How to Get ChatGPT and Claude to Recommend You

So if you're not helping these large language models learn something new, it's unlikely that you're going to win in these AI search engines. I met with one of the engineers recently that we used to expect free traffic from Google, and now it's really about like ensuring that your brand actually turns up. And there's some fundamentals that you need to have in place to make that happen. You used to have to worry about just your website or just kind of the things that you own, but now suddenly you've got this surface area, which is what's being said about my brand on other websites, including review websites. And suddenly Reddit's important. So LinkedIn floats between, you know, being the 2nd and 5th most sited website inside answer engines. Instagram, like, you know, is sometimes showing up, but for AI mode and Google, YouTube is the killer.

What are some things people can do to help them get found? Jeremy Cabral, welcome to How Do I Podcast.

Thanks for having me.

You are the co-founder of Finder, and I think some people will be like, why, why are you on this podcast? But there are some incredible things that I really want to dive into with you today. One of them is around search, how podcasters can use search, and even more importantly now, AI, AI tools, and all of those things to get found. The other side that not a lot of people might know, including myself before I started the research, uh, rabbit warren that I went down, is you are actually quite involved in the podcasting world.

Yeah, definitely have, uh, a few adventures that we can talk about.

I want to talk about that as well. Before we get into that, tell us, what are you doing in the AI space at the moment?

Yeah, so I would describe myself as a fractional co-founder that people call up when they've got this situation they need to solve. It's like they've got an executive that needs to work on a key important strategy. And in the AI space, I'm basically helping these executives firstly understand AI search and what matters there. But then also executives go through this transformation not only for their company but for themselves as individuals. I've launched something called Executive Arc. It's basically an online community where You know, people in my network can come in and share, you know, what's working, what's not in AI. And, um, yeah, having a heap of fun doing that.

So, and it's so wild, it moves all the time. I feel like you have to be a part of a community because you buy a course, a guide, or whatever, it's out. Like, my brother is, bless him, he's at uni studying AI at the moment. I'm like, dude, the second you finish that day, it's out of date. Like, you totally have to be in a place where you can talk to people.

My Whoop scores, honestly, the first half of this year were like abysmal. Like, just staying up till like crazy AM, just, you know, trying to keep up. It's been crazy.

It's insane. So I think like in the world of search, we've seen this change a lot in the last few years. But even now, like you type into Google a search term and it comes up with an AI-generated response. And like if you click on that, you can have a conversation with it like you're talking to Gemini.

Like, yeah, AI mode.

Yeah, it is incredible. So for someone who hasn't— and as podcasters, SEO used to be a huge part of our strategy. It still is. Like your title is so important, the keywords that you put in, like even your captions on Instagram, but your show notes are so crucial to being found.

Yeah.

What are some things people can do to help them get found?

Well, I mean, the interesting thing with, with a podcast is that it's a walled garden. Often Spotify and Apple don't make these transcripts publicly available. So I think the first thing is like probably having a website, which is kind of crazy to say it, but I think for discovery, like, you know, obviously you can do it natively inside a platform, but I think off-platform is really interesting. You know, having a website that, you know, has the transcript, you know, takeaways, all these sort of things, I think alongside the pod is really cool. With AI search, like it's changed a lot in the sense that we used to expect free traffic from Google and now it's really about like ensuring that your brand actually turns up and there's some fundamentals that you need to have in place to make that happen. You know, pretty much at this point you used to have to worry about just your website or just, you know, kind of the things that you own. But now suddenly you've got this surface area, which is, you know, what's being said about my brand on other websites, including review websites. And suddenly Reddit's important. It's not just where—

The most.

Yeah, it's crazy, right? So how do you go and engage on Reddit without getting your company banned and so on? And it's a bit of a nightmare, but I would say it's ultimately about, you know, setting up, you know, a surface area that can be indexed by these crawlers that informs what's being said about your brand. It's largely about consensus. So when multiple websites, multiple sources say the same thing, that's what turns up in the answers about your brand. So yeah, it's a bit of work.

Okay, cool. But with the world of AI, it's easier than ever to do that and not just do AI slop across everything, but like repurpose 100%.

And yeah, so I think that's the thing here, right? Is like, we in traditional search, Google was very much against programmatic content and like, you know, generating content. And I met with one of the engineers recently and he worked on a part of, I guess, the policy and kind of the way you communicate with webmasters where they're very much now on the front foot saying it's okay to use AI so long as what you do produces value, you know, genuinely useful information that isn't just, you know, regurgitating what's already been said on the internet before. The thing is, when you're producing content with ChatGPT, we're all using it and we're like, oh, how good is, you know, it's done, right?

Yeah.

But if, if you're just like prompting it and letting it use, you know, training data or things that have already been said online, that's not useful information. It's like, you know, for me, I always think about the board-level metrics inside a company like OpenAI or Anthropic, and essentially they're all battling to, to build artificial general intelligence. So if you are not helping these large language models learn something new, it's unlikely that you're gonna win in this AI search engine.

Yeah.

For me, it's about, you know, bringing that uniqueness, the context that you have about your business, but also the things that have not been published before or the unsaid about your brand and putting it online. So that's really critically important.

And even as consumers, like we've spoken about this in a couple of episodes previously, but around like your listeners can find any information online now. Like it's not useful to have a podcast that just tells them what they can already search for.

Yeah.

If you don't have a point of difference, if you don't have an opinion, if you don't have a reason to like generate even more conversation, Like, you're just not adding value anymore.

Yeah, it's so interesting because like companies are sitting on proprietary data, and I think that's the last piece that, you know, an AI company can't come for, is like how do you kind of decide between like, you know, what do you publish and make available publicly versus what do you put, you know, put behind like a, a gated sort of thing. And so strategy for a business at this point is very much, um, about making those decisions, you know, like if you're a a business that has like IP, like do you want to make that available to a large language model and then you've got like nothing left because they've trained on that in a few years from now? Um, or is that something you're kind of keeping behind closed doors? So yeah, very, very fascinating.

Okay, I want to talk about tools in the moment because you are the man in the know when we come to tools. But before we do, we're— so we're going to create some information online. So probably like Transcript Genius. You don't have to have like— I've seen it done really cleverly on websites as well, is like they'll have like a blog post which is basically your show notes, and they'll have a little button that says transcripts with a plus button, and you have your whole transcript in there. But then it's crawlable, right?

100%, yep.

What else should they be doing? Like probably posting on LinkedIn is another really good source it scrapes from.

Definitely. So LinkedIn floats between, you know, being the 2nd and 5th most cited website inside answer engines. So—

Wow.

Really, really important. Not only the posts that you put on LinkedIn, but also Pulse articles. There are actually some that are doing extremely well. Instagram, like, you know, is sometimes showing up, but for AI mode and Google, YouTube is the killer. Like honestly, like it is obviously like they preference their own, their own platform a lot. Um, but yeah, I think having not only the episodes themselves but uploading Shorts and so on, they're all going to be very, very important.

Yeah, we did an episode about this a couple of weeks ago on how we do research for guests as well, because what I was finding— and I've spoken to so many podcasters as well, like even maybe yourself might have seen this— you go on lots of podcasts and it doesn't matter what type of prompt they've put in, you're almost getting the same questions. Yeah, now because like it's just giving like, oh, this is Jeremy, this is what I found about him, this is what you should ask him.

Yeah.

And like sometimes it's a little different, but like the bones of them are the same. So we actually started using Gemini for everything because it's crawling YouTube. It has so much more unique data that is just not like in the news or anywhere else like that.

Yeah, it's interesting because like I actually, in the, you know, when I'm going into a pod, it's like I always think, what's a new story I can tell? Something different. I mean, even just the start of this episode's already like, you know, we're entering new territory. So you've done well.

Good.

Um, and yeah, I think it's like thinking about like your own brand and like what the things you need to have out there like that would serve you, right?

Yeah.

Um, because yeah, like it can't guess what's in your head and it can't guess what's great about your company unless you literally are publishing that information in a source that's gonna be cited by these crawlers.

So yeah. I love that you brought up Reddit as well, 'cause that was mentioned so many times over Cairns Lions. That's just been as we're recording. And it's this tool that not only have you got this incredible opportunity to like talk about your podcast, but also to create a community.

Totally.

Like if you can create a space for people to come together, like we've seen podcasters do this really well with WhatsApp, with like even Instagram broadcast channels, but like Reddit is such a good place for anyone in the tech space They're already there.

Totally. And I think it's interesting because the problem with Reddit is that other people moderate. And so you can get your content like banned or your domain banned or shadow banned. Like you don't even get told it's banned. What I think a lot of brands are missing is the fact that they can actually have a subreddit that's like they own. And so alongside like, you know, naturally, organically, you know, people posting on Reddit, you could also have your own community, which will probably lead to more positive mentions and citations because, you know, generally the superfans will go and find you on other platforms, right? So yeah, I think that's a, Definitely missed opportunity for a lot of brands. And then, you know, yeah, there's a lot of, you know, don't go headfirst, you know, kind of figuring out how to do Reddit. Really do the research online and figure out how to engage on there. I feel like the key tip for me would be don't go and post a link about your brand initially. I think it's more engaging with the community first and foremost in a way that's not self-promotional, but like really contributing to the community. That kind of earns your karma points, so to speak.

Yeah.

And then, you know, if you have, you know, a post about your brand, it could be in the comments versus, you know, a dedicated post and potentially a self-post where it's more informational. Versus a straight linking to your site. Because effectively with Reddit, mods can do things at, you know, a domain level or at a user level and so on. So you've got to be careful in how you, you use the platform to be able to extract value from it.

Yeah. And that could be like, you know, maybe your first couple of replies in a thread is like your own opinion, your own thoughts, your own advice, your own experience. Experience more than anything else, I feel like, gets you those points. And then maybe it's like, oh hey, I actually did an episode with this person. We covered all of this plus this. Yeah, like let me know if you want to send them.

Yeah, and interestingly, right, on Google, like they started, um, there was a behavior that was observed where people started appending Reddit on the end of searches, and like people were seeking out Reddit ultimately as a source of information. Yeah, right. And so Google ended up switching the, the kind of results to show Reddit pretty often in the discussion and forum sections on page 1 and typically above the fold. So if you're trying to figure out where should I go to first and, you know, what threads matter the most, if you go to Google Type in the keywords that matter around your brand. You can see often like 2 or 3 threads that are literally appearing. And if the comment section is open, it's an opportunity to say like, you know, my wife's business, she rents out baby products. You can go in there and say, hey, we do this in Melbourne as well or whatever.

Yeah.

And you have a chance of not only appearing in the thread that's ranking in Google, but then appearing in the AI search responses as well. So.

So clever. Okay, so there is still such thing as traditional search as well.

Very much so.

Yeah. Is there anything that podcasters could do? Like all of this stuff is gonna help there as well. Is there anything additional that they should add on?

I think that PR really matters because it's the things that these answer engines are effectively doing. When you run a prompt, it has this concept called query fanout. So it grabs that prompt and breaks it into a whole series of sub-searches or prompts that run, and each of those go and hit the web and also use their training data to bring towards, I guess, a response that when collected altogether, they have a consensus based on that research, right? So the example I always talk about is best electric vehicles. for a family of 5 in New South Wales under $50,000.

Yeah.

You know, best, you have to look at ratings and reviews. You know, a family of 5 probably cares about, you know, cargo space and safety.

Yeah.

New South Wales, there's, you know, certain incentives of buying electric vehicles in New South Wales. And so all these things run separately as prompts and then come together to form up the answer for you. And so when that's all happening, suddenly like, you know, you've got all these websites that are brought in and like, you know, it might be review websites, you know, editors writing about the latest, you know, Chinese electric vehicle that's just been launched, and there's lots of them coming into the market. So with digital PR, you've got to look at, okay, what are the publications that are being referenced? How can I get on them? And, you know, working with a PR agency I think is probably a good move at this point. Yeah. Um, the hack I think to getting PR is often releasing some research, which, you know, you then get a media release, and then you've got a reason to be in the news. You know, every day it's like something new needs to happen to be able to write about it. So if you can manufacture something new, some new research, that's, you know, increasing your chances of getting in the news. So it's a bit of a hack.

Yeah, perfect. We have just— and I'll pop it in the show notes as well— just done an episode with Jessie Marshall, legend, who's the queen of this, all around like how you can get attention for your podcast.

Perfect.

And there's so much stuff like PR seeding, so like getting them excited before it even comes out, then going in with an angle. Like, there is so much value in that episode if this is something you want to do. So I'll pop that in there as well. Let's talk about tools. Yes, there are so many tools. There are so many LLMs you can use. What are some tools that you think the podcasters or creators, like content creators, could be using and leveraging?

I think that first and foremost, the right transcribers really matter. Like, you know, obviously when you're recording a podcast, you get your transcription, but I think that there's certain transcribers that do a really good job at understanding like who spoke, what was said, and like, you know, that kind of voice recognition piece. Yeah. And I think that storing really high-quality transcriptions in a database and then over time having this corpus of knowledge mean that, you know, you can understand from the data with your episodes, right, like what worked, what didn't. And if you're not storing that context in the right, in a very structured way, I think it's tough to do analysis and figure out like, you know, who should I interview next and what topics and so on. I actually had this idea once, probably about 6 months ago. I was like, if I wanted to pitch a podcast, like how would I go about doing it? And for me it's like, okay, I'd wanna look at who in my network's been on a pod, what topics did they talk about for that pod? Like what were their, you know, top rated episodes or most listened to episodes? And then based on my transcripts, What do I speak about often? And could they do a new episode that bring them up to date on that topic? Or is it a gap in their list of episodes where I can come in and talk about that thing? And so how could I get the introduction to the host? Maybe a person in my network, Taryn, can be like, hey, I was on your pod recently. You should chat to Jeremy. And it's this crossover between what I speak about, what did well on their episode, and what I guess the zeitgeist is. What are people interested in right now? Yeah, I think the transcribers can give you that context to be able to do analysis really, really well. I would say also outreach is an interesting thing. So data enrichment and using tools like whether it be Apollo or whatever, these tools that effectively allow you to get the contact details of people and understand who could I possibly speak to, what pods have they been on and so on. I think that's kind of an interesting piece. And obviously, yeah, you can't go an episode without mentioning Claude or ChatGPT.

Of course.

These deep research tools, and I think if you use them well with the top models right now, it's like having a management consultant in your pocket. It's insane. Carefully constructing a prompt for deep research, giving it access to the information it needs to do a good job for you, I think is going to be very important. But some of these prompts with deep research on can take, I think, 47 minutes to complete the exercise, but you get a 50-page report and here's what you can talk about. Here's the kind of gaps in terms of what we've analyzed all these podcasts. And I think that'd be very interesting. Also, there's one more, which is a cracker. It's called Apify. Apify allows you to scrape virtually any platform. So they have this—

I've used this.

Yeah, yeah, yeah. So there's effectively these actors is what they call it. It's a script that runs and can allow you to scrape Google Maps, Instagram, whatever. So if you can scrape not only your own podcast, like you probably don't need to if you've got the transcripts, but competitors like in social and whatever, I think that allows you to do competitive analysis at scale. And I think that's super, super interesting.

Yes. Okay, I want to break this all down because there's so much value in what you just said. So starting with transcripts, I think what a lot of people don't even realize, like if you ask an LLM, Claude, ChatGPT, to tell you something about your podcast and you don't have that context stored in there, it's going to use so many tokens and so much brain power going to find out about your podcast. But if you can scrape the— if you have the transcripts and you store them locally, and then you're giving that as a reference, or it's in the project, it's already got all of that information there, totally, which is so good. So a tool that we use is Descript. That's one that you can actually do like editing and everything in. There is a really good one that's like linked to an app. Do you have any others that you recommend?

I use Granola. I love Granola. Yeah.

So, okay. So, um, is there any other transcription tools you recommend?

Um, well, so it depends on like, I think with the key habit for anybody that wants to crush in AI is that you need to be transcribing. But the other piece of context that's really, really useful is what's happening on your screen. So if your transcriber has like video, like, or screen recording as part of it as well, I think like Read AI or, um, Fathom or one of those probably has that.

Yeah.

Whereas like Granola runs in the background, so it's more discreet.

Yeah.

So I think that, yeah, storing like context along with visuals is really, really important because you can actually give the recording of a video, like, you know, you say you had a call with somebody.

Yeah.

And Claude will like go and watch it back literally and take screenshots, you know, and understand, okay, at this second this happened. And based on the visual combined, like this is what we could do in this section of the episode. Actually—

Wow.

I did a fake podcast with my daughters on Saturday. I've got twin girls and, um, I film content every week for a couple of hours. And they were complaining in the car on the way and I was like, all right, I need to construct something so they can chill out. So brought them into the studio and they had a mic each. And, um, yeah, basically just did my wife's phone and, and my phone and we filmed them both. And they were talking about like sea monsters and all this stuff. And it was like literally a 10-minute conversation, but I got both videos put onto my computer and was able to basically get Claude to, you know, with some various apps connected like Remotion and others, do editing.

Yeah.

Um, auto-edit this podcast like a 6 or 7-minute podcast and it was phenomenal. Um, it found like, you know, a section where they referenced the sea monster and whatever. And then I had Higgs field connected and it auto-generated this little clip on Higgs field. So yeah, pretty cool.

Okay, this is my dream and there are some incredible people who are doing this and talking about this. We've been like trying to test it out and build it. There are little bugs that they all have. You're not gonna come out with a perfect outcome, but you've gotta keep playing with it.

Yeah.

'Cause they change so much. Like even Descript in itself, the most basic simple app, it has an option for you to auto-generate clips. And like you literally click the button, it goes, finds it, it makes the clips. They used to be horrific. They're now actually quite good. Like they're perfect for trial reels and they come in 5 seconds.

Yeah.

So talk to me about this workflow 'cause there will be someone listening. They're like, hang on, hold up. I use Claude to like tell me what to cook for dinner.

Yeah.

So how are we getting to this level?

Yeah, so I got the 2 clips, put them on my machine like locally, then I connected up Remotion MCP. So Remotion does like editing and so on. And the latest ChatGPT, SOL 5.6, you know, it was crazy. Went into plan mode and said, here's my goal. You know, my daughters, you know, did this thing and here's the kind of transcript. I had Hixfield connected as well. I was like, you know, I want you to create 6 or 8 different clips that you can kind of have play like as an overlay when they're describing this part of the story.

Yeah.

So that was just part of the plan as well. And so, yeah, and then basically like edit and cut between the videos when it, you know, made sense to. So I did all that. The first cut sucked. Like, basically the audio was out of sync with one of my daughters and it just didn't work. So then it's just with AI, I think the thing is about being persistent. You just got to keep giving the feedback until it nails it.

Yeah.

And so you just describe very carefully. It's like, you know, when my daughter Camille was talking, the, you know, the audio didn't sync with, you know, her mouth and whatever. And so it tells you, it's like, I'm going to run this type of method and we're going to do this kind of QA process because I think The key with AI is a lot of people talk about evals. Eval are effectively things that you can run to ensure that the AI output was of the quality that you wanted, and it allows that consistency. So post like me going through this process of doing this first edit and the second edit with the feedback off the back of it, I created this flow which you then save as a skill. And so it's like the next time I go and film with my daughters, literally like the two cameras or whatever, it will go okay based on all the feedback. This is what Jeremy wants when I'm producing an episode for them.

Yeah.

So yeah, I think it's like if you go and spend all this time creating that, you know, kind of tuning the, the output, always save it as a skill. You know, it means that the next time you go to do it, it's just so much faster.

Yeah, 'cause otherwise like you go back and you're like, oh my God, there was all this stuff I had to change. If you're not, it's exactly like a team member.

Yeah.

If you are just changing it and you're not giving that feedback, they're never gonna get better.

Yeah, totally.

The connection piece and like finding out how you're gonna pitch people. I love this. We had one of our clients interview Joe who's in the AI world as well. And his last name has escaped me, but I'll pop it in the show notes. Incredible. He built something similar and he was like, what sort of questions questions are they going to ask me? So I think as a guest, it can give you such good insight to like prep and research and like find that angle, which I think is really clever. The other part of this is obviously as podcasters, we need to find new angles for ourselves. So again, with that, the process that you explained sounded incredible. It sounded like a lot of work though. So how would you do that so that it could do that with like less effort?

It's interesting, right? Because like the problem, the problem you just said was, you know, I was on Mark Burris's podcast recently and Mark was like, hey, we've already interviewed Fred before. Like, I'm not sure if we can do an episode with you. I was like, okay, so how can I solve for this? So what I did was like, I got the transcript of the episode that Fred had with Mark, and then I had all of my transcripts on the things I've spoken about for like the last 9 months. And basically off the back of that, I texted Mark like, here's what we could talk about where there's no crossover between what we, you know, the two. And it's like, took me 5 minutes, and then Mark's like, let's do that. Like, quick message back, it was done. So I think it's like with AI, it's like just Preparing that context in advance. And like, it will take a lot more upfront, and so it may not feel like it's magic, and this kind of back and forth process is quite iterative. But, you know, once you've nailed that once, again, it's like, you know, that becomes your podcast pitching skill or whatever, right?

Yeah.

Um, and yeah, so for me, like, that flow was so interesting. Like, what I did was I, to get my network, I pulled from my Google Calendar every single person I've ever met with, ever. Plus my LinkedIn, stored that as context in a tool called Superbase. S-U-P-A, right? That was the first piece. And then based on the meetings that I had transcribed, I enriched that CRM that I'd effectively built with the transcription. So if I met with someone, I know what I spoke about at that time. And then I also pulled all my iMessage and all my WhatsApp. And from that, I basically have this like super powerful thing. So I effectively can quiz my CRM of like, Okay, well, like that was first piece, right? Like it's kind of the second was this tool called Listen Notes where I basically went in there and I said, okay, I want to go on like marketing tech or business podcasts in Australia. And I paid a couple hundred bucks, I think it was, but it got this export of literally 283 episodes, sorry, podcasts, all of the details like email, like, you know, how long it's been running for, how many episodes, average episode length, whatever. And which ones had the popularity scores by engagement and so on. And for that, I was able to see with my scraper, Apify, who's been on which show from my CRM, what have they spoken about, and therefore who can I get the introduction from? I can auto-generate the introduction on email because it's connected into my Gmail. And yeah, it's just a Claude code exercise that probably took probably 6 or 8 hours. You just got to keep at it. Um, but when you connect the right tools together, it's pretty magical.

So powerful. And I know there will be people listening that are like nerding out with us right now, absolutely loving everything you said. But also, this can be really simple. Like, you have an idea, it's not going to be as incredible as yours, but like you can just ask it, this is what I want to achieve, can we do this? And like, there's an incredible Australian podcast website called The Great Australian Podcast, and she literally lists any new podcast in Australia.

Oh, awesome.

It's really good. And so like, that's not going to be the level of Listen Notes, but if you're doing this for free scrappy, you just want something, get it to have a look through there as well, because I think there's really cool ways you can find that.

I mean, you know, going back to the AI search piece, you want backlinks to your website as well, right? So if you've got a podcast and you've got a website, that backlink I'm sure is going to be useful.

Yes, absolutely. Okay, what is one thing that you think we shouldn't be using AI for?

Creativity. I think that AI sucks at new ideas.

Yeah, those—

you know, some of the latest models are pretty incredible. Um, I think that Yeah, I'd still be very human in that process, like use traditional ways of brainstorming, like mind mapping, whatever. And also just speaking at a transcriber for an hour, just your own raw thoughts. I think building off the back of that is always going to give you a much better outcome, something more unique, something different. When I was building my website, I spoke at Granola literally for an hour straight and I was explaining everything I've done, everything I'm now doing, what I'm doing next. And It gave me this incredible transcript, which I then put into Claude and said, construct a mega prompt that I want to go and use inside Figma Make to design a website for me. Did that within 20 minutes. I had built this beautiful website, which was totally unique based on words I had said, but it was tuned for Figma Make system prompt. And Figma, the way I think about some of these things is like, if I'm using an AI, what is the DNA of the company and What is it good at? So Figma traditionally has been great at design. So I was like, if I'm building a website, I want to go and use an AI that is built from a company that's got traditionally strength in design. And so did that first off the back of it, had some code and then I shipped it like the traditional way, pushed it over to Cursor and we shipped the website. Nowadays it's very different that you've got Claude Design and so on. There's lots of different ways of doing that. But yeah, I feel like it's about just understanding The right tool for the right situation. And, you know, it's hard to pick often, but I think trialling a whole bunch of them, and then when you get enough signal that that was a great tool to use, building off the back of that is key.

Yeah, again, zooming out to look at that company is such a smart piece. There's— it's— we've only just got access as we're recording to a tool called Stanley Studio. It is life-changing because Stanley and Stan, the parent company, work with creators. So all the data, all the information that they have is all from creators. So in terms of like static text that goes on as a hook, it already knows what's performing. Like it's connected to thousands, like probably millions of people's accounts, seeing what actually performs. So what you do is you literally throw your clips in and it makes the video. But the cool thing about it is rather than prompting, you get to then like move it around like, like you're using CapCut. It's the hybrid of the two. It's only like, I've only played with it a couple of times, but they are obviously gonna be great at what they do 'cause they've got all that context.

Totally, yeah. And it's interesting, right? Like with my Instagram, like I only started it recently and like for me it's like when I'm trying to kick off a new project, I always think about who's crushing it and why. And studying the outliers I think is very like important. If you can figure out like what the outliers are doing, like those that outperform on the key metrics, I think it's really, really key. So I went and scraped Instagram, like, you know, key people that were already doing really well and investing like millions of probably of dollars a year into their organic content. And off the back of that, was able to find the patterns on like, you know, here are the kind of the hooks and everything that everybody's doing, auto-constructed scripts for me to then go and basically just be in front of a camera and say them. And my first video got something like, I think it was like 350,000 views.

It was wild.

Yeah, off like 800 follower account. So yeah, pretty crazy.

Okay, I think you, did you document this process and you've got it to share with people?

I've not fully, but I'm planning to.

Okay, when you do, if it's before this episode, we'll pop that in there as well because I remember watching you do this and it was so fascinating. And podcasters can do this right as well. Like, yeah, look at the YouTube, look at like Connect 1 of 10, have a look at like what's actually hitting for people.

100%.

What are these best podcasters doing? Like, one of the first things I did when I got access to ChatGPT is I put in like Stephen Bartlett, Mel Robbins, like Joe Rogan, like all of the top shows. And I was like, what are the psychological tactics they're using with their guests that like we could apply? And there were 7 of them and they were the most powerful things. Anyone that's actually implemented them, it's a game changer for their show. Like, that took like 20 minutes.

Yeah, it's crazy, right? Yeah.

So yeah, so I think that's really cool. So using Apify to connect it in and then like telling it what you want to scrape and pulling that together.

Yeah.

With your own content, is there anything else that you've done since then that you think has been the key to success there?

Yeah, I think, um, well, firstly, like committing to volume. I was like, you know, we do like 10 videos a week. And I think that if, again, like, you know, the way I think about something is like if you're trying to build a self-driving car, try and hire somebody that's built a self-driving car before. Like, you know, it's like I think was a great example, like when Apple was behind in AI, they hired like the head of AI from Google or something like that. It's like you're just hiring from people that have done it well before. So I actually work with some people that have already done, you know, some great work on viral organic accounts. And so like I'm working with people that are not starting from scratch. Yeah, I think that's really, really critical. And so the edits like work, like you're studying what's working like literally week to week versus just having this independent strategy that you just keep running it for months and then You're not looking at the data of what's working today. So yeah, I feel like very data-led in that front. I also do my content in a studio, so I think the lighting's really good, the cameras are high quality, and you know, like there's lots of ways of doing this, you know, right? But for me, it's like investing a little bit to be able to kind of get the ROI, I think is really, really important.

Yeah. And how quickly has that growth come?

Yeah, so over the last 3 months, I had a, I think my account was private kind of mid last year. I had like 300 followers or something, and I've been like collab posting with podcasts I'm on and so on. But I started, I think it was like March 15th, my first video, that one that got like 350,000 views and went from like 800 followers to like, I added 3,000 followers with one video. And I would've actually grown a lot faster. I've hit, you know, 10,500 followers or something in about 3 months, but I've actually created too much content for me to actually be able to manually post it in time.

Yeah.

Staying across all these AI tools and stuff. So I've now switched to actually scheduling the content. It probably won't do as well, but like at least it's getting out there. So yeah, it's been pretty crazy. But yeah, pretty quick, I get to, you know, to 10K in a few months. Yeah.

Yes, that is incredible. And something I know that you will probably do is look at the back end of this as well. So are you analyzing the performance and like feeding that back in?

I've got the data there, but I haven't spent enough time on it. Like, for me, it's like a— I feel like it's a critical thing that everybody should be investing in building out their own brand and like having content and so on. The data's there, but I probably should spend more time looking at it, to be honest. So I put up trial reels and I don't even see what worked. And sometimes, like, yeah, it's just like a— I'm more into the habit of posting. Yeah. Versus like, and yeah, if I want to get the most out of it, I really should spend more time in it. But it's more like, it just is always running process in my life. And yeah, I definitely need to invest a bit more time.

Yeah, it's so powerful. And like, even with, like, you want to know anything, if you're really serious about like, okay, cool, this has been amazing, let's grow even further, like that data layer. But with AI, tool too. Like, you don't have to go through and analyze it all. Like, feed it all in and tell— get it to tell you.

Yeah, totally.

We built the coolest dashboard for our clients, and we're like getting it to a place where we can share it with other people now. But I like literally will analyze them, and it's almost like that 1 of 10. But like, with podcasts, we're across like 4 or 5, 6 platforms now. It is impossible to get all of that in one place.

Yeah.

So we've built this to do that. But then having that level of insight, it does it in like real time now.

Yeah, crazy.

You can see like, okay, this is popping off. Like, let's create an extra reel about this. Let's push some more traffic. This is obviously something that's getting picked up.

So Yeah, yeah, yeah. I think for me, um, Fable 5 like came out recently and like remember how they were like getting shut down by the US government? The last thing I did in that last hour was like tell me everything about all the algorithms and like ran this crazy deep research. Yeah, and it's phenomenal. Like seriously, like a long-running research task using the most powerful model and like the highest thinking kind of reasoning or whatever. Yeah, um, is insane. Like it's very worth it.

Yes, I used it. Not Fable, but just even research. And you could probably still do this today, research on the platforms and their algorithms. Even in a basic model, it knows so much.

Totally.

There's so much that you can put into that. Love this. Let's pivot now because people may not know this about you, but you have, as I mentioned at the start, been involved with quite a few podcasts.

Yeah.

Tell us about the ones that you've brought to life.

Yeah, so, um, Two Bro Chicks, uh, was incubated out of Finder, um, and that's a, you know, pretty successful— very successful podcast. Alex and Sally have crushed it. Um, they've got like, you know, 800 thousand Instagram followers, 500,000 on TikTok, and are often like a top 5 society and culture podcast in Australia. So, um, that's a key one. Um, I've like had a very indirect role in kind of like helping inspire, um, Dune Rashid and like kind of the Female Startup Club pod. And, you know, we were on a flight, um, from Switzerland to London, and I remember having my laptop open and building a spreadsheet. I was like, hey, who do you know? And like, you know, I think that there's this cool book, you know, Tools of Titans, where If you go and interview all these people off the back of it, you could write a book and there's like, you know, interviewing and podcasting is an opportunity for you. I'm trying to think what others I've been involved in. Like we've had plenty of failures as well, you know, inside Finder in terms of, you know, podcasts that I think largely failed because I don't think people want to listen to brand, like company branded pods. And like for me, like we had like Finder money. I was like, I'm sorry, guys. Like, I don't think it's going to work. But like, you know, it's hard to be inside a company and realize that you have to sometimes create a new sub-brand to be able to succeed. So—

Yeah.

For me, when we built out 2 Broke Chicks, I was bullish on the idea that like, I think we should flip the pyramid here and like have almost like a Netflix channel, like all these shows, like, you know, targeting different segments and you build them out and then eventually, you know, there's kind of a benefit to Finder, but it should never be a direct benefit. And it's like you're basically building trust and audience and connection that over time people, I guess, will want to come closer into your world and eventually find the benefits. But it was more about, I guess, building a new, um, series of assets that would allow us to have distribution on more channels than just search.

Yeah, we see this all the time, especially with the studio now. We have so many people come in and they're big brands and they're like, we want to start a podcast. I'm like, okay, but why? Like, what is the purpose of the podcast? Who is gonna listen to it? Why are they gonna find it valuable? Like all those little questions that like a lot of brands don't think about. They're like, oh, well, we're just gonna talk about like, you know, tax. Like, no, no one wants to talk about tax.

Yeah, exactly.

And we saw massive success with this with like Swyftx, one of our original podcasts. Like we created Tapping Into Crypto. You couldn't even know that that was tied back to Swyftx unless you saw the Powered By.

Yeah.

And I know like Two Broke Chicks, when you did bring that to life, what was the thinking about that? Why was that something that you guys wanted to do?

For me, like, Well, firstly, I was obsessed with that idea, like, that we need to find a way to crack these other algorithms, you know, build, you know, audience and distribution on platforms outside of search. Um, so that was my first thing. I had a few attempts, some didn't work, or actually all didn't work. Um, but when— what made me realize, like, that this one would work is, like, Alex and Sally, um, are some of the funniest people I know.

Yeah.

And, like, I'm not their target audience. And, like, for me, like, an algorithm is like concentric circles, right? If I find them funny and like infectious, then, you know, I'm like out like kind of here and like their target audience will probably just find them like insanely funny.

Yeah.

And so for me, it's like the chance for them to go viral is like near guaranteed. Like, you know, they just connect with so many people. So that was the first thing I saw, their connection. I was like, they should work together on a project.

Yeah.

And so they started working together on some stuff and then they pitched me this idea of, um, I think the deck was called like Gen Z Proposal or something like that.

Love.

Um, and pretty much like, I think it was like them realizing they loved working together. And, um, yeah, they just went and built out this, this social platform. Initially, um, it was on Instagram and it was just like posting statics and no video. And I remember pushing them, I was like, I think you guys are really funny and you should be on camera putting out these videos, like, so people can know who's behind their photos, right? And they started doing it, they started doing really well, and it just, you know, just kept on pushing them along the way. It was like, I had this dinner with them, it was really random, like last minute got a call, no, there's someone from TikTok in Melbourne.

Yeah.

Sat next to this guy for like 2 hours straight. Um, turns out Yeah, he was the TikTok co-founder. Oh my God. And I think he's personally worth probably like $60 billion or something, I don't know. And he was talking about the early algorithm that he had hand-coded and he was talking about this concept of creation rate. And it made a lot of sense, like, you know, incentivising people to create a lot of content when you're trying to seed and get a new platform going, especially if you're trying to convince them to come across from Instagram or elsewhere.

Yeah.

Like that kind of incentive to publish a lot of content so that you get that kind of early follower growth was important. So I remember talking to Alex and Sally, I was like, we've got to go on TikTok, like it's time. And the way I would focus the metrics for them was like, um, not about revenue ever. Like, I literally said, I don't care if you make a dollar. Like, this is all about the inputs. So like, you know, I want to see like follower growth. I want to see, um, TikTok follower growth specifically. And so you just directed them towards that. Um, but honestly, like, I think they— it's just their connection. They have like, you know, so much subject matter expertise. And yeah, I mean, all their success is really credit to them. I just kind of created space for them to do it.

Yeah.

And like, inside a company like Finder, it was, not necessarily a logical play, right? Um, but I was just so bullish on the idea of us building out audience, and I just wanted to see them, you know, pull that off as a proof point to be able to do it in other parts of the business and replicate that as well. And, um, yeah, they just seriously far— like, just far exceeded my expectations. And, um, well, I knew they were like extremely special, but like they just went into a whole other level with that business.

So it's incredible. But it is such a good concept, and I love everything that you've said there, especially around like going into this with like, I am not trying to create this for my business. I'm trying to create an audience here. I'm trying to create something standalone that would actually give value that people would love regardless of the fact that it's associated with us. If someone's listening to this, they're like, okay, I wanna do this. How are we pitching this at a business level with the ROI combined in there as well?

Honestly, the only way it made sense was positioning it as innovation where like we inside Finder, when we wanted to do something new that was challenging the norm, like, Everything inside Finder is a very performance-led thing. So you spend a dollar to make 2 today sort of thing, right? Whereas this for me was like, we're trying to understand these algorithms, we're trying to build off platform and do things differently. I was like, this is a personal project of mine. I'm treating this as R&D. And so like, I'm gonna fund this thing to get to that goal of learning so that we can then ideally shift the culture and understand that there's way more opportunity beyond websites. And performance marketing. And actually, we need to learn how to build audience on these platforms. So that was it. And like, literally, at times, like, you know, I think that it wasn't, wasn't necessarily supported because they're like, you know, why are we still doing this thing? I was like, I'm telling you, it's the way we need to go. It's directionally so important for us to figure out what to do given that, you know, search is changing like we've been speaking about. You know, I was kind of trying to front-run all of this. Um, and yeah, I I'm a firm believer in the idea that, you know, it's kind of the number of hours that people listen to, like, of you speaking. Like, you— I think, um, Daniel Priestley talks about this concept, right? Like, if you— I think it's like 200 hours to become like a super fan. Like, so for me, with audience building connection with the creators over time, I think that you ultimately have much more influence over the decisions they make and what they do and don't do. And not in a controlling way, but like in, you know, using that for good, right? You want to be able to shift consumer behaviour to take action. I think that, you know, the direct route always doesn't always make sense.

But Two Broke Chicks, genius in this business model that it just like kind of encourages you to think about where could I get a better deal, which is like kind of tie that back to Finder. Like that's the whole point. Like we're looking at other areas, we're trying to see what else we can compare it to. Was there a direct business benefit for it from it?

Um, I think that's initially no, but But so for me, I am with Finder. Like one of the things I wanted to do was get beyond finance comparison. Like we had done all that and I was bullish on this idea of like entering online shopping. So we started off in promo codes and we're doing like ASOS promo codes, you know, Showpo, et cetera. And it did really well. And so the problem with that though is like if you're a promo codes business, like you are capturing demand that's already been created. So if you rank for, you know, ASOS promo code in Google, people rock up on your website and the brands don't really like the fact that you just rank for the keyword and you're taking a clip of the revenue.

Yeah.

So for me, I was like, well, we need to create the value and the demand in the first place. So that looks like going into editorial. And we had— we tried all sorts of things like restaurant reviews. Like, this is kind of insane. But I think at one point Finder ranked for like best Chinese restaurant in Sydney, like number 1 in Google. Yeah, I'm serious. Like, so, you know, we were good at SEO, right? Yeah. And yeah, so for me, it's really been about like being a part of everyday life for people. You know, we launched something called the Retail Awards for Finder, so you can see a Finder sticker on the side of a toaster inside JB Hi-Fi. And like, for me, it's like about being known for the trusted brand that does objective research, that is, you know, consumer-first and is, you know, as close to independent as possible, right? And, um, yeah, so that was, that was really the goal. And I think with Two Broke Chicks, they have such, you know, ethical standards, like in the way they kind of work with, you know, which brands we promote, not promote. Um, but they're also savvy financially too. Like, they're figuring out, like, how do you get a deal? And so you're not spending, like, huge amounts of money. Like, there's the, you know, the fashion dupes and, you know, beauty dupes and stuff like that. I had no idea about this sort of stuff, but yeah, they're absolutely expert in it. Um, so yeah, I think they kind of just found this excellent way to kind of solve for what their audience was looking for, um, using the tools that, you know, kind of, and the skills they learned at Finder. But then well and truly going beyond that, they just have like this mastery of like, what do people want to be listening to right now and how can we connect with them as people?

So yeah, and if you look across every incredible brand podcast that's done this well, the culture is the same. Like you look at the business, you look at the podcast, the culture is the same, but there's always that additional value, that separation piece that makes it stand on its own 2 feet, build its own community. And then like you become that trusted voice. It's like, oh, like they are associated, maybe I should go and have a look at that. Like, you know, of course there's advertising opportunities and everything you can do within the episodes as well. If you If you're going to give advice to a brand that is listening and they're like, okay, I want to do this, I want to have a branded podcast, what would it be?

I think that to my point before, like if you're creating a podcast, it's not meant to be immediate ROI. It's about building trust and connection with your customer base and future customers, right? Like I think often it just takes time for people to like, you know, listen to content and they're like in the, you know, the funnel, so to speak, deciding, you know, if they're going to contact you for a long period of time. interaction points over many months before they finally go and fill out the website form or drop you a DM or whatever to kind of work with you, right? So I think it should be first and foremost about being connected with your customer between those kind of interactions you'd normally have, like, you know, meetings and, you know, helping them out with whatever you, you know, sell or do as a business. It's about education and kind of being on their phone, like, you know, whether it be like they're out for a run or having a listen to you. Like, it's just that constant connection with your audience and your customer base.

So, and any client that we have that has a service their podcast is always their number one source of lead gen.

It's awesome.

Again, it might not be like directly attributed because the last place they're going to click is Google. So you think like, oh, it's coming from Google, but you get on a call with them, they're like, I've listened to all the podcasts, don't need anything, I'm ready to pay. Like, they are so bought in because they've got that trust, the rapport, they feel like they know you. We've had clients and they've had literally people on calls like reciting words back to them because you've built this parasocial relationship with you. They absolutely obsess with you and they love you. So love that advice. For you, you obviously have the talent within. If someone else is going to do it, How do you think they should approach finding talent?

Well, it's interesting because like we didn't necessarily have the talent within. Like I remember hiring and thinking like, you know, can this person do well on, you know, a visual platform? So like, you know, I'd hire certain people that had, you know, great personalities, sense of humor, like were, you know, yeah, like I said, this infectious piece. Um, I think that if you're looking to experiment and like, you know, figure it out. I think hiring people that are naturally like using the platform day to day is like definitely the number one tip. Like people that, you know, are native to TikTok or native to, you know, they've done something as a side hustle before, as a blogger or whatever it is. Like I think that's always like de-risking that investment. But I think that, you know, the key thing, you know, from my perspective is like I don't think that these channels should be measured like paid ads, right? Like where you're like, Like literally saying, okay, how much money did it make this month? I think it just needs to be an always-on, something you just invest in. Like, it's almost like, you know, you have an insurance policy, right? Like, it just has to be invested into as a business. Like, so for me, it's like, it just should be an always-on that every company makes that investment because it will always provide ROI, but it's not directly measurable.

And how do you think, like, obviously this person is going to be associated with your brand if you're bringing external talent in, or it's going to be like the founder or someone, or someone within the business. There's obviously that risk that they could leave the business.

Yes.

Or be doing stuff that doesn't align with your brand.

Yeah.

And obviously the girls now have gone on to do their own show.

Yeah.

Post Finder as well.

Yeah.

So how would you suggest managing that?

It's super interesting because, um, in naming something, it's useful if you don't call it like, you know, the, the name of those individuals that run the thing. I think that's the first piece.

Absolutely.

Um, also, I think, I think it's an amazing thing if somebody kind of takes something and, you know, is able to, to run with it, right? Like, you know, I think the way that people are incentivized really matters if they're able to feel really bought into what they're doing, whether it be like a structure, like from financial incentives or things like that, that allowed them to really want to be, you know, at night, you know, replying to comments on Instagram, right? Because like realistically, I think if you, you know, you're hiring someone as a 9-to-5 job, you'll just never outperform a creator. So they need to be incentivized in ways that go beyond a salary. And so yeah, I think that's, that's the key piece that unlocked it. Like we always had these incentives for people that were working on social content where They had skin in the game. And yeah, I think that's the hack for sure. Yeah.

And if you're doing it as a founder, you need to have that team around you. Because if you're not trying to build a community, if you're not replying to social comments and stuff like that, like, why would people want to be part of it? Like, they might, but that's such a huge piece that we always talk about as well. I have some final little questions for you, which are fun and in the AI space, a couple of them. So what are your thoughts on AI clones?

Look, I mean, I've been trying to clone myself effectively for like 9 months straight, like extracting all the context I can about, you know, as I said before, like all my meeting transcripts, blah, blah, blah. In terms of like using AI clones in content, are you talking about specifically? I think, I don't know. Look, I'm very much a believer in like human-created content and I think it'd be pretty sad if like everything we were watching was just like AI clone, right? So look, I don't mind the idea of like if you say sold an AI assistant that was like trained on your knowledge and it's very clearly on the box, like, you know, on the tin, it says like, you know, it's a clone of me and it's based on my knowledge, that's fine. But like in replacement of you and not being authentic and kind of transparent around that, definitely against it. But yeah, I do think you should I think you could definitely train up AI to like think as close as it can like you, but not necessarily pro AI-generated content versus human.

No, me neither. They give me the ick. They're just like the mouth and the mouth. And the fact it was so funny, one of my girlfriends, she did a clone just to like see what it was and test it and put it in the group chat. And one of the other girls was like, it's not you, it's 4 heads moving too much. Yes, it's the way we can tell. Look out for the lines. Um, if you could only use one AI tool, one LLM, what would it be?

Well, honestly, today I would've said Claude Code last week, but I think Codex right now is phenomenal. Like it is a little bit more unhinged than Claude is, so you can get it to do things. And I think that, yeah, you just get it to do things that like Claude would just kind of, it's too guardrailed at this point.

Yeah.

And the fact that its top models included in the subscription plan means that you're getting incredible quality versus Fable, which is about to switch off to be API credits only, right? So yeah, Codex is where it's at.

Codex is incredible. Like one-shotting things in Codex, like the amount of stuff it has built for me. Like I'll be like, I wanna understand this data. It's like, here's the dashboard. I'm like, no, but okay, thank you. Yeah, it's so good. Okay, okay. What's a tool that you wish you had? Something that you wish existed right now?

You know, I think the thing that is really hard to do is create high-quality diagrams. And like, you know, like when you have a concept, like we've spoken about a whole bunch today, To be able to visually represent a concept in a diagram is a really tough thing to do. There's a guy, Taki Moore, who's phenomenal at this sort of stuff, right? Like, if I could create Taki-level models and it looked great, I would just love that.

Yes. You have to ask Tony. I feel like they'll have a way to do that.

I'm sure they do.

Yeah. Yeah, of course they will. If there's anyone who's figured it out, it is them. And just, I love the way his brain works. But yes, genius idea. One question we ask everyone to the podcast, what is a podcast you keep coming back to and why is it on your rotation?

I, um, I mean, I've got to go back to the Tim Ferriss Show. Like, that for me was— he's the OG in podcasting. I, um, I met Tim probably a decade or so ago and like, you know, kind of saw him speaking here in Sydney. And like, he just is such an incredible guy, inspirational in that he's just an avid learner. And like, he's always deconstructing what people, you know, drives people, like what high performers are doing. And across so many different topic areas. I just think he's phenomenal. So yeah, Tim Ferriss Show.

I feel like anytime you listen to it, it could just be like completely different centers of the field. Like there's such different conversations, but it keeps it so interesting. Where is the best place for people to connect with you?

Um, jeremycabral.com is my website, um, but I'm on, you know, LinkedIn. I've got my Instagram going now as well. Um, yeah, just drop me a DM and we can chat.

Perfect. And then if you had to give someone who wanted to grow their show one final piece of advice, what would it be?

Study the outliers. I think that find the people that are already crushing your space, reverse engineer what they're doing, and just start from there outwards. And I know it's less fun because you're being less creative in the process, but it de-risks your investment and makes it worth your time.

Yeah, and you can put your, you put your own spin on it. It's not like you're copying word for word.

Totally.

Yeah, I love that. Thank you so much for joining us.

Cool, thanks for having me.