0:00Jordan Metzner previews a podcast CRM he built that generates outreach emails to other podcasts from templates, and calls Suno an AI music creation app that is basically text to song.
2:02Jordan Metzner says podcasts are syndicated through RSS, which supports an email address, so almost all feeds list one. He noticed after a podcast ranking service with a $10 a month subscription emailed him.
3:36Jordan Metzner says he scraped the top 1,061 business podcasts, each with an email address, and built a template system where AI writes a custom subject line and email per show that opens in Gmail to send.
8:51Jordan Metzner says Apple's podcast lists are publicly available through a robust URL schema, unlike Spotify's, and Replit grabbed and sourced the data. He wanted to send 50 to 100 emails and says anyone could build it in a few hours.
11:40Jordan Metzner says Suno is an AI music creation app that is basically text to song: you type a prompt such as an opera song about Built This Week and quickly get a high quality song.
13:34Jordan Metzner says Suno does not allow artist references, so he asks ChatGPT to write a Suno 4.5 prompt and lyrics with no artist names, then pastes both into Suno and gets a much better song.
16:07Jordan Metzner says Suno returns two versions of each song and lets you download MP3 or WAV files, generate a lyric video, remix and edit with styles, and get stems that separate the audio from the instrumental.
18:05Jordan Metzner says they used Suno to make the Built This Week jingle and loops for clips, producers like Timbaland are starting to use it, and it can produce a style-alike score that might otherwise need an orchestra.
20:39Jordan Metzner says he does not think AI music will eliminate human music but it will most likely improve it, and the big untapped opportunity is AI inside music production tools like Logic or Ableton.
23:57Discussing the American Federation of Teachers' AI training hub, Sam Nadler says AI can accelerate learning, as with his five year old learning space facts from ChatGPT voice mode, but fears it replacing critical thinking.
AI Podcast Outreach CRM and Suno Songs: What We Built: the key questions from this episode, answered with quotes and timestamps.
Jordan Metzner [0:00]: And So I present to you, it's called podcast CRM or podcast promo, podcast name, and then I just choose my template. I have various templates. I click generate email. Suno is AI music creation app. It's basically text to song. Guitar sounds great. Audio quality is fantastic.
Sam Nadler [0:23]: The American Federation of Teachers who says it will open an AI training hub with Microsoft OpenAI Anthropic. Hey, everyone. Welcome to Built This Week, the podcast where we share what we're building, how we're building it, and what it means for AI and startups. I'm Sam Nadler, cofounder of Ryz Labs, and I'm joined here with my friend, business partner, and cohost Jordan Metzner. How are doing today, Jordan?
Jordan Metzner [1:04]: Hey, Sam. How's it going? Excited to be back here.
Sam Nadler [1:06]: I'm excited too. So, obviously, we're gonna cover, our normal docket, our normal agenda, but we have kind of a fun take this week. We're gonna cover something that we built very recently. I think the slight twist on it is I don't know. You told me you have a kind of a surprise, tool that you built for our team, so I'm really excited to see it. You wouldn't share what it was, so that's gonna be interesting. Then we're gonna jump into a tool that we both love to use. It's one of those tools that I think, when you show someone who's not deeply ingrained into AI, it immediately kind of creates that spark of imagination, inspiration, and it's it's one of those tools that almost anyone can have fun with. And then finally, we're gonna cover some news. So I pass it over to you to show me for the first time this AI tool that you built.
Jordan Metzner [2:02]: So, let me give you a just a quick background. We started this podcast just a few weeks ago. I've never had a podcast before. Don't know that much about podcasts. But I got an email from a website that says, like, your podcast is ranking. Like, click here to check out the charts. And I clicked on the website and has a free trial, and then, they have, like, a $10 a month subscription or something like that. And it got me thinking kinda how did this guy get my email address already and started to think about, like, what the whole process was. And I don't know if you know this, but podcasts are syndicated using RSS. And RSS is kind of an older technology that used to be used, for a lot of news syndication, and it's been most extinct, honestly, except for in the podcast world. So, anyway, RSS is like a URL system where you can broadcast out your feed of episodes, and that's what's used by most of the podcast systems in order to get the new episodes as they come out. Well, supported within RSS is an email address, which is, like I guess, like, maybe wasn't previously supported, but is supported, and I don't know as of when. But like I said, not an expert in kind of RSS podcasting. But long story short, in all these RSS feeds or almost all of them, there's a list of email addresses. And so, I saw this site.
Jordan Metzner [3:16]: I saw what it did. I kinda saw this thing, and I thought, like, okay. Well, like, what do I wanna do? I wanna promote our podcast, and I initially wanna promote our podcast with, like, other business podcasts. And I wanna tell them about our podcast in a way that shows that I listen and know about their podcast, and I wanna get their contact information. So it's like a perfect, like, little AI job. Perfect. And so I present to you, it's called podcast CRM or podcast promo. And, what it does, I've scraped, the top 1,061, business podcasts. It starts with number one, Ted podcast, and goes all the way down the list. Now if you notice here, every podcast has an email address. Now some look like this, like info plus blah. Probably, nobody's there to answer it. If you look at this one, number two, podcast at teams, okay. Maybe nobody's there to answer it. But, know, as you start to go down, you start to get, like, more customized addresses. Like, if you look at this one, this is with Ray and, here's Ray's email address. Right?
Sam Nadler [4:26]: So Yeah. The h b r, video@hbr.org is probably a valid email.
Jordan Metzner [4:31]: Yeah. Exactly. So, we have all the podcasts. I can sort by name and rank and, I can search like Ted example. And here, I can see all the stats, how many podcasts, how many have email, top 50 ranked. And then, what I built is an email templating system. So, I have this email template. I've already built it, so we can go into, like, edit mode here. But, this is the name of the template. And then I have thought of podcast name when we ship this as, like, a sample, but we can change that up. And since I don't really know the name of the podcast host, the only, like, kind of dynamic information I have is the podcast name. And it says like, hey, we're shipping stuff, blah. Here is like a template. Right? Perfect. And so that's like the template. And then I can go back to all the podcasts and, like, from my dashboard the same. And let me show you kinda how this works a little bit. So let's jump to, this first guy, Ray, that, like, has, like, a real name. So, let me just click, like, generate email.
Jordan Metzner [5:35]: And so it says, like, this is the podcast. This is the outreach email. This is, like, who I'm going to. Podcast name, and then I just choose my template. I could have various templates. And I click generate email. So look. Inspired by, Home Business Profits with Ray. And Home Business Profits with Ray is the name of the show. So now I've got, like custom, specific subject lines. Now this isn't like a highly engaging, like, click through rate subject line, but, I'm sure we can improve that. And then this has all been written by AI. So it says like, we're real builders shipping stuff built this week. Home business profits with Ray's been a big inspiration to us. We're not selling anything. We just wanna let you know about our podcast. Like, love get any feedback. And then Perfect. As you know, we use Gmail here. So you just click open and here you go. Here's your email. Here's the subject. Here's the whole thing. And then you can just press send.
Sam Nadler [6:29]: I love it. And then, like, ideally, they're, intrigued by your email. They take a listen, potentially reach out for collaboration or guest opportunities. Who knows? Just getting the word out there and, trying to establish a relationship.
Jordan Metzner [6:48]: Yeah. I tried not to like ask for anything as I thought that, asking
Sam Nadler [6:52]: Too much.
Jordan Metzner [6:52]: Stuff is what people try to do all the time. Yeah. So I felt the kinda humble approach is a little bit better. This guy has an agricultural podcast. Right? So we could just click here for the Damian email. Click like this, generate the email. This guy, we know his name is Damian. So in a lot of them, I don't really know if the name oh, so here we go. Like, it filled the name out for me. Right? And it says, like, this podcast has been an inspiration to us. We're not selling anything. Just wanna make sure if you like it, let us know. Amazing. And yeah. Just can click like that, and I'm in a Let's send one. Can we send one? Yeah. So this is to Damian Mason.
Sam Nadler [7:34]: And this is from your Jordan Metzner or the info at built this week account?
Jordan Metzner [7:40]: Well, I think I just happen to be in my personal Gmail address. Yeah. But if I move this over, we can send it. Got it. We'll send a schedule send just to, like, show the example. But, let's just, like, glance a look at this podcast just before we send it just to make sure we're not, making this all up. But here's, the business of agricultural podcast. Here's his email subscription, the latest episode. And, we could go back and, like, totally customize this email to say, like, reference the latest episode in it if we wanted to. Here we could probably learn a little bit about the podcast, etcetera. But, I think you see overall, like, that's kinda how it works. And, even on his website, like, here, you're not seeing his email readily available yet. We already have it. Oh, here it is actually. So
Sam Nadler [8:29]: Right at the top. But it's not for every podcast, and this kinda centralizes everything as well.
Jordan Metzner [8:34]: Correct. And this says, like, Damon Mason office at Gmail. Whereas, like, if you go to look at, like, my email, it's Damian@DamonMason.com. So actually, mine is probably much more accurate.
Sam Nadler [8:46]: Love it. Yeah. Great tool. Creative thinking. And, you know what? I didn't catch. So you caught this by being pitched a service to do what exactly?
Jordan Metzner [9:01]: The service was saying like, hey. You can track your ranking of your podcasts.
Sam Nadler [9:07]: Just pay $10 and you can track your ranking. And you're like, how did they get my email? And then you figured it out.
Jordan Metzner [9:12]: I was just like, yeah. How did they get my email? We're just a brand new podcast. Like, obviously, scraping podcast lists. But then I was like, well, that's probably, like, publicly readily available information is the podcast lists. And, the Spotify lists are not, like, readily available, but the Apple ones are. And Apple has, like, a pretty robust URL schema, so you can, like, I guess, lack of a better term, like, download the list of, podcasts by your subject matter category and, get the URL and a bunch of info about them. Yeah, obviously, we could build, a v two of this where we have a little bit about each podcast and maybe a link to the podcast. And, we could probably integrate this into a CRM and get some feedback loop and other things like it. But, from a high level, I just wanted to crank out, like, 50 to, a 100 emails to podcast operators, to podcast creators, and just tell them about our podcast.
Jordan Metzner [10:06]: And See where it takes us. Yeah. Hopefully, just, one or two people would say, like, oh, that's really cool. I like it. I'll tell my viewers about it. And, just a few of those could probably go a really long way. This one's called dude dudes doing business. You know? So
Sam Nadler [10:21]: Sounds like us.
Jordan Metzner [10:22]: I mean, sounds like us, except the 27 ranked, podcast and business podcasts. So, I used Replit to build this. I didn't actually have to go script the data myself. Like, I was able to find it, and then, Replit was able to, like, grab that data and source it for me. As you see here, know, I have a button to check for new podcasts. I can check for updates of the rankings. And then I have my templates here. And then I have some fake pages kind of as that I just built out in case, like, I need to add some additional features. And then I have this one little feature here, which is to see, when built this week is in the top 1,000. Unfortunately, we are
Sam Nadler [11:04]: We're not
Jordan Metzner [11:04]: Married yet. You're not in the top 1,000. Keep growing, it says. So anytime I want to, I can just refresh the rankings or check for new podcasts, refresh the rankings, then see if we made it yet this week. So anyway, yeah, that's a little bit on the podcast CRM tool. I think, we'll definitely use it ourselves, and I think anyone can build this in, yeah, probably just a few hours.
Sam Nadler [11:28]: We'll have to celebrate when we cross over the top 1,000 milestone.
Jordan Metzner [11:33]: Okay. Cool. Sounds good. We can come back to this clip.
Sam Nadler [11:36]: Come back to this clip. Yeah. Yeah. Okay. Yeah. Just to transition to the tool of the week, like mentioned, it's one of our favorite tools to use. We used it when kind of getting the different elements of our first podcast ready. We both use it from a professional point of view and a personal point of view. I mean, I've done it for multiple kind of personal engagements. It's just a really fun way to share AI with friends and family, especially those who like don't use AI on a daily basis or are scared of AI. So the tool is Suno. I would argue, Jordan, you're a power user. Why don't you kinda walk us through, a, what the product does, how we've used it, some of the cool features about it.
Jordan Metzner [12:22]: Yeah. Cool. And I think that'll lead into kind of like, Suno in the news as well. So Suno is AI music creation app. It's basically text to song, and it's as simple as that. So, you can go up here and type something pretty simply, and, immediately, you'll get a song. Now there's a lot of things you can do to actually make the songs, much better. But, just to get started, I think, in the purposes of the show and the episode, we could say something like, make me an opera song about the AI podcast built this week, and it's just gonna go make it. And it knows about Opera. It probably knows a little bit about our podcast, and it'll do a decent job of building this. And let me just take this prompt with us. And it'll go and create a song. And I think, like, what's amazing about Suno is just, like, how fast it works, the high quality of the audio, and the ability for it to just make amazing music.
Sam Nadler [13:34]: Now I was just curious. In the prompt, can you reference, specific artists like Taylor Swift or Snoop Dogg or something like that?
Jordan Metzner [13:41]: That's a great question. So in Suno, you're actually not allowed to reference artists, and they do that, like, to protect the artists, I guess. But, where you can reference artists is in ChatGPT. And so what I have found is, using ChatGPT to make songs has been a really awesome way. So what I'll do here is go to ChatGPT and say, like, write me a prompt and lyrics for, Suno 4.5. And then I have the prompt I just made, and ChatGPT will write me both like a prompt as well as lyrics for the song. And I find that when I do that, I get a much better song. Now here you can see it's in the style of Verdi or Puccini, and actually that's not acceptable, for Suno, but I'll just let ChatGPT know that, and, it should be able to fix that.
Sam Nadler [14:43]: And you'll paste in the entire lyrics into the prompt or just the Suno prompt that ChatGPT gave you?
Jordan Metzner [14:51]: No. So it just depends on, like, what I'm making. But, yeah, I just tell it no artist name so it'll make this, kind of the prompt. And then I'll paste the lyrics in as well. Okay. So you copy over here your description from ChatGPT. And then here we take the lyrics. And like I told you, I'm not a big classical music fan, but I think it'll illustrate our point quite well. And then you can click here create. And in just a few moments, Suno will start creating a classical music song, an epic one in the silence of the morning. And in fact, they've added a bunch of new features as well. But I think what will also be cool is to take the same lyrics and just okay. Make this a trap country song. Just give me a new prompt. And so I think what'll be interesting is here we can listen. One
Sam Nadler [16:06]: One question I have is like, so every time you, it gives you these two outputs immediately. Usually, I do is, either choose one or kind of tweak the prompt and continue working. Can you go in and within Suno edit elements of the song, like add a baseline or, different kind of musical elements? I'm, not musically inclined enough to do that. But theoretically, can you go and tweak little bits and pieces of the song?
Jordan Metzner [16:39]: Yeah. So they've built a ton of new features. So first, when you produce the song, you're gonna get two versions of the song. As you heard, we listened to the first one. If we just listen for a second, we can listen to the second one and, we can In
Sam Nadler [16:52]: The silence of the morning.
Jordan Metzner [16:54]: Can hear it very different. Begins to sing a spot But we can jump, later into the song. Future's taken wing. So I think the first one might be better for us. Let's just say here as an example. And so then Suno's added a bunch of new features. You can download the song as an m p three or a WAV file. You can actually generate a video with lyrics, and then you can also now remix and edit the song. So they have an editor. You can use styles, so you can kinda like remix it. You can also get the stems, which is a new feature, which means you can separate out the audio from the instrumental. So it really has been quite a robust tool. And, it sounds maybe a little boring with classical music, as we've made here. But if I just take my new prompt using the same lyrics, I'm just gonna go back to Suno here, and we'll paste in my new prompt, which is country trap hip hop with the same exact lyrics. We'll create that and in just a few moments.
Jordan Metzner [17:54]: And like I said, it Suno is just so fast. Here, we're gonna get the same song, but like in country trap. And that just shows the diversity of it. Yeah.
Sam Nadler [18:04]: And then while making country trap music is, with, specific lyrics that we generate is fun, party trick, share with friends, whatever. How have we used this, in business?
Jordan Metzner [18:18]: Yeah. Well, I think, Justin I think before we talk about, like, business, what's crazy about music is it touches people, like, all across different ages and generations. And what's amazing about Suno is you can make all types of music. And so, I've showed this to a lot of non technical people, just made a music song, on the spot based on, maybe something they're wearing or some of their interest or something like that. And, based on a style they may like and immediately, hear the song, and they're just blown away. And, I've shown, my 80 year old aunt and made a song in Italian for her, and she couldn't believe, the quality of the music, the speed, You know, even in, her personal, like, home dialect language, I could make a song and, they don't even make music in that language anymore. So, yeah, it's just a incredibly diverse product, that, is really fun to play with. And I think it works, like, across all genres. Like, kids love music. Adults love music.
Jordan Metzner [19:19]: Everyone loves music. So in the case of your question, yeah, of course, we used it to build the Built This Week jingle for our show. And we continue to use it for loops on top of our music of our clips and for other videos we're using. This is just the beginning. I you know, some producers are starting to use it in their, like, music production, like Timbaland and some other ones. When certain TV commercials or movies are made, sometimes they'll hire an entire orchestra to reproduce a style of a song. So, if you wanna make a spy movie, then, you might make a song that sounds like James Bond but isn't James Bond. And, here, you can just simply go into Suno and, type in that kind of style and get something that'll probably work, and you could just keep going over and over again. I mean, I think we should just listen to a quick clip here of what this trap music is. We can, blend it in post, but, this is the same lyrics as we had a second ago.
Jordan Metzner [20:19]: Guitar sounds great. Audio quality is fantastic. Lyrics sound great. Real beats. I mean, sounds amazing. Sounds really, really awesome.
Sam Nadler [20:38]: It sounds so good. It actually leads us to kind of our next topic, which is I don't have the exact headline in front of me, but it's about a band that has you know, I wouldn't I don't know if they outed themselves or it became public that it is completely AI generated, Velvet Sundown. It's got a it's received a lot of popularity, listens. I've listened to it myself. I totally see me vibing to Velvet Sundown on a long road trip. And overall, it's really good. So what are your thoughts on kind of an AI complete AI band becoming so popular? And does this mean the end of, humans leading the creative inspiration of music? Or is it just, something people liked and there will be lots of, creative humans contributing in the future?
Jordan Metzner [21:41]: Yeah. I mean, think it's super interesting. I'm just kinda surprised it took this long for an artist to, like, break out. I know there was an AI artist. There was an artist who did, like, a Drake AI song, maybe about a year ago or so that kinda got pretty popular really quickly. And, yeah. Is the music from the Velvet Sundown, like, amazing? Maybe. Maybe not. You know? But the story is picking up speed, and so it kinda makes you wanna listen even more. So but, yeah. I think AI artists, AI actors, AI characters, just the same way, people love animated characters like Homer Simpson. Right? So, I think we'll definitely see, more of that in the future. Will, AI music eliminate human generated music? I don't think so. Will it improve, the quality of music generated by humans? Yeah. Most likely. I think, like, where we're gonna see some big opportunity, and we haven't really seen it yet, is really, like, using AI inside the music production studios. So, like, all of the DAWs, like, all of the digital workspaces, so whether that's, like, Logic or Ableton or any of the other ones, haven't incorporated AI to the point, like, Suno is. And, once you see, you're making a song and you say, oh, that sounds good. Put a drum track on it. Okay.
Jordan Metzner [23:02]: Have it add a heavy bass. Okay. You know? And you're you're making music while texting, I think, is an opportunity that, is upon us and is probably coming really soon. And that's gonna start to really improve the quality and speed in which music gets produced. And, it goes kinda back to, like, that video production stuff we talked about in episode one with Veo 3. If I wanted an orchestra to sing a song, an opera or whatever we made or even a country music song, like, it would take hours to produce that, plus the skill set, plus all the instruments, plus, recording it line by line. So, the amount of time and money saved by leveraging this art form is incredible. And so, hopefully, we'll see artists leverage it as a form or a mechanism for them to produce more greater, faster, better art, just like we've seen, like, what Photoshop did to artists, digital artists, kind of the same for digital video editing.
Sam Nadler [23:56]: The last story I wanted to cover, which is, I believe, within the last day, is about the American Federation of Teachers who says it will open an AI training hub with Microsoft OpenAI Anthropic. I have two children. So, I do think about the impact of AI on their early education and, high school education. They're not there yet. But I'll give you my take from a parent, but what's your immediate reaction?
Jordan Metzner [24:30]: I mean, it's so obvious that AI has such a potential impact on, the education complex. I mean, whether it's from early education to late stage education to you know, I'm obviously far outside of school, and I use, AI to educate myself on a daily basis. How we're gonna see AI impact the classroom, I mean, who knows? But, I mean, it seems like it's a wide open space right now. But, it's one of these things that obviously can be hyper customized to each student and their learning abilities and what their interests are, and that sounds like right up the alley of, proper education, I would say. I'm I'm no education expert. Don't really know much about it. But, we do know that, like, students are at all different paces, have all different interests, learn in all different ways. And, AI's ability to customize that, seems like a home run, but I'm sure there's probably some drawbacks too that could come with this.
Sam Nadler [25:26]: Yeah. I mean, just having two young children and being, I would say maybe compared to the average human, I'm pretty exposed to AI tools. I'm very AI forward in the classroom. With that being said, I do have some concerns. Would say kind of my mentality is to be proactive and proceed with caution. I think it can just be such an huge accelerant to so many different kids, different learning profiles, personalities. They all kind of learn at different rates, even with my own daughter in subject she likes, which is math. She has different parts of math she likes more than other parts of math. And she prefers, of course, to do the things she's good at. But in reality, she should probably focus a bit more on the things she struggles with at this time. So I think it I guess my biggest fear with AI in the classroom, and this may be a little bit higher level at high school or college is, I would want to make sure that in the implementation of AI in the classroom, it doesn't replace students' critical thinking to really, thinking, is this, just because I got the answer, does that mean that this is the right approach? Not just kind of a copy and paste approach and how to like weave in the critical thinking, really strong critical thinking skills. And I wouldn't call it mistrust, but, having that level of critical eye to anything you read online or anything generated by AI and having kind of the wherewithal to investigate on your own, I think, is a really important piece to this puzzle. But it's kind of like I see AI is kind of like the Internet, twenty years ago in the classroom. Like my daughter's entering in fifth grade in her fourth, third and fourth grade costume, they're already using iPads and different online tools.
Sam Nadler [27:42]: So, you can't avoid it. It's going to be there. How do we use it safely? How do we use it in a way that actually greatly benefits the students? I'm not an expert here, but I think, you can't ignore it. And I think it's really important to be proactive in looking for those solutions, and, proceeding with caution as mentioned before.
Jordan Metzner [28:03]: It seems like ChatGPT so far has been great as kind of like a companion. And it seems like AI is very good at being companion like in that sense. And that seems like an area where there's a lot of opportunity, especially for children, kind of in supporting their learning and education. But just look at something like Suno. Like, how hard is it for a child in fifth grade to make an entire song? And now she can just speak it out, and she has a song, exactly what she wanted to make her song about. And, that that is incredibly liberating, I think.
Sam Nadler [28:37]: Yeah. And for instance, my five year old's really interested in space. And a couple times, I've just put on voice mode with ChatGPT and let her talk to ChatGPT about space. And my five year old can rattle off more space facts than I'm guessing most five year olds. So, just having, like, probably five space related conversations, You know, it's been a very valuable learning experience for her, and she's five years old. So I think it's very interesting. I think it's needed. Obviously, I think it's it's normal to have concerns, and we should have those concerns. But I'm, excited that it can actually accelerate and expose students to way more than they would have been in like a traditional curriculum. And that's kinda it. Jordan, anything else to wrap up the show?
Jordan Metzner [29:34]: No. Great episode. Fun to talk. Lots of things going on in the AI space. Again, crazy week. And, yeah. Hope you guys like the products we built and what we showed off. And looking forward to catching up with you again next week, Sam.
Sam Nadler [29:47]: Likewise. Take care, everyone.
Jordan Metzner [29:49]: Bye.
