things affect your conversion ratios. But looking at conversion [43:54.880 --> 43:59.600] ratios, you can actually control. You want to look at things like page docs or source to download [43:59.600 --> 44:05.200] conversion ratios, doc views to those who have downloaded to those who eventually signed up, [44:05.200 --> 44:11.520] if you can track that, that's great. I know that in the past, I've had to go through and do stuff [44:11.520 --> 44:17.360] like that where I've seen, I mean, actual example where it's like company A, there's really large [44:17.360 --> 44:24.400] company four years ago started looking at our blogs. And then they attended conferences, they [44:24.400 --> 44:29.120] did all this other stuff. And then two years, they called us up and had a conversation. And then [44:29.120 --> 44:34.480] a year later, they did something else. So there are breadcrumbs when you look that you can actually [44:34.480 --> 44:38.800] see what people are doing and what sort of activities help lead them to be a customer. [44:40.640 --> 44:45.360] Now, customers, when we talk about customer metrics, we're talking about standard business [44:45.360 --> 44:49.600] metrics. So these are the ones that most people who are running an open source company are familiar [44:49.600 --> 44:57.600] with the critical ones, things like the number of new customers or new logos, ARR, MRR, you know, [44:57.600 --> 45:02.960] stuff like that. So I'm not going to go terribly deep into this because there are MBA courses that [45:02.960 --> 45:08.560] cover a lot of these and I don't want to necessarily go in depth. But the ones that I look at typically [45:08.560 --> 45:17.360] the big three, number of customers, number of new logos, and yeah, the ARR, MRR. And then also, [45:17.360 --> 45:24.160] I look at the user to customer ratio. So if you know you have a million users and you have only [45:24.160 --> 45:29.440] five paying customers, how do you adjust that? How do you tweak it? How do you make that work better? [45:31.200 --> 45:37.360] And then I also look at the number of customers who are also advocating for us. From a retention [45:37.360 --> 45:42.640] and churn perspective, churn is the idea that people start using and then go away. So you want [45:42.640 --> 45:47.120] to look at those. You want to look at the number of instances that have gone away. You also want [45:47.120 --> 45:52.320] to do some competitive analysis, right? So if you have users and customers who are starting to use [45:52.320 --> 45:56.560] other open source projects or take a look at other things, you might want to be aware of that. [45:57.840 --> 46:04.640] So recapping, not all businesses, open source businesses are the same. Realistically, [46:04.640 --> 46:10.160] build an awesome project, focus on the funnel, and of course, measure it. All of these things will [46:10.160 --> 46:15.920] help lead you to open source success. Again, if you are interested in getting more details about [46:15.920 --> 46:19.760] any of the metrics or learning a little bit about what's been successful for other folks, [46:19.760 --> 46:25.280] check out opensourcemetrics.org. You can contribute to it if you think that there is something that [46:25.280 --> 46:30.640] is missing, something that you think we should work better, definitions, et cetera. We would love [46:30.640 --> 46:37.120] to have your feedback and help there. Details, if you want to reach out to me, be happy to have [46:37.120 --> 46:53.040] conversations with anybody offline. And that is it for me. Questions? Oh, we got one question. [46:53.040 --> 47:13.920] We're running mics to people in the crowd, it looks like. Oh, it's not working. Okay. Well, [47:13.920 --> 47:23.040] maybe you should shout. Maybe now. You mentioned collecting metrics from users, but because of [47:23.040 --> 47:29.360] the GDPR, you need informed consent that you're allowed to collect metrics. In your experience, [47:29.360 --> 47:34.160] what's the percentage of users that actually allow to do that? Is it like 1% or 10%? [47:34.960 --> 47:38.960] It's generally more than you would think. I mean, it's at least a third, maybe more. [47:38.960 --> 47:50.080] And beyond that, there are metrics that aren't tied to PII that can show those indicators going [47:50.080 --> 47:56.160] up and to the right. So just because you know that you have web page traffic or downloads, [47:56.160 --> 48:00.240] even if you don't know specifically who they are, you can still see those up and to the right, [48:00.240 --> 48:04.160] and those are just as important in a lot of cases. And that's why you have to pick and choose, [48:04.160 --> 48:09.280] which ones make the most sense. And in a lot of cases, what it is is it's about building that [48:09.280 --> 48:14.640] community and that relationship, so they're willing to share details. And how you do that and how [48:14.640 --> 48:20.320] you make sure is you don't do kind of like old kind of slimy sales tactics. Like you don't want [48:20.320 --> 48:24.560] to spam people, you don't want to call them. Look, let people raise their hand when they're ready. [48:25.200 --> 48:29.920] And that helps people a lot. And when people go to bat for you and tell you, you know, other people, [48:29.920 --> 48:33.120] hey, you should sign up for the newsletter, you should sign up for this. And then they give you [48:33.120 --> 48:40.960] that consent. That helps a lot. Thanks. Yeah. Oh, other other way, other way. Yeah. We had a couple [48:40.960 --> 48:49.600] more questions. We've got one in the middle, one in the end, one down here. [48:49.600 --> 49:02.960] Oh, hello. Yes. Okay. So I was one of the folks that raised their hands when you asked what, [49:02.960 --> 49:08.080] what do we know, open core. So I don't, but then you started explaining and I say, okay, [49:08.080 --> 49:13.200] that's like freemium. So I just was asking, could you explain the difference between open core and [49:13.200 --> 49:19.920] freemium? Ah, so, yeah. So freemium is where, you know, your first taste is free. And then, [49:19.920 --> 49:26.560] you know, you can, you can get more later on. Open core has been around a lot longer than, [49:26.560 --> 49:33.440] quote unquote, freemium. Okay. And typically from a freemium strategy, you're going to pay more as [49:33.440 --> 49:39.360] you go. And so having like a cloud service where the first, let's say, 100 gigs of data or something [49:39.360 --> 49:44.160] that you store is free. And then it goes on. That's more of that freemium strategy. But from an [49:44.160 --> 49:48.800] open source perspective, how I view open core is you've got a complete product that's running. [49:48.800 --> 49:53.040] You don't have to pay for anything. And if you wanted to augment it or change it on your own, [49:53.040 --> 49:58.880] you can. I think freemium typically comes in my mind and more into play on the cloud side, [49:58.880 --> 50:02.720] where you've got more of that, you know, kind of like ongoing service where you're paying for [50:02.720 --> 50:07.840] something where you can't necessarily replace it or pull it out. But it's a fine line. And it's [50:07.840 --> 50:13.280] really difficult to draw that line. And a lot of people might go back and forth. Oh, it looks like [50:13.280 --> 50:17.200] our time is up. I'm happy to answer your question if you want to come down and we can, we can chat [50:17.200 --> 50:38.720] afterwards. That's cool. I appreciate your time. I do. Thank you very much.