Peer Richelsen | Algora TV

COSS Office Hours with @peer_rich from Cal.com

Peer Richelsen

Watching now

zaf: 👋

ElTimuro: hola

McPizza0: yo

ElTimuro: sup

zaf: o7

zaf: thanks guys, was a fun session!

00:00 Welcome back everyone to the second episode of commercial open source office hours.

00:07 I'm excited for a new squad of people with hopefully interesting questions.

00:12 Last time we talked a lot about like early repositories and just getting started.

00:19 I can start with like an introducing of myself, introduction of myself. I'm Pierre, co-founder of Cal.com

00:26 and I'm joined by Renan, Tom and Ben

00:29 with some hopefully interesting open source questions.

00:33 There's maybe a couple more people joining,

00:35 but I think we can start already.

00:38 So that's me.

00:40 I'm working mostly on Cal.com nowadays.

00:42 I do a very, very small amount of angel investments

00:45 in open source, but pretty much most of my time goes into Calabcom.

00:50 Reynan, do you want to go next?

00:54 Okay, sounds good. So first, thank you for the opportunity of being here.

01:02 I'm Renan. I'm a founder of Scribble Dev, it's a small bootstrap startup from Brazil.

01:11 Our mission is to create the best full stack low code platform.

01:16 And to do that, we strongly believe that we will eventually go to the open source route.

01:23 But we have a lot of concerns,

01:25 maybe I'm being too close-minded,

01:28 but I think your input on those

01:32 will be extremely valuable for us.

01:37 So the first one would be going this route,

01:42 going full open source,

01:49 wouldn't it make it harder to eventually maybe licensing

01:54 or even selling the company and thinking about investment

01:57 and looking for investors? How was your experience with those?

02:01 Like, did it make it harder

02:04 or did you find the right people

02:08 so it really depends on your business to be honest um can you tell me more about the startup again

02:16 yeah we are a local platform so we start from building websites and mobile apps, but we do like automation workflows and AI chats.

02:29 So the idea is to be an all in one platform

02:33 to do all the digital, all your digital life,

02:38 the business digital life.

02:39 So from websites to-

02:42 You said low code or no code? What did you say? Low code, yeah no low code on local what is local so something like retool innocent yes yes

02:52 also because we tool is mostly to build it's more for internal internal dashboards and interface. So today we are more, we're working more with marketing use space,

03:07 like landing pages and websites and collecting leads

03:11 and through forums and integrating with CRMs.

03:19 Interesting.

03:20 Well, I mean, there's definitely a reason to go open source

03:22 mostly because I'd say if you're targeting developers, you want to make sure that you have a product that's just really good for developers.

03:32 And so I think open source has always been, you know, like very accessible for people, for engineers.

03:50 for people for engineers um but i mean licensing wise so it really depends on who you're selling to right like if you sell let's say consumers and prosumers i think what from my experience

03:56 open source has been pretty hard because when you have a free product that just like does something

04:01 very very good for free uh it's pretty hard to convince like a consumer to

04:08 upgrade to something pro-ish i'm not a big fan of prosumer anyway in long term so as long as you

04:14 find a way to have a free product and but also build a really strong kind of enterprise muscle

04:20 and you understand what the enterprise like larger corporations need they most likely never are happy with a free only product they always need more

04:31 support and and talk to and compliance and custom integration so I think as

04:38 long as you find a way to make money and like the tail end of the market I think

04:43 it's fine to be open source. Me personally,

04:46 I've never really seen a business that didn't work because it was open source. I think if you find

04:52 something that really works, people have always find a way to monetize a certain part of that

04:56 product. I've never heard of a startup founder who says, I wish I didn't go the open source

05:02 route after being wildly successful with my product, right?

05:05 Like everyone, if you find product market fit

05:08 whether it's an open source or closed source,

05:10 I think you're onto something, right?

05:12 And nothing else matters than finding product market fit.

05:15 Like whether you're closed source

05:16 or open source doesn't really matter.

05:20 Yeah, it makes sense.

05:21 And we can also go into that topic later.

05:23 I think other people will probably have a question about that one as well.

05:26 So Tom, you want to go next?

05:28 And sorry for butchering your name.

05:29 I didn't know you were.

05:30 Where are you from?

05:31 From Brazil, right?

05:32 Yes.

05:33 Yeah, Brazil.

05:34 Okay.

05:35 Yeah, my name is complicated.

05:36 Nobody cares.

05:37 That's okay.

05:38 Don't worry.

05:39 Hey, same.

05:40 I have to tell my name many, many times to some people.

05:47 Cool. Tom, you want to go next?

05:49 Sure. Yeah. I'm Tom. I'm actually also from Germany.

05:53 So greetings from here.

05:55 I've been coding for about 10 years.

05:59 Professionally, I would say more for like five years.

06:01 I started a company pretty young, years old doquito in germany here

06:07 not the best idea funding companies here um initially just building software for schools

06:14 to track kind of student progress um was something i was close to you know schools and teachers um

06:22 it's required by law here uh Then drifted more into agency work,

06:26 helped a lot of startups build, kind of build teams.

06:30 That's how I got more into learning about building,

06:33 shipping, scaling, and all the failing of startups.

06:37 Very interesting.

06:38 So then quit that after some time

06:42 and started just building my own stuff. Ended up more and more

06:47 on the route of building kind of an open source alternative to Google Workspace, exploring that.

06:52 In Germany, it's a very common topic, just data privacy, self-hosting, most businessmen

06:59 not really having any technical skills. So yeah, that's what I'm doing.

07:06 And so you said like an alternative to Google Workspace?

07:09 Or what did you say?

07:11 Yeah, so kind of mail, Drive, calendar maybe, maybe not.

07:19 Meeting.

07:20 These are like things that I explore.

07:22 Also exploring kind of what exists on the market.

07:24 What are other people building? a lot of people approaching that yeah and i think that also

07:29 brings me to my next question so my first question that would be in open source i think it was a part

07:38 of last stream when do you pair up with other people when does it make sense how do you pair up with other people? When does it make sense?

07:45 How do you figure that out?

07:47 Like in open source, like you see a good repository,

07:49 you could see good people working on something,

07:51 but they might have a different approach,

07:52 maybe different technology used,

07:54 maybe different philosophy on how to build a product

07:58 or all that.

08:00 So.

08:03 I don't think I've seen that. can you make an example of someone who's

08:08 done this in the past because i don't really have seen that for this case uh done not yet

08:13 um i think i only know omar who's approaching that there are things like of course next cloud

08:20 on cloud who have parts of the solution so they're often like parts of the solution

08:28 yeah um i don't know i don't think i've seen a lot of people like straight up just combining

08:35 multiple repos into one thing and being widely successful with that typically you just need to

08:39 build your own stuff a lot of times. Technical challenges and you know,

08:46 a lot of, yeah, it's just,

08:47 I think the complexities is very hard

08:50 unless it's like very, very similar code bases

08:53 and it's low maintenance to keep it up to date.

08:57 You're basically, if you take an existing repo

09:02 and you kind of like an exist into a new package

09:06 now you're like a fork technically of said product but you're also a fork of all the other

09:13 things that you like batched packaging in together so you have the challenges of being a single fork

09:19 which is already really hard multiplied by three repos.

09:29 So I don't know how you would do that unless you have buy-in from that project and like they help you run that,

09:33 maintain that.

09:36 So yeah, I don't know if that helps.

09:39 It does, maybe not fully my question though.

09:43 It's also interesting answer.

09:45 I think more of the interesting part is,

09:48 does it make sense to, for example,

09:50 build something completely new

09:52 if there are already some existing solutions

09:54 to some of the problems?

09:55 I see, yeah.

09:59 In general, I think it depends what you wanna do, right?

10:02 If you wanna start a company

10:03 and you have this really, really strong conviction about a problem and everyone else is doing it wrong, then you

10:08 should definitely do it yourself. But if you are just, let's say a consumer or an engineer,

10:14 you should probably just use what's out there and help make it better. I think open source

10:18 has always been like, it's let's make an example with like Node.js versus bum right like or or some other runtimes

10:26 like the engineers of bum you could have easily um just made prs for node and and hope they get

10:34 it merged but they've had a fundamentally different point of view how javascript runtime should look

10:39 and work and so they just started from scratch and rebuild everything themselves and try to compete

10:45 on a fundamental same level. So I think it's kind of like build versus buy and buy is like

10:52 contribute to open source. It's like, it really depends if you want to be part of a larger

10:58 community and you contribute, or if you want to say, okay, I'm doing a completely different, and I'm starting my own project.

11:08 For example, for cal.com, I don't think I'm biased, but I don't think it makes sense to start another scheduling solution just because a single thing doesn't work or a single integration is missing.

11:24 Because of all the other things now you have to take care of,

11:27 now you have this slippery slope of all the different problems.

11:31 And so, yeah, a lot of times it's just easier to fix what's broken

11:34 instead of restarting from scratch.

11:39 But again, it really depends on your conviction as well.

11:43 If you want to compete, you should compete with anyone know you should not be too afraid of larger projects out there

11:51 thanks um maybe something we can go into more

11:56 cool ben you want to go next yeah let's do that yeah my question was around

12:03 around when not to go for open source.

12:06 So now I'm not talking about generally.

12:08 I'm talking specifically within the stack.

12:11 So for example, my default is open source.

12:14 I just like it.

12:15 I enjoy it.

12:16 I enjoy when other projects are open source.

12:18 So it's just like kind of straightforward for me.

12:22 But I would like to pick your brain on, do you have a mental model?

12:25 Like, are there some parts of the stack

12:27 that shouldn't be open source?

12:29 And how do you think about this?

12:32 Yeah.

12:34 So I had this conversation with Omar earlier today,

12:39 and I don't believe in this kind of like BSD license

12:41 or the new license that like Sentry came up with,

12:44 where like you start with

12:45 a commercial core and after two years or so it becomes freely available. Me as a maintainer,

12:52 I never want to have users be on an old version because they have bugs, they have security

12:57 vulnerabilities. So it really doesn't make sense from an open source point of view to say like,

13:02 oh yeah, it's going to be open source, but like you need to wait two years like two years talent didn't exist right like that's the times

13:08 that we're talking of tech and so if you are now going to the free the very first version

13:14 it's dark and so um i don't want that for my for my community so um i think that's not an

13:23 option to start commercial and make it open source at all.

13:26 I mean, think about how bad ChatGPT would be if it has a two-year delay for open source.

13:31 Yeah, yeah, yeah.

13:31 Very fast.

13:32 So now the question is, okay, but how do you actually commercialize?

13:36 And to me, that's open core.

13:37 I'm a very strong believer in open core because what open core means is you have a 90 something percent freemium code base that like hobbyists or small companies

13:47 or startups can use for free if they set it up themselves

13:52 and no strings attached.

13:53 I do like AGPL because it means whatever you fork,

13:58 you need to keep public and contribute back

14:01 to the community and the community's work itself

14:03 is protected as well,

14:05 which means that if you're a contributor

14:07 and you don't like Facebook or you don't like Amazon,

14:10 you can be very sure that if that company

14:13 forks the code base that you just contributed to,

14:15 they won't make an AWS Elasticsearch or similar, right?

14:20 Which happened in the past,

14:22 where just a big corporation comes,

14:23 takes the MIT code base and makes a commercial product like happens all the time we say literally they can whatever

14:30 so and this to me yeah i'm sorry i just want to double click on this like open courting i think

14:36 to center kind of like this question like what is the core what is then the core you have a mental

14:41 model to this what is the core what is not the core basically i yeah i have a very good mentor model um and i it was partly inspired by figma which is

14:50 by the way not open source um but so figma has a very interesting notion as well and a very

14:55 interesting um go to market where like every single player feature is free for yourself but

15:02 the moment you need to collaborate and you need to pull in your team,

15:05 that's where you need to join an organization,

15:07 you need to pay for a seat.

15:09 And to me, when people are looking for open source solutions,

15:14 they typically have two reasons why.

15:16 They have maybe a cost attached to it.

15:20 There's many other reasons,

15:21 but let's zoom into the cost attached to it.

15:25 And they need something that they maybe can change code or like contribute to a toy around with, learn something.

15:30 For me as a business and take a look at Notion and Figma, they're doing great.

15:34 I don't really see that as a commercial customer. Right.

15:38 So whether it's open source or it's closed source, I'm giving it away for free already.

15:43 So if you can strike that balance, it's really easy to easy to think about okay what does an individual hobbyist engineer need if you're an engineering

15:50 focus maybe you're not engineering focused but then also what is it what is an engineering team

15:54 what is a a marketing team what is a product team and so um we've made for us internally it's a big

16:02 topic as well but we have this this thing where we say single-player multiplayer and like single-player is free

16:09 And then multiplayer is where you pay a sassy because that's where you have collaboration and for Calicom

16:15 It would be around Robin links and routing and team events and all that stuff and team availabilities and managing that

16:23 so from a mental model, it's really easy because

16:27 you just say, okay, whatever single player is free, take what you want and contribute back.

16:32 Same with single sign-on. We don't have single sign-on in our enterprise license because we hate security.

16:38 We just don't believe that a single hobbyist needs single sign-on.

16:42 Right?

16:48 that a single hobbyist needs single sign on right like i don't double click one more time into and go again a layer below um so i get the mental model i think i think it's amazing um and how

16:54 do you then like so does that mean that the cloud.com code base actually like is there really

16:59 a one-on-one like mapping right now on what you told me like the freemium features and and the codebase

17:05 that is live in the public so is that is that how it is right now so i would say very much um

17:14 as close as we try you know there's always um there are always components where you'd say okay

17:20 this is something yes this could be a single player but only for the head of

17:25 operations at a larger corporation so it's like it's not a multiplayer commercial it's not a

17:30 multiplayer collaborative feature but it's definitely not something that you need as an

17:34 individual hobbyist or single model that like pop store so i do think like advanced workflows

17:41 or routing forms technically you could use that as a standalone person,

17:46 but you would never just wake up one day and be like,

17:49 okay, I need this for my side project, right?

17:51 So, and the reason this mental model works out is because

17:57 if you were to build this product in closed source,

18:00 you would probably keep the same mental model as well.

18:04 It's like, how do I make my freemium plan?

18:08 And then how do I make my teams plan,

18:09 my collaboration plan,

18:10 and how do I make my enterprise plan, right?

18:13 Sure, sure.

18:14 And just to go last one,

18:15 which I hope will be shorter,

18:17 a last double click.

18:18 Does that mean like really like technically

18:20 you have like closed repos that cover those features

18:23 and then you kind of need to plug everything together from an engineering point of view you

18:27 kind of need to deploy everything separate it's like micro services or

18:31 something like this yeah it's a great question we actually had a similar one

18:34 last time we talked so the best way to go over this is we have a top-level

18:39 license from a like a license point of view that says everything is AGPL

18:45 unless the content in these folders, right?

18:48 And you have these.

18:50 Right, right, right, okay.

18:51 And you have these commercial enterprise edition EE folder.

18:56 And so if you go into one of these folders,

18:58 A, you'll see a different license.

19:00 You'll also see a different Redmi.

19:02 So you have like this enterprise edition and you can get a license key very easily and then there's a license

19:08 that is a commercial license yeah so technically it's available but it's not

19:15 open source which is also great for security audits it's all public but it's

19:21 any as you can see you know we have managed team events we have

19:23 organization impersonations we have teams we have workflows so it's it's, and as you can see, you know, we have managed team events, we have organization, we have impersonations, we have teams, we have workflows.

19:29 So it's very easy to also take a look at the code base

19:32 and see, okay, I understand what's commercial.

19:35 Yeah.

19:36 Right.

19:37 That was it.

19:38 Thanks, man.

19:39 And we've actually, funny enough,

19:41 we actually have started out with a private fork

19:43 and set like everything on calenso.com was our private fork and it had like a little bit more features

19:50 and things.

19:51 But it was just such a merging hell, like long term, it's just way too much overhead

19:57 and you're not winning anything and it feels weird.

20:00 And you have like, for example, when you have people who want to contribute,

20:07 they don't know what, like, how do they,

20:10 I have a bug here, but I don't see the code.

20:12 Right?

20:13 So you had like two remotes on Git.

20:15 Is that how you managed it?

20:16 Like you were kind of pushing it?

20:18 I don't even remember how we did it.

20:20 But it was horrendous.

20:21 I do not recommend.

20:23 Gotcha.

20:24 So I think even if i would start tomorrow

20:27 i would do the same split and gitlab has done this and they went public so i think it's a proven

20:32 proven path um they don't have agpl that's something that we brought into the mix they

20:38 have an mit license but other than that they still have an open source commercial split open core commercial split right right right thanks yeah cool but again it's always a i think that was the hardest struggle for

20:51 us to get going of like how can we actually commercialize products without giving them

20:54 away for free um the side effect is you still publish the code so people technically could

21:01 take it rip it out um this has only happened once from a company

21:07 that is at a stage where you could see of like,

21:10 okay, well, this is actually, this is actually bad.

21:13 A lot of times when people violate licenses

21:18 and this can be, you can make an example with like,

21:24 someone buys Tailwind UI and then privately shares

21:27 the key or shares the components, right?

21:30 Like happens all the time because tech is being shared and sometimes you just close

21:36 two eyes.

21:37 But the people who do that, they're not commercial large companies.

21:41 Like Microsoft would never by choice by later license they

21:46 like the engineer gets fired the head of engineering is like what the fuck's

21:49 going on the CFO is in trouble the head of legal comes in like there's just too

21:54 much risk to not to like go away DRM so the people you're selling to already are

22:00 in the in the camp of okay I have to get a license otherwise like i lose my job

22:07 the people who are like oh i found a way to like rip out all the license code and i can run it

22:12 without getting licensed that's the people who you're already not charging and right even if

22:17 you see them online and you see like oh they are violating my license you'll probably reach out to

22:22 them saying like hey how can we make this work? You know, are you low on money?

22:26 Can we give you a sponsorship?

22:27 Can we endorse you?

22:28 What's going on, right?

22:29 So I've never really seen a problem where you are like,

22:34 oh, but what if they take my commercial code

22:37 and run with it?

22:39 Yeah.

22:39 And worst case, you know,

22:41 you have a legal team at a certain size

22:44 and you just send out a nasty letter and say like, hey, what's going on? Like, please don't. So, you have a legal team at a certain size, and you just send out a nasty letter

22:45 and say, hey, what's going on?

22:47 Please don't.

22:48 So yeah.

22:51 Yeah, I have other questions, but I can do that.

22:53 I can relate it.

22:56 It's always an interesting topic.

22:58 And I was also, when I started, I was like, how do I do this?

23:01 And I was very private with the commercial code.

23:03 And then I was like, this is too much,

23:05 like this is too much effort.

23:07 So, yeah.

23:08 Cool.

23:10 Anything related Tom or any other questions?

23:16 To licensing?

23:17 Or getting started or how to figure all of that out?

23:21 Yeah.

23:22 Figure all of that out.

23:24 I don't want to use anyone else airtime, but I will.

23:30 We all have infinite time.

23:31 All right.

23:32 Yeah, of course.

23:35 Yeah, so other question is around sequencing.

23:38 If you have an opinion around like in the beginning,

23:42 you just have the software.

23:44 It's unimpressive, right, in the beginning. It's

23:47 the same for everyone. But if it's unimpressive and open source, and it's not like code for

23:52 developers, et cetera, there's like a question about how much value, will people pay for

24:00 this or not? So I wonder if there's a sense to eventually open source rather

24:07 to you know start open source because of you know at some point you just cross this intersection

24:12 where value in a value is provided and your thing is complex enough that like yeah it just reaches

24:18 this magical inflection point so I don't know if you have to. Yeah, I have a pretty interesting experience with that.

24:28 We started out with a paid only plan

24:30 for an open source product, which is very interesting.

24:33 So basically you have the open source code base,

24:38 but then every plan we had on the SaaS plan,

24:40 which we would host was paid when we launched

24:44 and A, you get revenue, which is great.

24:47 So you know what, if you're building something, people like it, they pay for it.

24:51 But also, you really only attract people who want to be early and gear, like, who are good

24:57 customers.

24:58 They leave high quality feedback.

25:00 They feel like they are part of a community.

25:03 They feel like they are part of a community they they feel like they are very close

25:06 to your product and you just have less people and you can really double down on their needs and

25:13 their problems and their features and their bugs then launching with a prosumer product which is

25:20 free and then you have like first off you probably have a 90% drop

25:25 off after like day one of people just signing up and clicking a couple buttons

25:29 and leaving again just because people want to try new things and see if it

25:34 works for them so and as an engineer or as a founder the early months and years

25:39 are very intense and you really want to double down on the people who like what

25:44 you're building and then eventually you want to figure out how can i find more of these people and then

25:49 eventually want to expand your audience and you know go into more later later stage um adopters

25:57 so i don't know i haven't read anything about it but it felt really really right to launch paid

26:02 only and then once we had a certain stability

26:05 and stage of like okay now we can actually we have a support team we can you know fulfill the needs

26:10 of more people we we actually we make every paid plan free right so you get an email and it says

26:18 like your 15 month plan is free now like that's sick yeah like And we did that at a point where we already had like somewhat

26:28 of enterprise traction, where it's like, you're not even using, you have a small dip, but

26:33 you're not like exploding in terms of lack of revenue. And it's a really cool email to

26:38 send because it says like, hey, like, thank you so much for being part of this. But like,

26:43 you're in it for free now because of these changes we were we are commercially more excess successful in

26:49 enterprise and mid market companies like us so here's all your free plan and it

26:57 was very interesting because our free plan is as good if not better as our

27:03 competitors pro plans because we

27:05 don't do the prosumer focus we say okay everything individuals is free no matter

27:10 how good that feature is like infinite calendars infinite whatever reminders

27:15 workflows links these are limitations that other people put on you as a free

27:19 plan and they hope you you convert to a pro plan but like from a from an

27:24 individual point of view,

27:26 it really makes no sense to upgrade

27:27 if the alternative is free, right?

27:30 And so that's been fantastic for us.

27:33 I would probably do the same thing again

27:35 and start paid only.

27:38 It's like long term.

27:39 It's only with open source

27:41 or paid only and closed source and then open source?

27:44 No, no, no, always open source from day one. Okay, always open source? No, no, no. Always open source.

27:45 Okay. Always open source. Right, right, right. Yeah. Yeah.

27:48 And what's interesting also is like, you're not really,

27:52 you're not really rejecting people who want to buy a free product because

27:56 they, they can take the open source review and run it themselves for free.

28:00 So it is,

28:01 it is more steps and you need to understand how Docker works or railway or

28:04 Versailles, whatever, but like, you're not really excluding people and and and that way you might

28:10 even drive more self-host or adoption um and more contributions because now people are looking into

28:16 your code and you drive more eyeballs to the self-hosted version because um you don't make

28:21 it like a one-click free plan on your website, on your SaaS plan.

28:26 So it's like if you want to go fast pay us, if you want to set up the repo locally, get going and set up yourself, you can always go free.

28:35 Actually, that touches on a good segue into another question, which is like, yeah, if everything is super easy to deploy,

28:41 especially in the beginning, then it's like, and then you have not that much to deploy

28:45 and you click with Railway.

28:48 I mean, I was like thinking,

28:50 do you do that in the beginning?

28:52 Do you give this one click deploy

28:53 or should you wait a little bit to give this option

28:57 to people because I want people to deploy my thing,

29:00 but eventually, eventually.

29:03 No, so we have a two command line

29:08 to get it up and running locally

29:10 and probably like a three command to run on a server.

29:13 The reason being is, and I thought about this as well,

29:17 whether other people do that or not,

29:18 but like, I think you really wanna make it

29:20 as easy as possible to get going with your software

29:24 because of contributions,

29:26 but also eventually you have paying customers and they pay you for a license and they don't

29:31 want to be dealing with all that hustle.

29:34 So for us, TTV, time to value has to be as short as possible.

29:39 So we have had Deploy in Brazil, Deploy in Heroku, Rest in Peace in the very early days

29:46 just because we wanted it to make it super easy.

29:50 The only caveat is a lot of these platforms

29:53 are so fucking different

29:56 that a lot of these deploy buttons

29:58 stop working eventually.

30:01 And so you need to put in a lot of testing

30:03 and community help to keep them up to date,

30:06 you know, when like, environment variables change, or how large a serverless function can be,

30:12 like every service has to be. So that's been something which we've been,

30:17 some we've been really struggling with, because of how quick our system works. And, you know, we can only test on our infra, really.

30:26 Like, we can't have, you know, 10 deployments.

30:30 And so some people were like, oh, I clicked the button,

30:33 but I got an error.

30:34 And I'm like, oh, fuck, Heroku shut down.

30:36 You know, like, fuck that.

30:39 So that's something to be mindful of.

30:44 But that's also something you can outsource

30:46 to the community.

30:46 If you find someone who's like a Heroku fan boy

30:48 or a railway fan boy, and you tell them like,

30:51 please click this button every two months

30:53 and let me know if it still works.

30:57 Right.

31:01 We still haven't done all the buttons,

31:02 but by now we have a lot of buttons that deploy Calendly.

31:05 Cool. Tom, are you going next? Oh, yeah?

31:11 Oh, sorry, yeah, I'm muted. There you go, I'm muted.

31:16 It's me, I don't know, sorry.

31:18 Yeah, okay, okay. So going back to the licensing,

31:29 do you use like some type of software that do the tracking of some infringement

31:32 of those licensing or?

31:38 We actually, we did have telemetry baked into the product,

31:44 into the code base initially

31:46 with a way to easily remove it of course,

31:48 with like an environment variable.

31:51 I don't remember if it was opt in or opt out,

31:53 but I do think we actually removed it

31:55 because it was kind of buggy.

31:56 I don't think we've brought it back.

31:59 So we have somewhat of an idea

32:01 who hosts cal.com on their domain,

32:03 because it just, you can report like, I'm using this on my own, whatever the domain.

32:09 We've never really invested time to see if they actually have a license or not.

32:13 As I said before, the large companies don't violate and the small ones you don't charge.

32:20 There's probably a ton of people that we don't charge, probably. But the total amount of revenue is neglectable to a single enterprise

32:29 to be able to close.

32:33 I think I saw a statistic from GitLab's last quarter,

32:40 or maybe two quarters ago, where 90-h somewhat percent came from like the top one

32:47 percent of customers and then rest of the 10 came from the rest of the 90 maybe 80 20 i mean

32:52 typical pareto stuff but that was really like unsurprising to me um how to me prosumer doesn't

33:00 really work and like mid-market and i mean just think about like to make a

33:07 like how many subscribers you need to make a million a year in revenue like that's just

33:11 fucking insane like if you have a prosumer product like i've seen sas apps where it's

33:16 like three or five dollars a month and i'm like are you gonna have like i don't know 200 000

33:23 paid subscribers like that's fucking rough

33:25 and so um yeah i i've never been a big fan like especially once you build something that

33:32 um people really really want and like and are comfortable with um companies are nothing else

33:38 than a bunch of individuals collectively sitting in one room or remotely. So I try to get like six figure deals coming in very early,

33:47 you know, like starting at like a hundred something,

33:49 maybe 75 K in the beginning, going up to like 150 or so.

33:55 And you also need to really be confident in your product

33:59 that maybe you raise some funding and you have some runway.

34:02 But like, I would rather start the enterprise

34:05 or like you know six figure sales muscle earlier than later because you need to learn so much as

34:12 a founder that like if you postpone it you just postpone your learning and you probably get a ton

34:17 of rejections initially but at least they reject you because of the price or maybe because you're

34:22 not like early enough and then that way you just

34:25 reach back to them a year later and say like hey listen like we're one year older we're still

34:29 growing and and we are 20 cheaper now because of xyz or maybe you're more expensive because you

34:36 have more features right um so as long as they reject you based on like you're not old enough

34:41 and you're like um too pricey or so that's good reasons if you want

34:48 to be rejected because your product is right and then you make your product better um but if

34:53 that's the two reasons i wouldn't like yeah i would just go for enterprise way way earlier

34:59 than others if you have a commercial open source product that makes sense for that market right

35:01 than others. If you have a commercial open source product

35:03 that makes sense for that market, right?

35:08 Yeah.

35:09 Okay.

35:10 Sounds good.

35:11 Tom, you wanna go next?

35:12 Yeah, sure.

35:13 That's, I think the community building

35:16 is an interesting part,

35:17 especially if you're not building a developer tool.

35:20 So how do you build a community

35:22 for non-developers tool software?

35:24 I think cal.com kind of falls into that category as well.

35:30 Yeah, it's kind of like, how do you define non-tech?

35:33 Because today I think every industry,

35:35 like there's not like tech and non-tech.

35:37 Like non-tech to me is like people using a fax machine,

35:41 but like if you're a carpenter,

35:43 you probably have a CRM of people. If you are a carpenter you probably have a crm of people if

35:46 you are a dentist you probably have a dentist booking system if you are a whatever come up

35:52 with an industry yoga teacher so like i do think every industry has its own set of engineers and

36:00 maybe it's an agency that sells to these industries, but I don't really know any industry where tech does not play a larger role in it.

36:10 And so you always have somewhat of people who are in that industry interested in that.

36:16 Especially now with low code and no code and Zapier and all that stuff as long as you have these kind of like integrations into zapier and and n8n and and make

36:27 um like we have lawyers who have never written line of code but like embedded our embed into

36:34 their wordpress page so um yeah i think everyone not github i get you that so like github is mostly for engineers but i do think that um you can build in

36:47 any industry because any industry has technical people so it's just a matter of like how big is

36:53 that industry yeah um just i agree um and the the developer community on github is probably harder

37:03 and maybe approach differently.

37:06 Because I feel like if it's like a very technical solution,

37:09 like it's a very technical tool,

37:14 hyping up a community about it, for example,

37:16 is much faster or that is much easier,

37:19 assume compared to like a calendar solution.

37:24 Like there are a lot of technical challenges,

37:27 but maybe not as obvious or as exciting for developers.

37:30 So how do you get developers excited?

37:34 I think I would even say the opposite is true.

37:37 I think very few people are skilled enough

37:40 and interested enough in really,

37:41 really hard engineering problems such as BAM noise,

37:44 or what is the programming language called? Zip noise, something similar. enough and interested enough in really really hard engineering problems such as bun noise or

37:48 what is the programming language called zip no something similar

37:58 like sure technical problems attract technical people but i think you know so many of us work in

38:04 industries where you deal with time zones where you deal with digital signing like document or forums

38:05 or workflows or whatever Salesforce alternative 20 is a great CRM as well so

38:15 I think people want to be people just want to work on interesting problems and

38:20 interesting problems show up in any industry you can probably launch a open-source analytics product for wind parks right and you would attract you

38:33 know the freaks of the wind park industry that might just die hard about

38:37 your solution it's a smaller niche but nonetheless it exists and and people will you know like for example we have 500

38:47 something maybe 600 um contributors and part of them care about calendars but part of them also

38:54 only care about like hard engineering problems and you have them in most industries so you don't need

39:00 to be very technical you don't need to work on like a database like super base or

39:05 or bun or um whatever trpc to be to be an interesting guitar project and also last

39:14 thing i also think the more technical problems are actually way way harder to monetize because you're

39:21 pretty much always only working with engineers and engineers don't want to buy things and because we're just used to like yarn install

39:31 anything or bond install anything and we're like oh it's free you know someone

39:36 has made it why should I pay for it and so if you're just a package if you're

39:41 just a runtime I mean kudos to bond i'm very happy for for the them to

39:47 to exist and we're really trying hard to get done integrate to run calicom on bun but um

39:54 trpc as well it's like it's a really hard um product to monetize really because it's

40:03 it's or react like who the fuck monetizes react

40:06 facebook won by far because of their tech influence and their hiring potential and unifying

40:12 their tech stack among apps like i think that's the biggest driver of this but like

40:18 i don't think you can really charge for react Like imagine you as a developer would need to pay a license for that, right?

40:28 So, or TypeScript, or TypeScript, like no.

40:35 Don't see that being successful.

40:38 So I'm a big fan of, me and JJ called it APFA,

40:41 where it's like you build an application

40:43 and you sell the infrastructure.

40:50 And this is true for, it's also true for private companies, closed source companies, you could make the argument that open AI's chat GPT is a great example of APFRA because they

40:56 have a free, almost free consumer facing app. And then they sell the API to any developer on a per-token business model, right?

41:07 And to me, they are only as successful

41:12 in the developer space because of all the brand

41:14 and also all the improvements done by the community,

41:17 by the consumers, right?

41:19 So for us, we didn't start selling the infrastructure

41:22 to larger companies before we had like a million

41:25 bookings done by consumers, right?

41:28 And so we were comfortable saying that, okay, whatever we sell to business will work because

41:31 of we've tested almost every edge case.

41:36 And so if you have a pure developer product and you're only selling to developers, I think

41:43 you're also only charging developers.

41:45 And so that's something which I find

41:47 very interesting as an industry.

41:49 Me personally, I really like AppFarm,

41:51 building free apps and selling infrastructure.

41:54 That's been really fun.

41:58 Maybe on that note, also interesting,

42:03 I think we talked about it before,

42:04 but approaching like individuals

42:06 and charging them in the beginning might be good but in the long run not really worth it

42:12 so is it maybe as a strategy to say okay right from the beginning we approach bigger corporations

42:18 and the prices we try to get deals with them maybe even before we have the actual product

42:24 try to get deals with them, maybe even before we have the actual product. And then with the more guaranteed revenue builds and invest in that.

42:31 What are your thoughts on that?

42:33 I mean, selling to companies is interesting, larger companies is interesting because the

42:41 lead time is taking time so you can start selling before it's ready

42:45 but um eventually you need to do demos and and get procurement on and build like give them a trial

42:53 and handhold them so like um and also the biggest your biggest enemy starting a project is actually

43:01 your immaturity as a company no matter matter how good the product is, a buyer looks

43:08 at your company and says, how many engineers do you have? How big is the team? Can you sign an

43:12 SLA? Do you have a support team? Can we do redlining? Who's your lawyer? Who's your ex?

43:17 Who's your, what's your insurance? And so your biggest enemy is actually time. Like, even if

43:22 you have all of those, they'll look back in your company registration or they ask you and say okay yeah we're like six months old they're like

43:29 all right okay let's come back in two years so like that happens a lot we're now closing a lot

43:35 of contracts that we've been in conversations for two years just because the buying person

43:42 is another employee who just doesn't want to lose their job, right?

43:46 They don't have any incentive to move fast.

43:48 They have an incentive to keep their job.

43:51 And so for them to take a risk, they don't reap the benefit.

43:55 A founder reaps the benefit.

43:56 If a founder takes a new product

43:59 and gets a 50% performance boost, that's awesome.

44:04 Or saves, I don't know, 20% of costs.

44:08 But if a generic procurement manager

44:11 or even head of engineering, head of product

44:15 from a larger company, as you said,

44:16 like looks at your product, they just see risk.

44:19 They don't see the upside.

44:21 And so your enemy's really time.

44:23 And the way you beat that is to have a good shipping

44:27 cycle we have a monthly cadence of every 15th of the month by the way in nine days go go subscribe

44:34 important time in every 15th of the month we launch a version and we let the person know here's

44:40 a new version here's what's better, cheaper. And that way you keep them

44:46 in the loop. And then eventually they realize, oh, wow, you have SOC 2, ISO, HIPAA, all of that

44:52 stuff. And you have several existing customers who have testimonials and all that stuff. And

44:59 then the buying decisions are easy. But if you're just getting started, it's really hard.

45:05 You still should try, you should still jump on every call

45:07 and learn as much as you can and build the features

45:09 that people wanna pay for.

45:11 But don't think you can make revenue early

45:15 by closing an enterprise deal upfront

45:19 and then not having a product ready.

45:21 So that's really hard.

45:23 Thank you.

45:26 At least from my experience. There's probably people to pull that off so we have three more

45:34 minutes yeah was pretty quick was a quick quick session this time. It felt like time flying. Anyone has a final burning question before you go to sleep?

45:49 No?

45:52 Oh, yes, you're up. Okay, so going back to my first question, I don't know if I

45:59 got it or not, but what I was trying to say is going open source wouldn't eliminate the possibility

46:06 or make it much, much harder to maybe sell the company or even for investors like getting

46:15 a seed round or stuff.

46:17 Yeah, I remember postponing that question.

46:21 Good thing you brought it up again um i have not felt any problems

46:27 raising i mean also was a different market but even today i think open source companies can can

46:33 very raised i mean it's one of the largest um uh one of the largest industries in yc right now and

46:42 i think yc is a good training indicator of what investors like to invest in

46:46 um it's a good startup index and so i would not be afraid of investors

46:51 when it comes to m a um i think the industry is just very young i know open

46:56 source is very old but i think commercial

46:58 open source is very young so we haven't really seen much

47:02 m a or or ipos besides you know GitLab and I mean

47:06 there's many others I'm not saying there's none but like when it comes to

47:10 SaaS I think there's a lot more it it really depends what the mindset of the

47:16 buyer is like whether they understand open source I think a lot of companies

47:20 understand the power of open source and acquire open source because of its open

47:24 sourceness.

47:30 When you deal with a traditional buyer that has no idea about open source, they probably see it as a risk. But to be honest, as a founder, do you want to be selling to that type of person

47:36 in the first place? Maybe if your company is dying, so you take any deal. And if your company

47:42 is dying, you probably leave on bad terms so the company who's acquiring

47:47 the company the open source code base will do whatever they want anyway like so they take it

47:52 private they commercially they change the license they kick you out you have no controlling stake in

47:58 the company anymore so if you're winning no one cares if you're losing, no one cares. If you're losing, you're losing anyway. So like, yeah, I don't think I've ever seen someone

48:09 being afraid of buying an open source code base,

48:11 to be honest.

48:16 If anything, if you're successful and open source,

48:18 I think you have a very, very strong mode

48:20 because you're the only one doing something.

48:23 Open source has very strong network effects

48:25 like think about next js right like if reverse cell eventually whatever is being bought by

48:32 something they buy probably the largest react framework ever in history um with the largest

48:38 community like you would probably pay a really really decent dime to to get that much uh engineers right um so i don't i'm not

48:49 concerned if you're winning i think you're in a very good position and then if you're doing if

48:53 you're going sideways or losing you probably are in a bad day in a bad place whether you're open

48:58 source or not so yeah it's very very For fundraising, there's probably people who don't understand open source