Podcast Expert Explains How to Have Explosive Growth Next Year

Everything that could ever have been tested, they have tested. Should the captions be on the top? Or should they be on the bottom? Should we do a split screen? Should we do a full screen? Are there things that are like universally true for all creators? Everything I build is hard to build because if it weren't, I'd be screwed. How are you thinking about the future of podcasting? Jimmy's famously said, "If you make a video 10% better than mine, you'll probably get four times the views." On today's episode of the podcast, one of the most respected people in the entire industry who very few people know unless you are as important as, say, Mr. Beast or Steven Bartlett, these are two people that rocks, codes, has actually worked with very, very closely. And in today's episode, we unpack his working relationship with them, as well as the tool that he's working on right now that helps make podcasts better. Okay, rocks. Welcome to the show. I'm so excited to have you. Please give me your incredibly illustrious career summarized in like 60 seconds. Okay, so first I used to have a restaurant startup and that didn't really work. And so I left the restaurant startup depressed. COVID happened, said, "I'm going to build startups for myself." Started twitch streaming, myself programming. Became a full-time twitch streamer for three years. Built tools for twitch streamers didn't work. They have no money. Started dating a YouTuber, built tools for YouTubers, worked. They have money. Thumbnailtest.com, A.B. Testing Thumbnails before YouTube did that. I built that for a while. The Mr. Beast team found me. They said, "Cool stream, want to work for us?" I said, "We'll try it." When out there, built basically version zero of what would become view stats from them, decided to go back off on my own. Finish Thumbnailtest, sold Thumbnailtest, got in touch with Stephen Bartlett from Diary of the CEO, who was a beta customer of Thumbnailtest. He said, "Let's do something." I said, "Bet, we started Flightcast, which is a video first podcast hosting platform together, and I've been running that for two years." Dude. I love speaking with you because you always make me laugh. Anytime I see you at a conference, you're surrounded by like 15 people, just making everybody laugh because your responses are just so funny. So firstly, the one that really intrigues me is you worked with Mr. Beast. How did that come about? Yeah. So when I was Twitch streaming, I was basically live coding, like eight hours a day. And I did this for about three years. The first year I was working on a different startup, and then I got to press sold everything, move to Thailand. When I did that, I started building my own tools. So I built the tool for Twitch streamers, and then I started building Thumbnailtest when I met April in. And at the time, I'm still streaming, and I'm still tweeting about all of this. So when I launched Thumbnailtest, it was a very public launch on Twitter, which is where the majority of my audience has been for years, I guess. Now it's maybe more LinkedIn, weird. But at that time, I launched June 30th, and I think within about six weeks, I wasn't touched with the Mr. Beast team, and they were asking about me, and asking me to come work for them in some context. So I like to think of it as I was building in public, and the Beast team stays pretty tapped in. So you ended up going to North Carolina and working with the Beast team, is that right? Yeah, yeah, I went in person. Too cool of an opportunity to not go in person, you gotta do it. Right, right, right. What was your impression of the way that they operate? They are obviously the top of their game, they were back then as well. What are they doing that's so impressive and interesting to you? As a guy who prides himself on what he builds, and the quality of the thing that I make, and trying to push the boundary, there was kind of no bigger opportunity than getting to be with the people who have pushed every boundary. Right, the people who define the industry to some extent. Because then we could do. I'm obsessed with the concept of building world firsts, and I think that you'll never really do something for the first time, 99.99% of the time in this world, especially as I say, every iteration of everything has been created. But it was like, we sat down and we were like, "Great, let's do the outliers." I learned that there. That concept wasn't mainstream until maybe 23, before 23, I think like 100 consultants maybe knew about that. At least to me, it was a very new concept. And it was like, okay, great, outliers are your thing. Let's build tech against outliers. That was crazy. That whole concept was introduced, and then it was great. Let's find the best videos in the world to make the best videos in the world, and let's automate all of this in a time where automation was much harder. Because again, you were writing it all by hand. So they picked me because I was very good at building stuff quickly. That was kind of my whole MO. It was small builds fast. Thumbnail test was a 30-day build originally. I started it on June 1st. I shipped it on June 30th if I didn't. I had to pay my friend a thousand dollars that I did not have. Oh. So I really had to ship it. I really didn't want to work there actually, because I was so scared of becoming an employee and not building startups. And I think if it hadn't been for my aspirations to make bajillions of dollars, it would have been like a trivial decision. But it's the same reason I ended up leaving in the end was the desire to do startups instead. But I loved my time there. Those guys are geniuses. They really care. They care more than I could tell anybody else I've cared. Is that like the primary characteristic that leads to excellence, just like caring more than other people in the industry? Like what else did you learn while you're there? What do they care about? Well, everything. There is no stone that must go unturned. There is no. We shouldn't do this. We have to do this when it comes to the content at least. Sure, maybe we shouldn't do burgers. We should do chocolate. Let's bail on burgers. But if it has anything to do with thumbnails, let's test 1,000, right? Let's pre-vis 100. They had teams doing 3D models of sets ahead of time to figure out if they were feasible and if the video would look how they wanted to look. Right, like every single thing that they did was just taking this. Whatever playbooks we all do and whatever concepts we all do, whatever figmas and average creator might get to create in such a concept if any. And it was just the eleventh degree of of of it all. And that was amazingly the theme of Steve as well. Right, and Diary of the CEO is they are also obsessed with they call it 1% gains. Just how do I make the same 1% better, 1% better? Jimmy has famously said in like a recent podcast. If you make a video 10% better than mine, you'll probably get four times the views. Because that's how the math of YouTube works. The best video just keeps going and keeps going and keeps going. How do you apply that if you're not working for a team as large as Mr. B store has Diary of the CEO? How do you pursue excellence at all costs? There's so many things you could do. And I think to some extent it is mostly just becoming comfortable being a student and learning and then chasing learning. Like chasing every opportunity that you have to figure out something new. We have a feature in flycast that makes clips. But we post the clips to a separate channel. And everybody that comes on goes, "Okay, well what the hell is the point of that? That doesn't grow my channel is in the point of this to grow." And I say, "No, this is not about growth. This is about learning. I'm going to post 120 clips of your podcast a month. I'm going to spam them to a separate YouTube channel. And then I'm going to just tell you which of the clips got the most views. And the only goal of this entire thing is to get some signal of which 30 seconds in this episode was good. Because otherwise you would have to guess. With thumbnails, maybe testing exists. Okay, great. Take every experimentation opportunity, every lever given to you by YouTube. You have to do thumbnail testing. You have to do title testing. If you're already doing that, then start testing your intros. Start looking at that retention and seeing what you can do. Ask for feedback, send it to more friends. Like, yes, there is a laundry list of things. But there's also like diminishing returns. The thing that I happen to love is being at the top of the mountain of creators so that diminishing returns are considered equal. Where a 1% gain is all that's left to gain. So we must pursue it. We must do every little thing. When you're an average creator, there's so many 10% gains. There's so many 5% gains. There's learning different software or visualizing something better, giving people like Aprilin makes PDFs of her frameworks that you can download after the video and uses that as an email capture. And then that goes into her newsletter. And then the newsletter later goes into a course since she's created this whole business out of it that all started from just her making a framework and then deciding, ah, you know, I made this whole thing in Figma. I'll let them download it and I'll just block it behind email. Right? Every ounce of your whole whole stack could be tested, could be iterated, could be improved to some extent. And I don't know, maybe I played too many video games as a kid. But man, I love to see number go up. So that's where I go, right? Look at the easy things, look at the obvious things, look at the things that people are teaching. It also sounded like you were going to say, you know, talking with other people in the community or like finding people to connect with and learning from them through their experiences and other way to just find those incremental gains. Yeah, and feedback. I mean, feedback is probably thing number one. Learn thumbnails, learn titles, get feedback, truly. Like I could not say more how much, and you know, I always bring up April and it's the case study. The number one thing she ever did was first become a student. Learn thumbnails, learn titles, learn intros. And then got a mastermind set up of her and a few YouTuber friends. And they called every week and they all capped all over each other's videos. And they all capped all over each other's thumbnails. And they capped all over each other's titles. And they gave notes and they gave notes. And they said, I don't like this. I didn't understand this. And then they did it again, right? Even the concept of being willing to redo something in a video, right? Oh, I already filmed it. Oh, I already did the voiceover. I already animated it. I can't redo it. Great. Cool. You now have cost yourself an opportunity to learn. One thing that I wanted to add in was you mentioned using, you know, putting out 120 clips, using Flightcast to use it as a learning opportunity to see what's actually objectively good from your episode. You know, I work in the educational space. And we try and cover all ground, right? We cover YouTube, Instagram, all the different platforms. And so we've been thoughtfully spamming Twitter. And then every 30 days we get a report and we see like what of our content is performing really, really well versus what's not performing really well. And then from the ones that win, those get turned into like Instagram carousels or maybe they get promoted into shorts or like a YouTube community post. So that's really interesting that you do that with Flightcast and that you use like a second channel as a way of learning. I think that's so clever. I mean, testing your cross channels is clever. You know, I, Steph told me a little bit about that kind of side of things that you do. And I like, I never really even contextualized that for my. Diaries CEO, they'll test 200 thumbnails on Facebook, paid ads on Facebook, 10 bucks, whatever, on every thumbnail, to see whatever click through it. And then that gives them their top five or 10. And then top three, go to the AB test on YouTube. But even the idea of, I'm gonna tweet something every day, look at the performance of the tweets, I'm gonna take the top five tweets, those get to become carousels, they'll take the top five carousels, and then two of those become shorts, and then whichever short does better, I'll make a whole episode about, right? There's pre-validation is a whole kind of like advanced creator playbook. And the shape that it takes is always so interesting, because it really depends on the creator, it depends on where you consider your testing ground to be. Like some people do it organic, some people do it paid, some people do it, trial reels, right? There's all these different ways, but all we're pursuing constantly is the idea of some signal that a thing is good before I put in way more effort to it. Which is really crazy, because if you super-duper think about it, that's actually what outlier theory also is. Outlier theory is that with other people's content. It is pre-validation through market proof of not your content. Do you wanna give like a background of what flight cast is? - Flight cast is a video first podcast on the same platform. Basically one shot upload to YouTube videos, Spotify video, Apple video, RSS video. It pulls in all the analytics from all the places, which is a really annoying problem for a lot of podcasts. So it's all your YouTube data, your Spotify data, everything in there. And then it creates very complex graphs, like overlaps between episodes of audience. So you can see if people are returning in the way that you expect and go deeper than just new in returning. It's got an AI chatbot to talk to your data. If you're like, ah, there's not this graph that I want. I have a question or you're on the call with like some sponsor. And they're like, what's your 112 day episode average views? And you're like, what? You can ask the robot and then you can answer it. And then we also do all the stuff with dynamic ads. We also help you monetize on Spotify. And it's on other cool stuff like the clips that we were talking about. And I co-founded it with Stephen Bartlett from the diary of a CEO. - So like our show is hosted on Flycast and has been from day one. And you know, I'm YouTube native. And I consider myself pretty podcast illiterate. It's so much more intricate to like actually host something somewhere and then to get it all across these different platforms and whatever. But then you just also subtly drop the fact that your co-founder is Stephen Bartlett, which is crazy. - We didn't, I didn't have this idea before beating him. We met and he was like, we should do something together. And I was like, what? He was like, I don't know. And so he, no, it's not true. First he had you, he pitched me on. Well, he was a beta customer of thumbnail test. And so he had my WhatsApp from asking for some features early on. 'Cause again, he was, you know, very day zero about a lot of this kind of stuff. And he pitched me on this other idea. I said, I was selling thumbnail test other time. I was like, sure, let's try 30 days working together. I tried this other idea. I was like, I don't love this idea, but I like working together. Let's try something else. He sat me down one day. He's like, "Frox, what's the best idea?" Like what's just the best thing that we can build together? And I, out of my pocket, I pulled podcast hosting. And he's like, why? And I was like, I don't think a really cool podcast host has been made yet. I don't think a video podcast host has been made yet. Now everybody else does it. But we were like, really the first like video first, where that was the whole topic. And most importantly, podcast hosting is really interesting because it is a necessity. You have to use something to host your podcast. You have to use some tool here. So if I can build the best thing, it is a need. And I have never really built software that was a need. And the opportunity to build a need and then stuff it full of wants was a very exciting premise. - I imagine you have a lot of data. Are there things that are like universally true for all creators that you've learned? - Well, I can't see everybody's data legally. I do wanna say that. But for those who share it with me, which is a lot, adding Apple video probably grows your show like eight to 10% on average right now. We're the only guys doing it in 4K, I think, or one of the very few doing it in 4K 'cause Apple's gonna put it on TVs soon. Apple video is really complicated. Anyway, Apple video grows you like eight to 10%. Spotify's pushing video, I'd recommend it. If you're used to the one of 10 thing on YouTube, which shows like how good this episode is relative to your recent ones, that like episode performance, we do that for Spotify, YouTube, and Apple separately. Not a huge correlation. Generally, Spotify and Apple don't pop off. They don't like have viral episodes as much as YouTube would because they are not algorithmic platforms, which is a very, very, very fundamental understanding of what's different between the YouTube industry and the podcast industry is algorithms. That's kinda why it's like a little weird to do video on Spotify and you know, on Apple because why? You do a video on YouTube because videos, the only thing, it is a video platform and there is an algorithm. Doing it on Spotify or Apple, the reason right now to mainly do it is either better connection with your audience. Thumbnails having a bigger physical space on the screen, which is a big deal and just the platforms happen to be pushing them. So like let's take that eight percent, let's take that 10 percent, why wouldn't we? But episodes don't pop off in the same way because of that lack of algorithm. So that is very telling of still YouTube being the primary platform. Podcasters generally monetize on RSS though. So usually at the highest level, monetization comes in a lot of different formats. I've learned a lot about what that looks like. Hotcasters make way more money than YouTubers do with the same view count usually because of dynamic ads and being able to sell at a CPM your own ads beyond just a sponsorship. So we have these words of baked in versus dynamic that doesn't really exist on YouTube right now. What is a good growth strategy for a podcast on YouTube? I'm finding it more challenging to get high numbers and I think it's partly because podcasts are very, very long. So they see the timestamp and maybe they're not interested. If maybe also harder to deliver on time to value in a podcast episode compared to like a talking head YouTube video. Have you seen anything that's like working really, really well? I think podcasting is really, really hard to grow on YouTube from zero. I think especially because we talk about expertise. The expertise of the guest becomes so huge. So like, you know, expert strategists explain how to go viral on YouTube. Okay, well, I don't know who that guy is necessarily, even though he's, you know, and so on, he's the legend. The average pod person doesn't know who he is and then how many other videos have a title that sounds like this and how many other videos have a thumbnail that looks like this is kind of a hard question to always ask. And if the answer is a lot, then you're in a tough spot. And I think what happens with podcasts a lot is they start to look the same and that becomes a tough game. So the value of it is harder to convince people of because they are inherently doubtful of anything about like creator economy being valuable because people love to talk and then try and sell them something. That's like kind of an example in this case at least. So usually guests have been much more important than I ever thought they would be. If you look at a lot of YouTube strategists refuse to work on podcasts because all of their aspirations and all of their knowledge applies maybe 10 to 20%. And the other 80% is the guest and the content being good, which is hard. And then good takes a very different definition because weirdly, you have to compare like this podcast to an April and alter video. For example, if any of them is talking about like YouTube growth or something, right? How does Mega Channel keeps going viral on YouTube, right? Like that kind of video will be on the same algorithm as April ends, but April ends is doing a framework, a detailed education, a highly edited, clearly hundreds of hours of work video, which is on the same platform as a podcast that probably filmed for an hour, had a few hours of editing, had a few hours of thumbnail and title. So total time in is like 10 hours, right? And that's not a perfect comparison, but I've always liked that kind of mindset because it really forces you to think about, how do I make my 10 hours of work feel as valuable as somebody else's 100? Because if I can't, I won't win in the algorithm because the audience won't show up. Because that same 30 minutes is arguably better spent on the higher effort deeper dive thing. Whatever that thing is. So a lot of like Diary of the CEO comes to bringing the smartest guy who oftentimes has not yet been heard of. So they're like very often the first or one of the first people to interview somebody. Generally they do it at a higher quality than anybody did it. They generally do more research. So it's the best interview of that guy and then the thumbnails and titles are obviously crazy. Every single time. So it's so, so tough. So we can talk about optimizing trailers and I would say that that is a pretty big deal, right? - Yeah, I've always wanted to interview the guy who cuts the trailers for DOAC because we did it one time when I worked with Mark Manson's team and on YouTube it led to like a pretty steep drop off. So that either means the trailer was bad and we thought it was good. Or I've been really curious, like is trailers the right way to begin an episode? Or do you just get right into it and ask a spicy question and then go from there? - Depends. It's all to say, but depends. And the answer is it not only depends on the podcast, it depends on the platform. My friend Seth Silver's found this out. He puts less sound effects on trailers on audio platforms because people complained and said, what's with all these weird beeps and boops and page turn sounds and all of this? This means nothing to me in this two-minute intro. It's like being bang, bap zoom and it's like, this is nothing. I'm not seeing it. And my eyes are closed. This sounds like a Batman comic and I just assume they're fighting. You know, it's not providing anything. You know, audience expectation, right? You have a microphone in your thumbnail. You are signaling this is a podcast. That's great because it sets up that expectation. Maybe the trailer sets the wrong expectation. Maybe the trailer is a little too hype. But also, if you see huge drop off, the same question I have to ask is, well, did that episode get way more views? Because if that episode got way more views and had a lot of drop off, that's why. It's because actually this episode went viral and virality looks like worse retention. Once you go viral, all the stats reverse. Everything goes backwards. Your CTR goes down and that's a good thing because you're growing, you know, the retention goes down a good thing because more people are seeing it and that is just the aggregate audience having a little bit less of an appetite but still interest even with like some tests that we've run on like different titleing strategies on audio and everything hard to find any good patterns in there because people click on Spotify more like they click on Patreon. They're really just their week over week generally speaking. It's odd for an episode to pop off on these platforms like I said which also means weirdly every episode is like a five out of 10 like every episode is doing about the same as the other episodes. If you're like I get a thousand downloads per episode you say that with some confidence. On YouTube we'd be like yeah I get between a hundred thousand and like two million you know is is how that feels which is in Spotify would be insane it'd be insane to get a hundred K followed by two million it happens very very very rarely. So that's one tool that you do inside flight cast what are some other things that make flight cast the best of the best. There's a lot of little things for example does Markman's team have an editor a video editor like a podcast that yeah does the video editor upload the file no it goes to a social media person great how does it get sent Oh Google drive great so flight cast has import from Google drive and the reason we did that is because that file for the whole edit of that five hour podcast guarantee you somewhere between 25 and 60 gigs massive download massive upload don't know who your social media person is doubt they have five gig internet right at best they're losing 20 minutes that worst it's overnight our servers are 10 gigs per minute so you can pull a 30 gigabyte file in in three minutes if it's on Google drive which for a lot of podcasts you work with it also pulls from frame.io or Dropbox or Lucidlink which like Markman enterprise one those kinds of little workflow things that end up saving a huge chunk of time are a lot of the stuff that we've done as well yes you can publish everywhere and if you want YouTube to go out two hours after Spotify goes out you can just do that you can just set different times if you want them to have different titles you can just do that if you want a different file in the YouTube when you want to say subscribe in the Spotify when you want to say follow great put in two different files you can do it the right one goes the right platform so you are a builder in the creator economy space and I feel like this is becoming an increasingly saturated market what have you seen in the past couple of years how do you feel about this whole everyone's building these little micro tools kind of sick kind of sucks for creators I love to see them make their own stuff and it forces people like me to be so good at what we do and so dead focused that I have to both think of things nobody has thought of implement things nobody else could build and then sell them to a set of people who have to largely act the same and YouTubers do not largely act the same podcasters do which is very helpful a lot of creator people know this a lot of people building for creators don't know anything and they're just spamming AI tools and I swear to God if I see an eight hundred opus clip take I'm going to lose my mind if I see oh it writes your scripts for you oh it pulls in this and then prints out that pulls in this prints out that I'm just going to lose it Steve wanted to do flycas as the AI powered podcast platform and we started and I said I will bet you a million dollars that that video first beats AI first and we tested it we literally tested it and we ran a ton of ads and I was right and I was like thank God I'm right because I really didn't want to do AI first so really I so so deeply at this moment don't believe in building companies where AI is a foundational part at least for me because for me that means that at some point it can all kind of be redone at some point you know somebody companies died to chat GPT overnight all the time that can't happen to me I have deep complicated video infrastructure I have an insane amount of analytics polls that are all gross and then reorganize and reformat it and very efficiently done so it doesn't cost too much like everything I build is hard to build because if it weren't I'd be screwed and I think that is missed because it is so easy to build now we keep getting pieces of software that are a six out of 10 and I just never want to use a tool that's a six out of 10 and I think maybe I've made an eight and a half and I'm hoping in a year I've made a 10 so for people like me living in the year 2026 in the podcasting space how are people living in 2028 or 2029 how are you thinking about the future of podcasting oh dude it's so crazy multi-multiplot form beyond what we've spoken about today broken TVs fire TV Netflix to be all of these places are starting to ingest some of them openly some of them less openly I'm going to kind of talk about premium because I guess this is the the 2028 of it all right it all is still experimentation but depth how how far can we take this baton of of experimenting how much can we test where testing was not possible before right how much can we invent that was not invented before to do so right steves team made up this Facebook ads thing and it worked for them it worked really well for them what else can we make up what what what other playbooks can we find and or create and then copy paste playbook to other podcast scale it up so Steve made I have a CEO now flight story has a few shows most of them doing very well that opportunity then expands how do we eventually do ads better than anybody else that we eventually do sponsorships better than anybody else can I take inspiration from data from one place and use it to inform another like die of a ceo talks about experimentation so much but the number of posts everything that could ever have been tested they have tested and continue to test the captions be on the top should be on the bottom should we do a split screen should do a full screen should start with a guy should start with an I boy should start with a title should start horizontal but it just like horizontal video in a vertical video or should it be full vertical video or should it be top and bottom what color should the captions be should they be single word should be full sense every single thing that could ever be tested gets tested gets tested again gets tested per platform is then going to be kind of the oh my god your brain explodes which is the logic of like what works on Instagram doesn't always work on tiktok and the audiences are different and then the recognition of that implies that the content must differ and so for mr. beast they don't even usually repost stuff across platforms right a mr. beast they had years and years ago hit the point of instagram gets their instagram videos tiktok gets a tiktok video different completely different and that was so so so interesting so I think podcasting goes the same way but I think really the expansion of scope as podcasting becomes for lack of a word cheap TV to produce and then people who are TV or want TV such as the two pieces of the world and the netflix is of the world and all these guys begin to attempt to ingest these signing these deals with whomever you know who knows where I was he was going to show up next it's on planes right it's on a lot of planes and there's a handful there's like probably a few dozen podcasts on planes but that's crazy who even gets you on there what screen is that but that opportunity is the plane people needed content you know else needs content to be you know who else needs content Netflix you know he needs content Disney right you know who's making podcast everybody that needs TV who wants to spend less on production cuz it's like two people and maybe three cameras and it's pretty good the diary of a CEO is largely show run by like one guy maybe two people behind camera yes they have a couple of editors just they have a couple some people all of those things but if you're Disney you can have multiple people doing it I think that's where it goes in 2028 and this concept becomes how do I build the most perfect show that then can exist so perfectly across all of these platforms non-problematically such that no matter where my content is the audience gravitates towards it and then the opportunities for me as a creator expand so much because I can cut a Netflix deal and a deal with to be and a deal with and a deal with and a deal with and that presents a whole new monetization wrong that we as youtubers and never even really really considered similar to how people on YouTube started talking about shows like Michelle courageous challenge accepted which is a show which can be sold and that's that's what I think happens in 2028 thank you for coming to my Ted talk. Rocksman it is such a pleasure to speak with you the way that you think is just so enlightening I think so highly of you and also just to boot your the funniest guy I know so thank you for coming on the show. Where can people find you flightcast.com twitter.com/ out rocks codes rocks with the next codes like programming linked in dot com slash in slash rocks codes rocks dot codes is my personal website and I bet you want to do a shout out to your favorite YouTube channel the gimbal podcast. Every single video you ever need to watch about learning thumbnails learning titles learning intros at the advanced level and the beginner level it goes all the way through she is a dynamo she is an unbelievable talent and I like her not just because her videos are cool but also because she thinks I'm cool and I'm very lucky in that but you should watch her genuinely she's so so so goaded she's the best at what she does in my opinion. Appreciate you Vinnie. Thanks a lot. Thank you so much for watching or listening to this episode if you're new here please consider subscribing or leaving us a review on the audio platforms it helps us as a very new show. If you want to check out another episode just look through our library it's starting to build up we're at like 40 episodes now and we have a new one coming out every single Thursday so thank you for being here and I will see you then. [MUSIC]