320 - John Nunemaker of Fireside.fm, Box Out Sports and Very Good Software

In this episode I talk with John Nunemaker about his journey with Ruby, acquiring Fireside.fm, and his ventures in software focused on podcast hosting and feature flag services. We also explore the potential impacts of AI in the development space and John's vision for the future.

Links:
- Box Out Sports
- Very Good Software
- Flipper Cloud
- Fireside.fm
- Nonsense Monthly

Transcript

The Snail Mail Newsletter Pitch

SPEAKER_01 0:00

Hey, it's Jason, host of the Code with Jason podcast. You're a developer. You like to listen to podcasts. You're listening to one right now. Maybe you like to read blogs and subscribe to email newsletters and stuff like that. Keep in touch. Email newsletters are a really nice way to keep on top of what's going on in the programming world. Except they're actually not. I don't know about you, but the last thing that I want to do after a long day of staring at the screen is sit there and stare at the screen some more. That's why I started a different kind of newsletter. It's a snail mail programming newsletter. That's right. I send an actual envelope in the mail containing a paper newsletter that you can hold in your hands. You can read it on your living room couch, at your kitchen table, in your bed, or in someone else's bed. And when they say, What are you doing in my bed? You can say, I'm reading Jason's newsletter. What does it look like? You might wonder what you might find in this snail mail programming newsletter. You can read about all kinds of programming topics like object-oriented programming, testing, DevOps, AI. Most of it's pretty technology agnostic. You can also read about other non-programming topics like philosophy, evolutionary theory, business, marketing, economics, psychology, music, cooking, history, geology, language, culture, robotics, and farming. The name of the newsletter is Nonsense Monthly. Here's what some of my readers are saying about it. Helmut Kobler from Los Angeles says, thanks much for sending the newsletter. I got it about a week ago and read it on my sofa. It was a totally different experience than reading it on my computer or iPad. It felt more relaxed, more meaningful, something special and out of the ordinary. I'm sure that's what you were going for, so just wanted to let you know that you succeeded. Looking forward to more. Drew Bragg from Philadelphia says, Nonsense Monthly is the only newsletter I deliberately set aside time to read. I read a lot of great newsletters, but there's just something about receiving a piece of mail, physically opening it, and sitting down to read it on paper that is just so awesome. Feels like a lost luxury. Chris Sonnier from Dickinson, Texas says, just finished reading my first nonsense monthly snail mail newsletter and truly enjoyed it. Something about holding a physical piece of paper that just feels good. Thank you for this. Can't wait for the next one. Dear listener, if you would like to get letters in the mail from yours truly every month, you can go sign up at nonsense monthly.com. That's nonsensemonthly.com. I'll say it one more time. NonsenseMonthly.com. And now without further ado, here is today's episode. John,

Meet John And His Companies

SPEAKER_01 3:53

welcome. Hi. So this is your second time on the show. Your first time was a long time ago. I don't know how long ago. But um for anyone who doesn't remember who you are or hasn't heard that episode, I personally don't remember who you are. So if you can re-introduce yourself to me and everyone else.

SPEAKER_00 4:14

Yeah, sure. I'm John Inamaker. I've been programming forever, uh probably like 20 years, and almost all of it in Ruby. So a little bit of other languages at the beginning, but about 2005, you know, I saw the famous um, you know, blog post in five minutes or whatever video, and uh went down the rails rabbit hole. And I've tried everything else since then and just keep coming back to Ruby. So big fan of it. And I've got uh a couple companies, so one is Box Out Sports, it's like sports marketing, social media graphics, all that kind of stuff, serving high schools and colleges, and then another one is very good software, and that has two apps in it right now. One is FlipperCloud, which is feature flags, and the other is fireside.fm, which is uh podcast host.

SPEAKER_01 4:58

So oh, interesting. Okay, I knew about that, but then I forgot about that. That is been a while. Uh-huh. Yeah. Yep. That is uh relevant to me as somebody who hosts a podcast. I'll have to check that out.

SPEAKER_00 5:13

Yeah, sure. I I'd be happy. We're doing a lot of stuff right now. Um we just redesigned the studio. Um, so that's like gonna be probably in the next day or two, it'll be opt-in for people, and then you know, in a month or so we'll just flip the switch after we've got all the possible kinks worked out. Um so that'll be big, and then new marketing site coming after that, it's almost done. And then we have a whole bunch of like little things that we've been then then we'll be able to actually work on flow, which I'm really excited about because I've got some really cool ideas about workflow after having had a like a podcast of my own, 10 episodes or so. And you know, it's just there's a lot of tedious stuff around it, so I'm really excited to kind of try and programmatically make some of that stuff easier.

SPEAKER_01 5:52

So yeah. Okay, well, this is a topic near and dear to my heart because yeah, tedium is is part of my life with the the podcast processing and stuff like that. Yeah. So that's that's something that I have opinions about. Um,

Why He Bought Fireside.fm

SPEAKER_01 6:09

but what made you want to start a podcast host in the first place?

SPEAKER_00 6:13

It was honestly, it was not even like uh on my radar at all. I think what happened for me is um I work with Garrett Diamond a lot. Garrett uh actually worked on Fireside for Dan Benjamin, who owned it originally and created it. And because of that, uh Dan had been like, he'd had it for several years and just kind of like I think lost interest and moved on to other things, and and it was just sitting there kind of chugging along, and he was like trying to find someone to buy it, and it was like too big for some of the smaller players, and it was too small for some of the bigger players, and so it was like in this weird position where it just didn't make sense for like anyone to acquire it, and so he had reached out to Garrett, and Garrett was like, um, he was like, I don't know, I'll ask John. John knows some people, maybe he would, and then I was like, I I that actually sounds kind of interesting. I was really I'm a huge podcast like consumer and fan of the format, and like had been wanting to kind of start my own, and so then I was like, Oh, this this seems like an interesting way. So we got it for you know a decent multiple, and um I so I felt good with that, and so went down that whole path of acquiring it and um yeah, all of that. So it was uh and so really like complete happenstance. It just was like, oh, it's making money, it's profitable. Um we can we're at a si like a size where we could do it, so let's just do it. Um so yeah.

SPEAKER_01 7:38

And and were you seeking a business to buy or or how did that because not everybody you know would find that a natural move?

SPEAKER_00 7:47

We weren't. Uh like I was I was actually paying Garrett to work on Flipper because I just I was like, I need to do something on Flipper, like not doing anything just didn't feel right. Like I was working on you know on the open source gem and I was you know occasionally working on cloud, but like just not really pushing it forward, and I felt like let's just really push it forward. So it was a couple years ago, and I was like, let's just um you know pay him to work on it. He was looking for stuff anyways, and so we did that for a while, um, but probably six to eight months, and then the fireside thing just kind of came up, and I was like, well, shoot, if we buy it, then you know that pays for you to work on both of them. Like, so why don't we just do that? Um, and so that's that's kind of how it came. It was it just kind of worked, and so it it made it so that I was not, you know, taking money out of my pocket every month to pay, like to, you know, to get him to work on Flipper and push that forward. Instead, it was like, well, if we just acquire this, then it can just have you know automatic income from it, um, and that can kind of pay to work on either of them. Um, and then you know, I I have I get paid from box out and stuff like that already, so I I didn't really need to have like you know a large everyday, every month thing from it. So for me, it's always like a long-term thing. I'm just always pushing on like what's the long-term result of what the work on that I do right now is. Like I don't care as much about trading hours for money and stuff like that. And so from a long-term standpoint, I I just thought, you know, podcasts aren't going down, I don't think. I mean, they just seem like more and more people are creating them, more and more people are you know finding value in them and stuff. So it seemed like a good wave, and I felt like a lot of the same things that involved in a podcast, analytics, content hosting, things like that. I've done them all before with gauges and harmony and things like that. So it just seemed like kind of a natural fit. And then, you know, I I mean I had good fortunes at GitHub and things, and so there was there I was it was possible to do. Um, and so yeah, I wasn't looking for it. I would say now, like I would be I'm more kind of like looking for it. Like I keep an eye out for like Rails apps that are for sale and stuff like that. I mean I haven't bought any more since, but if the right thing came along, I definitely would.

How To Find SaaS Deals

SPEAKER_01 9:55

So yeah, I've kept my eye out for stuff over the years just cuz um it's hard to find something that like speaks to you. Um because a lot of stuff, you know, there's like buy bizell.com or whatever it is. Um and I and I look both at software so buy buy bizell, whatever it is. Um I feel like that's mainly like uh brick and mortar businesses. Um I've I've looked on there, like looked at laundromats and uh tool and die companies and whatever, and it's like, oh hey, maybe I'll buy a tool and die company. I don't know. Um I did which I did not do. That's a good call. Yeah. And then uh, you know, there's the more software-oriented ones, I forget the names, but I've I've looked on there, and a lot of it is like, oh, here's like a supplement uh uh e-commerce site or something like that, and it's like eh, I don't know, probably not. And so it's yeah, it's hard to find something that A speaks to you. It's like, why would I do this? Um and like the numbers all look right and it just looks like a good thing and all that stuff. A lot of things have to come together uh in order for it to make sense.

SPEAKER_00 11:10

Very true, and I feel like a lot of people, I mean, I know people who have have dove into you know random businesses. I mean, my dad's a farmer, but he like randomly invested in like a machine tooling company. Um, and there was just like a ton of issues, like because things can look good on paper, but if you don't know where to look, that you know, there can be problems and stuff. So I I always say like stick to what you know, and so then finding something that you know is also very hard. Um, but yeah, there's like uh Quiet Light. I get emails from them all the time and FE International. There's all you know, a whole acquisition.com or or whatever the the one by uh Gazdecki is, but yeah, there's a bunch of them that are out there. I feel like those are not usually where you want to go to find them. Like the the way to find something like this is is network. It's like people that you know that know somebody that you know they're throwing it up for sale. And that's I I would say from the ones that I've seen, those have worked a lot better. Because, you know, you the people that throw them on those sites, they're like they expect ridiculous multiples. It's like, you know, you have whatever, you have a dying SaaS and you think you're gonna get four to eight times profit or revenue, and it's just like sorry, that's not it's not gonna work that way, you know. Like if you have a SaaS that's going down, you're probably 1.5 to 2, or maybe 2 and a quarter, or like it's not gonna, it's not gonna be that high. So yeah, finding that I think it's more just about like reaching out and talking to people a lot more and going to conferences and talking to people there, uh, you know, in various communities online and stuff, than actually on those sites. It's possible on those sites, but usually you're you're getting more of a job than just um you know, than than what you probably are looking for if you're do if you have a regular job and you're just looking to buy something on the side to maybe switch to it or whatever. So I will say the one thing I have learned is they all take the same amount of work, so make sure that it's big enough that you know, like all the paperwork and all that stuff is the same for something that makes fifty thousand or five hundred thousand or five million, like there's no difference. And so you might as well go for a bigger swing. Um, as long as you know, that's I think that's the one thing that I looking back, maybe I would point as like a learning experience.

Financing With Loans And Equity

SPEAKER_01 13:14

And I used to think that in order to buy a business you had to have like a bunch of money, like just a bunch of cash. Um but I found that that's not really the case. You can it it seems counterintuitive, but it's like you can get a loan to buy a business that makes money and then pay back the loan with the money that the business makes. It doesn't seem like it should be possible, but I understand that is a way you can do it.

SPEAKER_00 13:39

Yeah, we kind of did that. Um, like we it's funny because we were not like you when you were like, hey, do you want to come on and talk? We were not gonna talk about this at all, but like I love this stuff too, so I'm happy to go as far as you want. So yeah, we what we did is we I ended up um I my house was paid off, and so I had a bunch of equity, and I was like, I've never done a loan for stuff like this. So I was like, let's let's see what what you can do, you know. So I looked at like small business loans, but honestly, like that's to me, that would be like if I have no equity to use anywhere, that's what I would do because um it's basically I would say for that, it's like a 70, 30, 80, 20 thing. So like you can get a loan up to like 70 or 80% of it, but you need like 10% down. You need like they typically want the seller to have like a 10% seller note or or 15% seller note or something where like they're invested in in it succeeding still as well, you know, from from like a getting paid back standpoint. So you have to find the right person who's okay with not getting all the cash up front um and having some kind of like an earnout or buyout or something. So it's a little more complicated from that standpoint, and then and then the additional thing that makes it hard is the interest rate is so high. Like, I mean, I was looking, you know, so when we got fireside, that would have been like end of 24, September 24, um, October, right around that range. And the interest rates then, you know, for like a bank loan, I think we're in the fives or like low sixes, and like an SBA loan would have been like I want to say it was like 10% or 12%, like it's significantly higher. Yeah. So you want to have make you want to make sure you have like lots of a very healthy profit margin um to be able to handle that. And the loan payback period is like five years, or maybe like you know, six, seven, eight years. I think it's theoretically possible in some sense, you know, instances, but like with the what I did, um I was like, well, I have a couple different routes because I have some farmland and I have um my house, and I was like, so I just talked to the banker like straight up. I was like, well, what's you know, what can I do with either of these? And they were like, well, for farmland, we would do like 50%, like up to 50% of the equity that you have, we'll give you a loan for it, you know. So whatever. If your land's worth $100,000, they'll give you $50,000, kind of a thing. Um, and with a house, they'll do like 80%. Like, because they know houses really well. So they're like, if you want to use, you know, your your house as the leverage, then they'll give you you know up to 80% of whatever equity you have in your house. So same, same math. So if you have a hundred thousand dollar house, it's eighty thousand. They'll give you as a loan. Like the next day, no questions asked, like sign some paperwork, it's done. Um, it's they're very good at you know dealing with houses. They know exactly what they can do. It's an established, you know, houses move a lot more. They, I mean, they they uh you know change hands a lot more, so they have more updated value on them, uh, versus like land might sell every 10 years, a house might sell every two years or five years. So they get a lot more, I think, accurate numbers on what it's worth. And it's just easier for them to like you know, rent or sell or things like that. They just uh better burst. So I ended up doing the house route. Um I didn't do 80% of my house, but I did some percentage of my house, and um, and it was super easy, and I was able to get 15 years instead of you know five years or something like that. And that doesn't mean I'm gonna pay it back in 15 years. I mean, I would certainly hope to pay it back in five, six, seven, eight. That's the plan. Um, but it's nice to have the lower payment just in case, you know. Um, so yeah, that's kind of how I went about it and definitely learned a lot doing it. I'm glad I did. There's times where I'm like, I wish I didn't have this stupid bank loan, like I should just pay it off and like move on with my life. But I feel like it's it's kind of good to force the business to have at least at a minimum that much profit. And then it's also interesting because, like, you know, like we said earlier, like the business is just paying for itself. So as long as I never the business doesn't go straight to zero and I have to start paying this, like um, it's literally like I I did nothing. I stapped my fingers and and acquired, you know, like a a good chunk of uh profit cash flow automatic every month. And that's kind of neat.

SPEAKER_01 17:47

And

Assessing Fireside And Market Risk

SPEAKER_01 17:48

when you bought Fireside, what exactly did you buy and how did you approach that whole thing? Like, did you buy it and kind of assess the product and the business and stuff like that and say, okay, here's where I want to take it from here. How did you approach that whole thing?

SPEAKER_00 18:04

Yeah, it was pretty, I mean, probably more nonchalant than it should have been. Um, like I think because I knew Dan and I uh a little bit and I knew Garrett, who knew Dan a lot, I knew the business and I knew what you know what the story was. Like that it's basically, you know, it was it was going down, it was um you know not being as actively maintained and things like that. So there's not a lot of changes going on. And so because of that, I thought, well, okay, I can look at the trajectory that it's on, I can kind of estimate what it's worth to me. So just from a cash flow standpoint, you know, it made sense. I was like, you know, if I if we just didn't pay ourselves at all, like it would pay itself back in like two years. So like I was like, that's that's fine. You know, I know that it's not gonna go to zero in two years, and we can pay ourselves, and then ideally we add new things to it, we work on it, and then we change the trajectory so that no longer is it going down, but now it's flat, and then eventually it's going up and all that kind of stuff. So that was kind of the one of the main things, and then you know, just looking at from a cash flow standpoint, I was like, it won't be a poor investment from that standpoint. So then it was like, okay, well, what can we do and can we turn around? Can it be a you know a a really nice business? And like, I don't know. I mean, I I just felt like with podcasting going the way that it is, the the only threat to that would be Apple and Spotify get in some kind of big fight and like decide that they're gonna try and take over the whole market, which Spotify is kind of trying to do.

SPEAKER_01 19:29

Um, and that's really like what do you mean by podcasting going the way that it is?

SPEAKER_00 19:35

Oh, I just mean uh like there's always more podcasts, they're getting more popular. It just seems like a trend that's increasing, not decreasing. Um, so looking at the trend overall of podcasting, seeming like it's on the uptake still, um, you know, the it was meteoric, you know, around COVID and all that kind of stuff, but it's still, you know, uh I feel like a very hot or warm trend. Everybody wants to have a podcast these days. So that's what I mean by that. And then, you know, on the Spotify front or the Apple front, like Apple just wants to be the directory and they just want people to buy iPhones and use more things. They're not trying to like dominate podcasting or anything like that. Spotify, I think, kind of is, and so they're buying up, you know, they bought uh Anchor, they bought like a bunch of different um megaphone, like different podcasting companies immediately kind of internalize them into Spotify. Um, so they're I think they're probably the maybe the I don't know if I'm accurate, but um, I feel like they're kind of the biggest threat because they could, you know, be like, well, we're not gonna like distribute indie podcasts anymore. So like RSS doesn't matter. You have to be on Spotify's platform. If they got to that size, that could be, you know, an issue. Um, but I think there's a lot of the indie, you know, transistor, buzzsprout, like, um, and there's several more of them that are not um in that ecosystem. There's hopefully enough and they provide enough value that people will keep using it and and that won't happen. That's really the only threat that I saw would be like, you know, horrific or would maybe really be a problem for me. Um but for the rest of it, you know, I was like, uh it's just a good business and it's fun. Uh it's like fun people to serve. Like I think that's the thing that's you always want to think about when you have a business is like who are you serving, who are your customers, and are they, you know, is it fun to do things for them?

SPEAKER_01 21:22

Yeah, yeah. So I talk about this sometimes um with a business I used to have, scheduling software for hair salons. It's I mean, those people are fine people, but they're like not my people, you know? Um and now I'm building this CI platform and developers are my people. It's like who I'm talking to all the time on my podcast, and I already go to conferences and stuff like that. Makes so much more sense for me to do this than than to do scheduling software for hair salons. Uh so I totally get that.

What A Podcast Host Actually Does

SPEAKER_01 21:55

Um what is the scope of the product? Because like I use for Hosting, I use Buzz Sprout, I have for some number of years, um and and it's fine and good for that. Um I recently started using Riverside because a couple people pointed out that when we use Google Hangouts for the recording part of it, um the the connection can drop and really negatively impact the audio quality and the listening experience and stuff like that. So I'm like, okay, I gotta switch to something that that gets around that and you know this product records locally so that connectivity issues don't have that effect. Um and once I got into Riverside, I'm like, okay, this is interesting. The the like topography of this is much different from Buzz Sprout. They're like not trying to do exactly the same stuff. It's not exactly that it does the same stuff and more, it's that it does overlapping but kind of different stuff. It looks like they do hosting too, but the whole thing is just the like conceptual model of it is totally different. And so I'm curious about Fireside. What is what's the scope of that?

SPEAKER_00 23:07

Yeah, we would be more akin to Buzzsprout. So like the record we're not doing we're not helping with the recording or like any of that kind of stuff. It's just like make it easy to get a website, to get an RSS feed, to upload your things, put in details, you know, submit to Apple, Spotify, Amazon music, make make all those connections and things like that. So it's it's just and and also like workflow, I guess, you know, beyond that. Like, I mean, anyone could just make an RSS feed manually by hand and put you know audio files on S3 or R2 or whatever, and it would work. I mean, that would totally work. That just doesn't work for like non-technical people. So, you know, you're you're typically it's like non-technical people or technical people who aren't interested in uh doing all that stuff by hand. Um, and so like our our you know Riverside, I mean we used Riverside for quite a while for like first like seven or eight episodes, I think, to record. Um now we have um an entrepreneurship thing that I'm involved in here. We have like a co-working space and offices and all that kind of stuff, and part of it is a fireside podcast studio, and so we have a studio with a couple mics, and so now we've started using that instead, and that's definitely up. I'm actually not in that right now, and so like my mic is not gonna sound as good, and it's killing me on the inside. I need to get a good mic at home. I just have like a Yeti or whatever. Um, but um, yeah, so I would say we are not involved in that, so Riverside would have overlap in that they do hosting also, but that's just something new, and I think they basically just did it because like why not? Like, we're already producing all the things, let's just make it easy to like export it to our own platform. Um, and and probably a bunch of definition of easy, yes, yeah. Yeah, and probably people ask it for themselves.

SPEAKER_01 24:47

Yeah, no offense to Riverside, but like now that you say that, it makes sense because like I haven't seen it before like a couple weeks ago, and I'm like, okay, well, what if I just want to have one platform right now? I'm like recording on Riverside, yeah, and then uploading to Buzz Sprout. Can I just do it all in one? And it's like there's not really an easy path from recording to publishing it. Like, why can't I just like hit publish on the recording and like that's it, it doesn't seem like all integrated with itself yet. Maybe they're working on that, but it doesn't seem quite there yet.

SPEAKER_00 25:19

You would hope so, yeah. You would hope so. Like, that's what I assumed when I heard that they had hosting as well, because they didn't, I don't think they did when I first started using them like a year ago. Maybe they did, but it was just really not obvious. But when I first started using them, they had like export to transistor and export to captivate and other things. I don't know how those worked. I didn't look at that, but like they had some of that functionality. But like it makes sense if you're if you have like a recording thing, like and you have an editing thing, you know, the next step is the finalized, you know, distribute distribution of the end product. So it makes sense from like a thing, but it's also like a large surface area to solve. Like I think it's okay if there's just some tools that are recording, some tools that are editing, some tools that are hosting. I don't think that you have to do all of them, and if you do all of them, it's kind of like the PDA of old, uh, where like it doesn't do anything great, but it does everything kind of. Um so yeah, I I mean it's possible, I think, with the right team to do them all. I don't I don't we've talked about some of that kind of stuff, but I'm like right now, you know, we had a lot of tech debt, we had a lot of like, you know, just um things to fix and get better and improve and stuff like that. And because of that, we've just spent a lot of time on on that. And and so like now we're kind of coming out of that and we're starting to put our own touch on it and and things like that. So maybe I wouldn't say we're never gonna do those things. It's mostly just like um I haven't considered it to this point because we had enough on our plate. Um, the editing is that would be insane. I've thought about it a lot, like how I would build that and do it, because it, you know, you end up with Riverside, and when I first started using it, I had a lot of performance issues using it in the web and editing, like where you, you know, you'd be editing, you trim it and do some things, and you get part way through the podcast, and then you'd like come back in later, like leave, you know, get distracted, and then come back in, and then it would just I would have a horrible time with buffering and out of sync and all this kind of stuff. It's I so I'm sure it's gotten better. Um, but I did switch to when we started doing in our podcast um room at Momentum, when I started doing that, um, we switched to descript because I was like, I just need the editing, I don't need the recording anymore. And the only reason I was using Riverside's editing was because they had it and I had just recorded there, so why not? Um but then I switched to descript and I actually uh descript I think is a lot more powerful and performant um in my experience for the editing, but I don't think they have the recording, or maybe they do, and I just didn't notice it. Um so yeah, but there's it's interesting how those tools all kind of like some of them overlap, some of them don't, and they all have their you know special sauce.

SPEAKER_01 27:50

So

Winning On Personal Support

SPEAKER_01 27:51

yeah, and uh what is kind of the value proposition of Fireside? Why would somebody use that instead of some other product that does roughly the same thing?

SPEAKER_00 28:02

Yeah, I mean, it's one of those things I feel like uh it's a like a great question. Like, do I have like an exact answer right now? Like, no. Honestly, like I feel like for me, a lot of it is just like, well, people buy from people, and that's what I even think about with AI and all that kind of stuff. Like, in the end, you know, they're gonna trust a certain person and that's who they're gonna end up buying with. They might go with you know some other random thing for like a little while, and then that'll have scaling issues or it'll have something else, and and you know, so like a lot of it is just like, well, what is the experience you want to have? Like, how fast do you want people to respond to support? You know, do you want them to actually implement you know things that you ask for? Like, because that's kind of the state that we're in right now. We're like, hey, we don't claim to know everything about podcasting or have strong opinions. So if somebody wants a way to easily duplicate episodes, like because they do the same ones every year and they just run them on a loop, okay, we'll build a feature for that and we'll literally turn it on for just them. Um, you know, if they're a big enough customer and stuff like that. So I I mean, right now I would say like the biggest difference between you know us and um the bigger companies would just be uh the personal touch of like you're literally like I'm doing the support and I'm also the person who's building it. And so like if you send the right thing and I think I can just do this quick, I just do it before I even respond. I like I'll do the fix and then I'll reply and I'll be like, here's the new feature, it's out. You just use it. Um so that's kind of what I would say is maybe the difference right now. I mean, long term I have some really uh big ideas, and I know Garrett does too. Um and by long term I mean like this year kind of a thing. But right now, in this very moment, um, it's probably more just like, hey, you're just gonna maybe you have a better chance of actually getting something that you want built, built.

SPEAKER_01 29:42

Um so yeah, yeah, I think there's something to that idea of uh people buy from people and and taking that mentality. That's that's what I'm doing. Like if you look around at certain people in in the Ruby uh ecosystem community, whatever, um you know, it's not it's not exactly a meritocracy, you know. Um there's a lot of uh cronyism, but i I I don't mean those in a bad way at all. Like it's like a benign cronyism. Like in order for somebody to do business with you, like they have to know you exist, and there has to be some amount of trust there because there's risk involved. And you know, if I hire somebody to uh put a new roof on my house, uh I'd rather use somebody who did my friend's roof already, somebody I have a relationship with because that's less risk. And I think I knew Tom Rossi for some amount of time, the Buzz Sprout guy before I started using Buzz Sprout. Um and honestly, that's the only reason I chose that. Like I didn't pick it because like I did a thorough evaluation of the options and I decided that was best. For all I know, there was something better available at that time that I chose that, but I I picked it because of Tom. Yeah, and I look at like um Avo and Adrian Morin, and he's very active in the Ruby community, and it seems to me that that's that's part of it for him. You know, people use Avo because they know Adrian, and Adrian is an extremely likable guy, yeah. And so that they're they're probably not gonna do it just as a favor for him if they don't need it, but that sure like greases the wheels and and makes that a lot easier. Agreed. So it sounds like you you are thinking about that in similar ways.

SPEAKER_00 31:36

Yeah, yeah, I've definitely I mean I've thought about even just like uh you know, for like doing big discounts for Ruby people, or I don't know, I don't know what the right thing is. I mean, podcasting's not super expensive anyway, so it's not really a a big thing. But I I mean honestly, it's more just something that I'm like, I just really like doing this. It's fun problems to work on, and it's a nice little app, and the customers really appreciate most of them are not technical, and so they really appreciate like the you know the support that's like a real person that can help them. And I mean, I how many times I've hopped on Zooms and basically helped grandmas get their podcast going and stuff, you know, um, or get their DNS correct because that you know, stuff like that's hard. It's it's not stuff that's easy for if you're not technical. Like, I don't know how they expect people to to do that, you know. So sometimes it's like I literally have to just be like, look, just here's a zoom link. Like if you're around, like let me know, and I will just hop on and on screen, you know, share your screen, click this button, and I'll tell you where to click and we'll get it right, you know, and whatever you use. Um, so there's yeah, I think you know, people buy from people. I I know Tom, Tom's great. Like, I I've got nothing against Tom at all or or um uh transistor either. Like I think they're awesome. Um I'm in uh chat room with those those people and I've talked with um you know uh a few of them from there. We've had you know um chats and stuff. So and I mostly just from like uh they're like well, they're the ahead of us and like a lot of the difficult features like dynamic content and stuff like that. And so you know, having a conversation with them, and then you get like this proud smile, like, yeah, uh well, enjoy when you work on that, you know, because like they've they've gone through it, they've been there, you know. Um I I mean it there's a bunch of really cool people in the Rails community that have this, like have platforms like this, and so um, and I don't think anybody really I mean we I'm sure we compete at some level, but like it doesn't really seem more it doesn't seem like a a hard competition, it's more just like you know, what are you doing, what are you doing, and and that kind of thing. So yeah, it's pretty I would say very friendly overall.

SPEAKER_01 33:36

Yeah, it's interesting being friends with your competitors or or whatever. Like I'm friends with uh some of the RWX guys, and they're way ahead of me in terms of product maturity and stuff like that. Um and I find myself like getting ideas from them. They're like, oh, the the the the um test execution workflow is a directed acyclic graph. That actually makes a lot of sense. Maybe I should use that idea. Then I feel a little bit weird. I'm like, am I like stealing their ideas? But everybody's stealing. Yeah, yeah. Um so that's interesting. And and it makes yeah, it's it's and and with a lot of this stuff, it's not exactly a zero-sum game, like the pool is big enough that you're not really fighting over the same people and and stuff like that.

Marketing Focused On Retention

SPEAKER_01 34:34

Um what is your like what do you do for marketing? Because you're not like targeting specifically programmers or something like that. Do you go to like podcasting conferences or like try to connect with those people somehow?

SPEAKER_00 34:47

Yeah, so I mean right now we've been no, we haven't. Um I would say the biggest thing we've focused on right now marketing is just um marketing to our current customers that we are doing things and that things are different and that we're listening to them, we're making new improvements, and so it's a lot of like honestly, it's a lot of marketing emails and things like that. So it's like, hey, we did a thing, here's the new thing, try it out, let us know what works, what doesn't, and we've try and fix it as fast as we can. And just um, you know, our our goal right now is reduce churn, you know, keep uh I mean not churn's not bad, it's like super low, it's like 2% or something like that. But it's like we want people to stay, be happy, you know, that that kind of thing. And then the next wave I think will be um a little more aggressive on the marketing front. So it might be going to those conferences, it might be, you know, I yeah, I've got some some other ideas for like niches that might work really well, um, that could be easier to grow in, that maybe are underserved, you know, by some of the bigger uh current players in the space. Um but I mean right now I would say like tech and religious is probably like those are probably the top two categories, you know. There's a lot of like Linux shows, uh big Linux shows that use us and and then a lot of uh religious organizations, uh either posting sermons on Sunday or you know, Bible in a year or all kinds of different things like that. Um those are probably the two biggest right now.

SPEAKER_01 36:08

So um something you said I thought oh yeah, I I think you're taking a really smart approach with the kind of marketing to the existing customers. Um because churn is a trailing indicator, you know? Yeah. Um people churn only after they have like decided to churn. And then you find out that they left and it's too late to do anything about it. So like getting ahead of that and uh you know, maybe there's a certain rate of potential churn that exists, and you're like dialing down that potential churn so that it doesn't happen. Um I'm I'm doing an approach, and I'm actually curious to hear your thoughts on the approach that I'm taking with how I'm how I'm treating my funnel.

Bottom Of Funnel First

SPEAKER_01 36:59

So I'm I'm taking like a I call it a back-to-front approach, or you could call it a bottom to top approach. Um I'm focusing first like on the end of the funnel. So like when somebody is using Saturn CI, uh my my primary focus is like let's get this person to not leave. Um and then only once it's working for them do I try to get more customers. Um and I'm doing nothing so far to like make the onboarding process easy. I of course want that ultimately, but I'm not bothering that with that now because I can just manually onboard everybody and I'm gonna have a personal relationship with every single customer who comes in at this point. So investing in making that easy is like not where I want to invest first. Um and also the investments in my sales and marketing activities, like right now, all that is highly manual and labor intensive. I'm a I'm a big believer in the Paul Gram, do things that don't scale idea. And so like if if in order to keep you as a customer, I have to, you know, do 10 hours of calls with you and build special features and all that, like sure, I'll do whatever it takes to get each customer because there's more value than just that customer, it's yeah, adding value to the product and like all this stuff I'm learning, blah blah blah. Um and then I I'll try to get more efficient with my sales and marketing and cast a wider net and stuff like that. What do you think of that approach?

SPEAKER_00 38:45

Yeah, I mean I I think I agree with that as well. So I I would agree on going to whatever length, like I'm filling out stupid government forms that I always told my like I would never fill these out, you know. But it's like if somebody wants to use Flipper, and the only way they can is if I fill out these forms, then I'm like, okay, I'm I'll hop on a call, I'll listen to somebody talk about some you know dumb government thing that's unrelated to the customer, it's a middle mid middle person that I have to go through in order for that customer to pay me, I will do that, you know. So like I definitely will do those kinds of things. But I also believe very highly in like the most frictionless onboarding that's possible because it not only helps you know people who do just want to jump on by themselves without being handheld, it also helps everybody who does want to be handheld because you know now it's more of what you're doing. I think every time you onboard someone like that manually, you should spend a little bit of time to try and automate the bits that were not automated so that each one gets easier so that you can do that with more people, you know. So I've actually been doing that. Um so like flippers just about I've I like the the way you migrate currently from like flipper open source to you know the flipper, like one of the commercial versions, is basically like you click a button to export, and then you know, you click a button to import the file. So it creates an export and then you import it. So that's one way. The other way you can do it is like from your console. You can just, you know, flipper cloud.new, you know, your your token dot import and then like your current flipper adapter. Either of those work. But I was like, you know what, it just needs to be more seamless. Like a lot of people use the flipper UI. And so uh like two days ago, I was like, I'm gonna make everything like just completely seamless, like one click, automatic. Like it's it's like almost so slippery that you just went right down the waterfall and didn't realize it, and you're like pulling out your credit card and you're like, wait a second, what happened? You know, like that's that's what I've been obsessed with the last little bit, is because I've seen on Box Out, which is you know uh my my largest comp uh company that I'm involved in, is like the the better you make the onboarding, like the better everything is. Retention's better, everything kind of feeds, they're the same thing from two different sides. Like, if it's really easy to get into your product and use it and see value, then you're probably also not gonna leave because you see the value. So I feel like the two they're kind of two sides of the same coin, and so I've been really the last like six months super focused on onboarding. So with flipper, I basically did um I have a like CLI that's about to go out. So you could if you're in your app and you have the flipper gem, it's basically add the flipper gem, and then it's like you know, bundle exact flipper setup, uh, or you know, migrate. And so like if you're you're you can just be on a console in production, or you can be you know bash in production or whatever, or you can be in the flipper UI open source, and there's just a button. You just click the button, it creates a a stash of your features, gives you, redirects you to the flipper signup, and that stash you know has a code that goes into your session, and then when you sign up, it takes you to a page where it's like, would you like us to make you know this project from your Rails app name automatically and import all your features? Or do you want to make a new one or import into one of your you know, it does everything is just like this complete automatic flow of like just click a button and boom, your features are just magically all imported, or you can do it from the CLI or you can do it from the console. Whatever you want, choose your own adventure, kind of a thing. And then the same thing, I've I'm building um an just a like a really lightweight, cheap APM. Um, got some really cool ideas based on like there's just so much new technology that's out, and there's some I think some opinionated ways you can do things that will dramatically reduce the amount of data, and that's what makes those expensive. But no one who's doing it has an incentive to do that because they're making money on it. So why would they stop, you know? So my thing is like I'm gonna make you know this super, super uh affordable version. Cheap is probably like a horrible horrible word, scrappy and affordable, let's call it that. Um, and it it's like so I've been working on that as well, and it's just about ready. And it the whole thing is the same. It's like you know, it's gonna be probably called caboose, and like you just caboose setup and you're done. And that will automatically you know redirect you to a browser, and then you authorize, you're back, and you've got a token in there. It automatically from that one token, you can put it in all of your apps, or each app can have its own, whatever, and it just sets everything up and you're done. Like, you don't it's literally add a gem, add a token, and it guides you through the whole flow, and you have APM and you don't have to think about it. And so that's great. Yeah, go ahead. You're gonna yeah, you look like you got thoughts.

SPEAKER_01 43:14

Oh, yeah, yeah. So I I I have all kinds of daydreams of like products I could build after Saturn CI takes over the world and stuff like that. Um,

Onboarding That Gets Faster Each Time

SPEAKER_01 43:25

one is hosting, because like I feel like the demise of Heroku is leaving a gap in the world, and nobody has really stepped in to fill it as well as it could be filled. Um, and so that's an opportunity. Have you used Railway?

SPEAKER_00 43:42

Just side side note or not? No, no. Okay, I'm obsessed with Railway right now, highly impressed by it. Um, there's a little bit of like mental learning curve just because they do things differently than Heroku, but like it's uh as good better than Heroku once you get in it. Like it's it's legit, really good. Um, and they ship they're shipping. Cadence is just unreal. I mean, they're just putting stuff out like crazy. So highly recommend them. They just got like a hundred million dollar funding series, whatever, A or B. So they're like, they're legit. They're going to be around. Yeah, they're they're super impressive. I'm all on them. Well, yeah, our I mean, Boxout's core rendering infrastructure is all on Railway. We haven't moved the Rails app yet because the database and the database is state and state's hard to move. But anything that's stateless, I've already moved over there on all my apps.

SPEAKER_01 44:27

So interesting. Okay, I'll check that out.

SPEAKER_00 44:30

Sorry, keep going. Hosting, what's next?

SPEAKER_01 44:32

Yeah, yeah. Hosting. Um and you know, okay, so I run CI that requires like compute resources and stuff like that. Right now I'm like running a Hetzner uh dedicated dedicated box. And I'm like, well, eventually I'll want my own physical data centers. And then once I have that at scale, then I can start reselling uh some of my capacity and be like an AWS killer. Um so that's another idea. You know, crazy scope and all that stuff, but I'm I'm just Adrian. Yeah. And then uh an APM was was the other thing. Because like I use New Relic and it's like fine, it's useful.

SPEAKER_00 45:20

As long as you're by yourself. Yeah. If you add a second person, it's ten times more expensive. You know, if you add a third person, yeah. It's like for ten, it's ten bucks a month, you know, to get going, but as soon as you add a second person, it's a hundred bucks a month. And then if you want to add more people, it's a hundred bucks for like every person. It's it's insane. Yeah. So like if you have a team, it gets really expensive really quick.

SPEAKER_01 45:39

Yeah, and the the price isn't even the whole thing. It's like it's built for the enterprise and it really shows when you get into the UI. And it's like I don't know what to do in here. And I think probably um, you know, for people like you and me, um, who I mean I I don't have a deep understanding of your background, so maybe I'm making bad assumptions, but like I don't have like a deep history in application performance. Like I can performance optimize something, but like this is this is not what I do all day, every day. And I'm just trying to build a new product as quickly as I possibly can. Yeah. And I don't want to like futz around with my APM a whole bunch. What I what I wish I could have is just a product that I spin up, and it's like, okay, here are like the top three things you should pay attention to, and here's those. And it kind of like shows you what to do. Like the the interface suggests certain paths through it.

SPEAKER_00 46:37

Speak of my language.

SPEAKER_01 46:38

Yeah.

SPEAKER_00 46:39

Breach on.

SPEAKER_01 46:41

So it sounds like we're thinking about that in similar ways.

SPEAKER_00 46:44

Yeah,

Building A Cheaper Opinionated APM

SPEAKER_00 46:45

I so I have a lot of experience because that's what I did at GitHub for, you know, I mean, most of the years that I was there. Like the analytics team was like the first couple years, which is you know, store sending, receiving, processing a bunch of data. And then it was like, you know, I did a lot of the first dashboards and stats D stuff that we had, like just getting it organized, you know, and stuff like that. And then eventually, after like four years of that, we moved to Datadog, and I did the migration for that. So like I have a lot of strong opinions on like what do you need when like how what how can you build an opinionated APM? And if it's opinionated, then it's cheap to host, you know, because you're not storing everything so that you can ask any question you want. You're building pre-computed reports, which is like, I mean, that's there's no better way to build something performant than to say, I know the questions that you're gonna ask, so here's the answers to them, you know, automatically um on repeat. So yeah, I I could talk about this forever. Um, so I I'll let you choose your own adventure on that. But like um, I've got some very strong opinions, I've got some really good ideas, and I have some some friends who had built like um APMs in the past, and so I've talked to them and learned from it. Um I've been talking with you know Ben from Honey Badger because they do they kind of ingest everything, you know, especially with insights and all that. And so they have, you know, their problem is like, well, then how do we make it fast to query all this stuff when you're just receiving it all and you want to search through all of it? And you know, so like I'm kind of gonna go the opposite route where I'm just like, look, I know what it is. You want to know when something is slow, uh, when your app is slow or something is wrong, you want to know why is it slow? Did I get more requests or did the requests get slower? Like those are the two things you generally want to know. And if it got if you got more requests, which endpoints getting more requests, you know? And if it got slower, which endpoints got slower. Um, and if you know that at a at like a deeper level, then like you can you can do a lot of stuff with that. Um and that's very cheap and easy to store those things. So count sum and like error count. Um, so that's kind of like the first thing. And then you do the same thing for jobs, because those are like uh like what I would consider like an independent you know transaction type thing that's happening. And then inside of there, you just need to know the network stuff usually. So it's MySQL, Postgres, it's memcache, it's you know, HTTP requests, it's like all those kinds of things, view, cache, so view rendering, you know, uh cache stuff. So if you have like db, view, cache, and like some high level of those things, then you can usually you have almost everything you need. The last thing you need is like a couple of samples to look at. So you just need a couple of like slow requests. Um and if you have that, you have the whole picture, and especially in the AI world, you know, what you want is basically something that is API first that you can just point your you know, Claude or Codecs at and be like, here's here's how you get data, tell me what's slow, and then once it figures out what's slow, okay, now tell me, you know, get some of the samples, pull them straight from object storage, which is insanely cheap to store. They don't have to be indexed, they don't have to be any of that kind of stuff. All it is is just like user controller index, you know, and like you just need some of those samples. They can I I don't know anybody that I've talked to who's like, I want to see a span, an entire you know, span tree of a single request from like a month ago. Nobody says that. It's like let's not store that, let's just store it for like a day. Uh and so like, or like, you know, store the last five, whatever, whatever it is, um, something that's insanely cheap and easy like that. And that if you have, you know, these are uh over time the total count of requests and the total sum of compute that you're using and how many errors happened for every single endpoint and job, and you know, all the number of queries per table, like user selects and user inserts and all that stuff, and the same for every other network call, and you have all of that data and you have a few samples of each endpoint, like you can solve any performance problem, is my opinion. And so, or like any one, you can solve 80 to 90 percent of them. So, like for the people who really need everything, they can pay you know six figures and for their APM and they can do that. But for most people, I think you know that kind of a system is 10 to 50 bucks a month, maybe a hundred bucks for like some people who have like hundreds of servers or something like that. But for most people, it can be insanely cheap. And then you can focus on the thing that really matters, which is telling people when things get slow automatically instead of making them wait, you know, or suggesting SLAs or budgets or whatever, you know, however people want to think through that stuff. Um, so yeah, I started like a week ago. I'm almost ready to put my own apps in it because I've been going down the claw train pretty hard. And I feel pretty good about it. I just wrote it for Flipper, so it's it's running smooth already in production for you know all the flipper checks and stuff like that. So it's the same premise, you know, of like aggregating on the client per minute and shipping the telemetry up, you know, batching that, inserting it, partitioning it, pruning it, creating aggregates, like all the stuff that you would do to build an APM. Like I've already built it like on Fireside, on Flipper, on all the things. So um it's just building it one more time for this use case.

SPEAKER_01 51:50

So do you think eventually it might be the case that you use not just the APM tool on Flipper, but you use Flipper on the APM tool and you use the APM tool on the APM tool itself?

SPEAKER_00 52:03

Yep. Yeah, I'm uh definitely gonna do both. Yeah. I mean, I already have Flipper in the APM tool, and I was starting to think like, why am I building all this metric aggregation roll-up buffering and stuff like the eighth time? Like, why don't I just switch all the other products to just use, you know, the because this set this system has like four levels of like aggregation that you can kind of slice and dice in. And I'm like, if I just sent all the flipper telemetry here and then you know, queried it from an API, and I just sent all the download stuff for fireside here and queried it via an API. Like, it'd be way easier to just write this system once that can like aggregate and store all these things um and not have to keep rewriting it in every app that I do. It's really annoying. Um, especially then I would have enough throughput that I could use, I could justify using click house or something that is even better at doing those kinds of things. To date, I've you know, my little aggregations and buffers has been enough that I'm even with hundreds and hundreds of second or thousands of second, it's not a problem in you know my SQL. Um, or you could do it in Postgres, but I think it's better in MySQL for this kind of stuff. Um, but yeah, so I don't know. I I might it's definitely like this, you know, and it's not just cross from the standpoint of like they're both dev tools and stuff, it's also cross from the standpoint of like what do I think about? And what I think about is like how what's gonna be true in the future with AI, like you know, knowing what is happening in production and safely getting code out is gonna be more important over like the next year than it has been historically. And it was important historically, but like now with people looking at the code less or you know, any of that kind of stuff, I feel like it's just gonna be more important. So I feel like APM and resiliency, running you know, GitHub scientist type stuff, running two code paths at once and comparing mismatches and comparing performance, all that kind of stuff is I think the stuff that's that I'm super pumped about right now, and that's on my radar.

AI Driven Ops Without The UI

SPEAKER_01 53:56

Yeah, yeah, it's crazy. Like the possibilities are so enticing. Like uh right now I get an alert when my um server health is poor for whatever reason, and then I just take the alert and paste it into claud code and I'm like, hey, uh diagnose this and suggest a fix. And then if it suggests something that seems good, I'm like, alright, go ahead and do that. And wouldn't it be great if I just get like a text message or something? I'm out and about, and it's like, hey, your server is unhealthy in this way, it seems to be this, and I think we should do this about it. Yes, no. It's like, yeah, fix it. Uh yes. Yeah.

SPEAKER_00 54:37

Like you don't even need a UI. Like the UI just like honestly, it's there because why not build it because it's so easy to build. But like honestly, you just want an API with a CLI, and then you want skills, or you want, you know, um MCP or whatever on top of it so that that that you know, Claude or whoever knows how to you know speak your language and get the data out. Like I just did this with Flipper. I made a new endpoint for telemetry summary of every single flag. And so I did one curl to get all the you know gate data of like what's on, what's off, what's conditional. And then I did another curl and stored it in temp files of like here's all the telemetry data. And then I just went into Claude and I was like on Flipper Cloud itself, and I was like, hey, uh here's all the gate data and here's all the analytics over the past 30 days. Tell me which flags I can remove right now. And it was like, it did an amazing job. Like it was so good. Um, I got like, you know, here's six that like never have been checked. There's no telemetry data, so like you can delete these right away. Cool. Uh did that. Then it's like, here's another 10 or 15. And I was like, oh, four or five of these, like I still need, you know, like I can't um these are circuits. I want to leave them in, so like ignore those. Um, and you know, just like that alone. I cleared out 600 lines of code and like deleted like I don't know, 15 features that had just been kind of sitting around for too long. And I was like, this is amazing. Like I I that's what I that's when I built the APM because I did that and I was like, I want this for my APM as well. I want like I want the same thing. I want to be like in claw and just say, what you know, what's slow, or like get a text message or whatever, like you're talking about. I just want it to be more automatic, um, you know, with permission. Like I want to be like, yes, this is something I want you to act on and do your thing. Um, so yeah, totally. I I think it's gonna be really exciting the next yeah, three, six, twelve months.

SPEAKER_01 56:28

Yeah. Big time. It is a thrilling time to be alive. Um, all right.

What To Check Out Next

SPEAKER_01 56:33

Well, there's about 40 different topics that I want to have an hour-long conversation with you about.

SPEAKER_00 56:38

Sounds like round three.

SPEAKER_01 56:40

Yeah. Um, but sadly we're out of time for now. But um, of your mini projects and stuff like that, um, what's what do you think is most interesting to people listening to this podcast? Where should they go online and check out your stuff?

SPEAKER_00 56:53

Yeah, probably I mean, probably for people on this, Flipper Cloud is the first one, and then you know, caboose when it comes out. It's it's not ready yet. I'm gonna use it on my own. I'm gonna guinea pig first, and then after that, I mean everybody will know. I'll I'll be loud about it so uh uh when it's open and stuff like that. But for now, Flipper, I mean, I I've been shipping a ton of stuff to Flipper. It's really been cool to see um it pick up and grow. I mean, it's it's like doubled almost in the last year, and I think most of that was in the last three months, so I feel really strong on it right now. Um I think it's gonna be uh I think it's gonna have a big year, especially with AI. So I'm excited about it.

SPEAKER_01 57:30

So all right. Well, we will put that stuff in the show notes, and John, thanks so much for coming on the show. Yeah, thank you.