The Bootstrapped Founder

Arvid: 00:00

Hey. It's Arvid, and this is the Bootstrap founder. Today, I'll be talking about a successful marketing project within my own software business that turned out to be so successful that it spawned a business built on top of that business. And when this episode is released, it'll be one day since the platform has been live. So I'll be sharing the most raw and most timely insight into what happened right now, what the story is, how it happened, and how I went about building this particular project.

Arvid: 00:33

This episode is sponsored by paddle.com, my merchant of record payment provider of choice, who's been helping me focus on Podscan and the project that I'll be sharing with you today from day one. They're taking care of all the things related to money so that founders like you and me can focus on the things that only we can build. Paddle handles the rest, sales tax, credit cards failing, all of that. They take care of it. I highly recommend looking into it, so please check out paddle.com.

Arvid: 01:02

The thing I built is called the Podscan Ideas newsletter and on top of that newsletter now sits the Podscan Ideas Vault which is a platform that tracks, scores, ranks and analyzes all kinds of business ideas that it pulls from the full transcript of fresh episodes in real time of around 500 podcasts of all kinds mainly focused on the entrepreneurial business finance philosophy and science shows out there. Now I know it's not a novel idea to extract information from other media. Right? The whole idea of travel guides or rewatch podcasts for that matter, I watch and listen to quite a few of those, is pretty much an indicator that there is value just from extracting things from existing content created by other people. But Potskin, in particular, has been extremely well set up to extract information from a whole industry's worth of insights and content in a very, very easy way.

Arvid: 01:58

And I'm just gonna share how all of this came to be. Because I have several customers on podscan.fm that have lists of podcasts set up with a webhook. Whenever a new episode comes out for any of these shows, it sends the full info about this episode to their servers, and they send it through a lot of cleaning up. Right? They take the full transcript that I provide with every single show.

Arvid: 02:21

They analyze it, have an AI summarize it, and do a lot of stuff with it, different things for different people. And then they generate some kind of content from this data and send that out or make it available to their customers. Imagine, I have a couple of people just taking podcasts and summarizing them and turning that into a newsletter. It's really as simple as that for some. But there's one really interesting one.

Arvid: 02:44

I have a particular customer that works in the medical field. He's a doctor who summarizes very specific podcasts in a very specific subfield of his industry, like a subdivision of a particular kind of surgery. It's very, very precise, very specific. And he takes all of the podcasts in that particular industry that are run by experts. He extracts the most relevant information from all these shows and then even translates that into a foreign language so that the people who operate in that field in his particular country are able to get all the information from all these many hundreds of hours of podcasts that come out every week at a glance.

Arvid: 03:26

He even has the capacity to show only the episodes that are relevant to one particular person and their sub interest in that subfield of medicine. It's really, really cool. It's really inspirational, and it's inspired me to do something about it myself. Because I recently thought about what a success story that actually is for the platform. Right?

Arvid: 03:45

From my own marketing perspective, somebody is pulling all this information through my API, through our webhooks that we send out, the real time webhooks that are activated whenever a new episode comes out and we finish transcribing it. And we sent the full transcript to anybody who has an account on PodScan and is tracking to a particular show. So it's super useful. And I was like, this is a great example of how PodScan can be used. So why don't I tell my audience about this particular use case?

Arvid: 04:11

There's probably a lot of people out there which could use this, which I then did a couple of weeks later. I started working on sharing not only that this story exists, I shared that on Twitter, which is cool, but also promising that I would build an automation using the existing automation tools out there, things like Zapier or N8N or Make, just to show people how this can be done reliably using that platform and Podscan's API. Because I wanted to show, hey, Podscan can be easily integrated, and I wanted to reach more customers. Right? That was the motivation behind all of this.

Arvid: 04:42

So why not do development or coding or integrating as marketing here? There's a phrase called engineering as marketing, and I think this falls under that umbrella. I'm building something, something, sharing the process of building it, and then the final product so people can see, I could do this too. So I started building an automation on n eight n, which is a very highly flexible but more technical platform to build integrations in. I think Zapier, probably I could have built this too, but n eight n is this whole flowchart of all kinds of things.

Arvid: 05:11

It's quite developer y. Right? If you're a dev, you immediately understand what it is. If you're not technical, it might be more complicated to get into. You'll still figure it out, but it has a couple of components where you need to understand the underlying technology.

Arvid: 05:24

Other platforms are more plug and play. So I took my webhook that came from my podcast list where I just added a couple 100 podcasts from all kinds of interesting industries, And then I took this webhook, pulled it into n eight n, and generated a summary for each of these episodes and saved that in a Google Sheet. It's really that simple as a basic functionality. I just used tools that I already had active. Right?

Arvid: 05:47

I didn't have some kind of software service or a back end or database. I just used Google Sheets, OpenAI's APIs, and everything else that n eight n offered. And that was just more logic, like if then else and switch and summarize and all of that. I wanted to create a newsletter, just like my customer, out of this data from the podcast. And I thought, well, my audience isn't medical.

Arvid: 06:08

My audience is entrepreneurial. Why not take all the good ideas that happen to come from all of these podcasts? Because people talk about oh I wish this would exist or hey this industry has these kind of problems somebody really should solve it and here's an idea I had I'm not gonna build it but you know maybe look into it people talk about this all the time so whenever that comes up an AI can find these things right can find these occurrences and take out the idea flesh it out a little bit and present it so I could turn it into well this could be a solution here you could build I don't know, an AI powered marketplace that sells these things to these people because it knows exactly who they are or because both are trying to look for sellers and buyers and it combines them in an interesting way with things like you could turn this into an info product because people say, I wanna read about this. Well, why don't you just build something that facilitates that and gives people that information? There's a lot of good stuff out there that people need and talk about.

Arvid: 07:03

So in my workflow on n eight n, I took my webhook and I generated this idea based summary just using OpenAI. I posted the full transcript into an OpenAI prompt, asked it to take the best three ideas out of it. And then I would write those ideas one by one into a Google Sheet. And once a week, I thought initially, but it turned out to be once a day, I would take all of the items from my Google Sheet, rank them in some way, score them, take the top 10, and turn those top 10 into an email, that markdown document that also was generated by an AI, and feed that markdown into kit.com, formerly convertkit, kind of an email service provider, and just create a new broadcast. That's what they call an email, and send it out to everybody who subscribed to it.

Arvid: 07:47

That was the idea. Right? Get webhook, get summary, save to sheet once a day, pull all the information from the sheet, and put that into one email. Now Kit has a really wonderful integration. So you can have a landing page where people can easily sign up, and you can easily tag them so you know exactly who wants to read this newsletter.

Arvid: 08:07

I very quickly got something working here that just created a nice list of the top 10 ideas and had additional information for each idea, like how feasible it is, how much money you would need to spend to get there, what the challenges are, the opportunities, and how you can potentially measure the idea if it's implemented right. All of this turned into a newsletter ready to be sent out automatically scheduled even just from my incoming webhook information that has a transcript for each podcast episode. And a couple hours later, you know, just kinda ironing out the kinks, I had a fully working workflow. I took a screenshot of this, just the n eight n workflow page, and I shared that on Twitter and got a lot of positive feedback. People were like, I didn't know that this API could be used in the first place, but I didn't know you could even do this, like, pull information, generate a newsletter out of it.

Arvid: 08:55

And people were signing up for PodScan to begin with to see if they could themselves build something without knowing exactly how to do it because I only shared the workflow, not how I built it. But people were just trying to experiment with that, but that was already a win for me and for PodScan. Right? I already had some interest from the community on, I could see myself using that API for something, people sign up. And I got a lot of DMs on Twitter, and that's the interesting part, from people that would really like to see me walk them through it so they could set it up themselves, something like that for their own industry.

Arvid: 09:28

So I decided to make a video out of it. I recorded myself explaining the already setup process in the edit and workflow. Every single thing, I talked about the full configuration, why it's that way, how it happened, how you can pull it together. Here's where the webhook comes in. Here are the fields you can expect.

Arvid: 09:46

Here's how you can create a summary. I even shared the prompts for each of these steps. It's in the YouTube video's description, both the summary and the newsletter generation, and I shared which APIs I use, what the kit.com API looks like, how to deal with that, and what Google Sheets can be used like. Obviously, anybody could use whatever they want, but I just shared how I approached it in the video so people could rebuild the exact same thing. And I shared that video on Twitter to great success.

Arvid: 10:12

A lot of people signed up to PodScan just to see if they could build something like this now that they have this blueprint. But here's the thing. A lot of people were interested in the actual outcome. That newsletter it wasn't just an experiment to show how Podscan worked people wanted to actually get the business ideas too and they told me so I doubled down and made the newsletter generation tool better and better I refined the prompts and all of that just made it more cohesive but it was still Just Newsletter. And people started signing up for it.

Arvid: 10:41

It's really great. But Just a Newsletter is wonderful because it's always gonna be this way for me to advertise Podscan being the source of all this information in the newsletter, but there's so much missing. Right? In 500 plus podcasts that I currently track for the newsletter, there are a lot of ideas. Most shows have an idea of some sort.

Arvid: 10:59

I get hundreds of ideas a day, but only a couple of them make it into the email because 10 is a lot to read through in an email. But what do I do with the other ideas? People wouldn't be able to get them because I just take the top 10, but the others might still be interesting and they might be super niche, but somebody might have the perfect idea for them there. Right? So I thought, how about I build the platform where I save all of these other ideas and make them available searchable to anybody who's interested and obviously this being an extension of Podscan because it's sitting on top of Podscan's APIs and Podscan data I could likely monetize that a little bit as well the newsletter would always be free Cool ideas should be out there.

Arvid: 11:38

But 10 of them for free a day, every single day, that's thousands a year, that's wonderful. How about I make everything free that happens over the last twenty four hours, both in the newsletter and on the platform? That was my initial thought. And if it's older than twenty four hours, it's pretty much an archived idea and something that came up last week and you wanna look at it, something like that. I will make it available for a fee through a subscription to the platform and then some.

Arvid: 12:04

That's kind of the basic idea for the Podscan ideas vault. And once I started thinking about this, I very quickly came up with this idea of the newsletter vault. And not just newsletter vault, the idea vault. Because a vault would be a place where all the good ideas are stored and stashed where you can search for them, search for the ones that are related, do deep analysis on one idea. Maybe you could even chat with an idea to see what the potential problems are beyond just a couple of paragraphs that are in the newsletter, like, really dive deep, could do a deep analysis where you have a really smart system, look into the potential pitfalls, all the potential little challenges along the way.

Arvid: 12:41

You could even have the opportunity to click a button and have the idea be built by tools like LaBaba Bold, VZero, any of these tools that allow you to prompt your way into a product. Well, I have the prompt because I have the full idea, right, in in great detail, and I have a system that could easily make this available. Well, it's probably not gonna be a big surprise. All of this is possible right now because I built this as a software as a service platform on ideas.podscan.fm and it solves exactly this problem. It's the Podscan Ideas vault.

Arvid: 13:11

It's a really simple Laravel 12 project that has been completely coded by Cloud Code. I wrote maybe two lines of code in this manually and everything else was fully automatically prompted into existence. I told it to use Laravel Spark, which is the Paddle integration that I use and Paddle being my payment provider of choice was the thing that I went for immediately because I thought, yeah, I wanna monetize this as quickly as possible. So gonna use the thing I know. I used the default Tailwind design because it's also very data focused, and I don't need to have fancy graphics or anything.

Arvid: 13:46

The whole platform is very much focused on the data of ideas and the concepts that I'm interested in and that founders might be interested in. And I integrated a couple of webhook endpoints where now that I had this platform running, instead of saving data into Google Sheets, which I ran into, like, rate limiting issues on occasion, this application would now host all the ideas. I would just save them to my own database, which is all hosted on Laravel Cloud. So it's a serverless Postgres database in there, very cheap, very reliable. It's very easy to set up, and Laravel supports this stuff very, very quickly.

Arvid: 14:23

The database would also host all the past newsletter episodes. It would be an archive of both the ideas and the newsletters that they were in. So my n eight n automation could then slowly be either turned into a completely internal process. I'm not there yet. Or at least I could use these endpoints of the Ideas Vault to store and fetch information from and not rely on the Google Sheets anymore, which made it a lot easier.

Arvid: 14:45

It's much easier to query information than having to sift through a Google Sheet. Even though it's possible, this just made it easier. I built this over two or three days. I created a new account on Paddle, which was verified a couple days after, and then I started integrating all the relevant secrets into the project. Now you kinda have to test your payment system, and then you go for the production system.

Arvid: 15:05

Very easy to do with Variable Spark. Just a couple of environment details that I had to add, and that was it. And now it's a fully functioning software platform that hosts, at this point, almost 1,500 ideas that were extracted from actual individual podcast episodes with quotes from where that thing was mentioned, with the actual location in the conversation where the idea came from, and what the idea is and what founder it's for, how much it might take you to get to any revenue, how much time it might take, and what the initial investment might be. All of this constantly automatically extracted from podcast episodes as they are being released in hundreds of shows. It's a really cool system, fully automated, and I added a function for people to suggest new podcasts to track that I can then add to the system.

Arvid: 15:51

I added functions for people to chat with ideas, to even favorite ideas and have a conversation with those. You can see past newsletter episodes. All of this is pretty much available for free for anything that was added in the last twenty four hours or so. Yeah. It's it's, I think, exactly twenty four hours that stuff stays available.

Arvid: 16:09

Any idea, any newsletter episode that is twenty four hours or younger is completely available to anybody who's interested, both if you're logged in or not logged in. In the moment you want to go to an idea or a newsletter that is older than this, you're asked to purchase a $10 subscription. It's been very simple to build because Claude is highly capable of building these kind of CRUD applications, right, to create, read, update, delete apps that just interact with the database and show things from a database, and then you add comments to it, have a couple relationships there. Cloud is really good at this, so it was really easy to prompt my way into. And this has now been operating for a week.

Arvid: 16:44

I was just testing the system, see that it is reliable. And yesterday, as of this episode being released, the newsletter itself contains a link to the ideas platform, an actual link to the idea hosted in the ideas vault for each idea in the newsletter. I started sharing this on Twitter. I'll see where this is going. Like, you're really in the middle of this right now, so I cannot tell you more than I know because I don't know anything about it just yet.

Arvid: 17:06

But I think this will be an interesting source of revenue for PodScan, which it is because, this is a database that just keeps growing in terms of the value of the content, the ideas. I have vector search happening in there so people can find ideas for keywords that aren't even mentioned in those ideas but finds it by proximity. I think it's a really interesting project just to source information from what's already out there. But it also has the capacity to show people how easy it is to set up something that sits on top of the Podscan API. So that's the win win part here.

Arvid: 17:37

Right? All the podcasts, all these episodes, all these transcripts, they have so much valuable data in there, and people get to see this. This is just one way of showing how this can be used. So it's a business on top of a business. I think I'll look into building similar things, other ones just like it over the next couple of weeks.

Arvid: 17:54

But because this is so heavily automated and it's such a daily injection of fresh ideas into the community, I think it's gonna be really a helpful little tool to pull information from podcasts out there and inspire people to do this for their own industries. And in my own numbers, in the prospects on PodScan itself, this has already shown an uptick. People have been signing up. The idea of sharing, the whole concept of taking podcast content and reworking into condensed extracted information that you then share periodically, that's not just my doctor customer's idea. It's not just what the Podscan Ideas Wall is about.

Arvid: 18:28

That's a generally applicable concept that can be used for podcast data in almost any industry. So if you have experts in the industry that you serve or that you're part of, you can pull all this information that they have, that they share on podcasts out, compress it, and share it with the yet to be experts in that industry. Any industry can benefit from this, and that's so wild. Right? You could have construction people that do, like, a certain kind of architecture, or you can have people in biology.

Arvid: 18:57

You can have nerds like Star Trek fans, all of this. Like, there's so much in any industry. Podcasts have some kind of info that people are willing to actually pay money for. You could turn this into automated TikToks and monetize through that. You could turn it into newsletters, obviously, paid ones too.

Arvid: 19:14

Could turn it into individual emails that you send out to individual people. It really doesn't matter. The data source is the same. That's the Podscan API with all these podcasts. We even have a Firehose that gives access to all the podcasts as they are being released.

Arvid: 19:27

So you could technically serve every single industry if you were building the right platform to determine what belongs where. So there's a lot of opportunity here. So that is what I'm trying to do right now. It's engineering as marketing and business building as marketing. And it's both an opportunity to get people to have a subscription on IdeasVault and to see the power of PodScan and become a subscriber there.

Arvid: 19:48

So, yeah, that's my latest project. I think it's really fun. It's been a lot of fun to build. And please do check it out at ideas.podscan.fm. You'll see all the latest business ideas sourced from these shows there for free.

Arvid: 20:01

And if you're looking for business idea, these are really, really cool. And all of them link to the podcast where the idea was discussed, so you can check that out. They're all scored. They're analyzed. Might be really interesting to find the thing you wanna build.

Arvid: 20:12

I've seen a couple really, really great ones. And, course, there have been challenges, technical challenges in making this tool work, making it interesting, making the ideas good. I think I'll share more about this over time. Already had to implement quite a lot of things to make ideas pop more, to make them more valuable, which is, I guess, trade secret at this point. But, you know, I'll share more over the next couple weeks as I learn more about my customers here and what people wanna see, and I just wanna see where it goes.

Arvid: 20:37

So let me know what you think. I'm always excited to hear, particularly in the super early stage, where this could go, where you wanna see it go, and all of that. So I'm excited about the Podscan ideas vault and the power of Podscan being shown in a real project just sitting on top of the API. And that's it for today. Thank you so much for listening to the Booster Founder.

Arvid: 20:56

Find me on Twitter at obid kah, a r v I d k a h l. If you wanna know what everyone's saying about your brand on 3,800,000 podcasts, podscan.fmtracks mentions with a powerful API that turns podcast conversations into actionable data. And if you're hunting for your next business idea, get them delivered fresh from the podcast world at ideas.podscan.fm, where AI extracts the best startup opportunities from hundreds of hours of expert conversations every day. Share this with anyone that needs to stay ahead of the conversation. Thank you so much.

Arvid: 21:27

Have a wonderful day,