How Leo Morejon Builds AI Tools with Lovable Cursor and Vibe Coding
Leo Morejon · January 16, 2026
vibe coding
**Introduction to AI Insiders**
(0:05) Hey everyone, welcome to another edition of AI Insiders with AI Marketers Guild by March Media. I am David Burkowitz, your usually friendly host. We've got a much more friendly guest today, Leo Morejon, a multiple-time colleague, multiple-time collaborator. He was with me from the start. He actually built the first Ad Markers Guild website.
**Leo Morejon's Entrepreneurial Journey**
(0:32) He's been with us in prehistoric days, and Leo has been more and more entrepreneurial, always doing cool stuff beyond these amazing hats he's worn on the agency and brand side. He can tell you all kinds of things about winning Guinness World Records for cookie brands and crazy stuff.
**Focusing on AI and Leo's Background**
(1:00) Today we'll focus a little more on the AI front and the stuff that Leo is able to ideate and come up with and market fast. It's cool. So why don't you share a little about who you are and anything I didn't quite cover yet?
No, of course. I appreciate being on. Thank you so much for having me. I appreciate everyone joining to hear me
**Leo's Career Philosophy: Gold vs. Shovel**
(1:23) speak and hopefully have a really fruitful educational conversation that really supports you and your work and everything you're doing. Berky, thank you for the kind words and for having me. I look at my career in two different ways: either I am looking for the gold or selling the shovel. I started in big ad companies from WPP, JWT to 360i at Densu.
**Career Shift: From Hunting Gold to Selling the Shovel**
(1:46) I look at that, I'm looking for the gold. I'm out there hunting. I'm out there digging. I'm out there shuffling, looking through what can I build and do that is amazing and impactful. Then I went into the technology world after that, still within the realm of social media and marketing. In the MarTech world, I look at that as selling
**Enabling Marketers: Entrepreneurship and Hormel**
(2:03) the shovel. I am giving the tools to other marketers so they can go and find the gold. From there, my experience goes into entrepreneurship to working at big brands. My day job currently is at Hormel, the makers of Spam, Applegate, Justin Peanut Butter. I'm running all social media, influencer marketing stuff. As Berkie mentioned, I
**The Best Time to Be a Builder**
(2:26) am always building. I'm always learning. I'm always testing. The last few months, I've shipped faster and built some of the coolest things and most successful things that I ever have in my whole life. I think right now is the best time to be alive for someone who's a creative, for someone who's a builder, or any marketer. If you're not getting up and
**Embracing AI Possibilities**
(2:48) getting excited every single day, maybe a little scared, you're really not testing and looking at all the possibilities that are out there. David: Why don't you dive in because we could ask why you're not so scared about AI taking your job, but I feel we all have our reasons and some of us might
**The AI "Aha!" Moment**
(3:10) even want AI to take some of our jobs. The stuff you are building, what's great with it is, was there even taking one step back, was there an aha moment for you where you're, "This is fun now, or I can do things I couldn't do before now?"
Leo: I remember it very distinctly. It was about my son's about to be 3 years
**Pre-ChatGPT Discoveries**
(3:38) old. It was right before ChatGPT came out. There was the playground so you could still play with GPTs. I remember going in there and building simple calculators or word replacement tools, and if this then that, just for text. I was profoundly amazed and felt I discovered gold. The only other time I ever felt that was
**Social Media and AI: World Changers**
(4:03) when I discovered social media in one way, before it was called social media, just doing some stuff on MySpace. I'm, "This is going to change the world." I saw the future. I'm, "This could do marketing. This could do coding. This could do essentially anything." I still remember that, and it coincides so well with my
**Personal Connection to AI's Emergence**
(4:23) son being born. There's a lot of positivity in my life looking at all this, and just being able to play with it is incredible.
David: So he's around the same age as ChatGPT, huh?
Leo: Yeah. Yeah.
David: That's super fun and just something that sticks out in my memory. But yeah, I remember working, and then
**Chat GPT's Accessibility Shift**
(4:42) when ChatGPT came out where you could actually talk to it, that was also incredible. This is going to change and make it a lot more accessible to everybody else, not just this playground that looks almost not a terminal, but just a little form that you work in.
David: And then you mentioned this new
**The Vibe Coding Phase Shift**
(5:02) phase shift or something to that effect happening. You said it several months ago, and what happened then? Why? What did it for you?
Leo: I think all these vibe coding platforms from Cursor, Lovable to Replit, to anything else that's out there. I've always built things, but I am not a coder. When I say I've built things, I probably should say more like I've
**Producing Ideas vs. Coding**
(5:23) produced things. I've always had a lot of ideas. I've hired. I've partnered with technical people. I've gone out there and tried to put scripts together, but it would take so long. It'd be so costly. The resources would take up so much energy and time that things would ultimately fail. Even if I had successful things, I actually had one of the first successful-ish
**"Prompt to Product" Tool**
(5:49) products when it comes to combining ChatGPT with some kind of marketing thing. It was called Prompt to Product. What I did was I allowed you to get a WordPress setup plugin where you could write a prompt but then turn that prompt into a feature. We actually used it for one of your earlier websites, Berky or David, we
**Translating Resumes and Recipes**
(6:10) looked at your job, right? Or your resume, and then we translated it to be more marketing. I had built one that is put in a recipe and then make it gluten-free or make it vegan or whatever. This is just on the back end, a few forms that you would fill out and say, "If someone inputs a recipe, make it
**Maintenance Challenges**
(6:30) gluten-free." You could put it into any WordPress site and make it a feature or a tool. It was super successful, but then it would break.
David: Have to pay someone to fix it.
Leo: Then they would change the API, I'd have to pay someone to fix it. I want the new features, I'd have to pay someone to fix it. While I was making money, I was
**The Rise of Vibe Coding**
(6:51) also losing money. Then just about the tail end, I was, "I can't do this anymore because I'm funding this and actually losing money."
David: Vibe coding really came to be. You could always vibe code and Chat-B, right?
Leo: Vibe coding as a vertical and a focus came to be, and I'm, "I could update this myself."
**Coding Independently**
(7:12) I can code this anytime something breaks or changes.
David: Yeah. This, by the way, one of the things, I've got this developer on Upwork for doing the kinds of things you're talking about, and he wrote me the other day. He's, "Did the servers change for your site and my High Caliber site?" I'm,
**Taking Control of Site Edits**
(7:36) "Oh, I forgot to tell you. I had given it to him to go, and then you put on WordPress to make it a little more stable or something." Then I said to him, "Yeah, I forgot to tell him. I want to go and make a bunch of edits myself and start screwing around with it more." So I took it back to Base44, one of these vibe coding platforms.
**Constant Iteration and Ownership**
(8:00) Now I'm constantly iterating myself. Even if WordPress can be more SEO friendly or some of these other benefits, this to me, I get to own this. They're great microsites all the time and things like that. It's fun.
**Sharing Vibe Coding Examples**
(8:22) Leo: It's so much fun.
David: Yeah, it's totally fun. Maybe I could show you.
Leo: You know what?
David: I was just going to say, why don't you share some examples of?
Leo: Yeah. And I can maybe vibe code, but do I know how
**Introducing Jingle My Brand**
(8:38) to use Zoom?
David: That is always one of the biggest tests of your technical acumen here.
Leo: Let's see here. Okay, we got Jingle My Brand. Yes. One of the most successful things I've ever done and one of the most fun things I've ever done. This was built essentially within a weekend. It took a couple of weeks, but
**Jingle My Brand: A Weekend Project**
(9:03) just as far as hours go, in a weekend.
David: And I will preface this and say that I spent eight hours in sessions training marketers. I built 250 slides of material, not including live demos and all this other stuff that I built for this massive training marathon on marketing stuff. This was their favorite thing. It was
**Everyone's Favorite AI Tool**
(9:33) in a way, the least important, right? It wasn't they're going, "Put this on TV."
Leo: And it was everyone's favorite, hands down. So now, David, you can explain it.
David: Yeah. No, of course.
Leo: All right. My wife and I always talked about jingles and how amazing jingles were, and it's so sad that after, I don't know, 1998, it just
**How Jingle My Brand Works**
(9:54) went away. We're, "It'd be cool if brands brought back the jingle." One weekend, I started coding this and developing it. All you got to do is go to the landing page jinglemybrand.com, type in your brand name, type in your product information. I'm just using a pug perfume, pug cafe. Type in all the benefits. Choose your style. You're going to choose 80s 90s pop rock,
**Free Jingle Generation**
(10:19) 80s new wave, classic jingle, the length. What's cool right now, you can generate your lyrics on the fly based on that information. If you want to use this, this is completely free right now. I am funding this all with my own money. Anytime you create something and use it, I am happy to pay for it for you. You don't have to do anything. It's
**David's High Caliber AI Jingle**
(10:38) just free for you to use. Then within a second, you generate music. I already have one loaded up. I loaded up David's here. Here we go. This should work. Ready?
David: Yes. Yeah.
Jingle: Smart and true. High AI here for you. From strategy to making it fly, they'll help your business reach the sky. High
**Create Your Own Jingle**
(11:05) caliber AI. It's the real deal with David Burkowitz. You know how good it will feel. He's an author and exec with a brilliant mind. The other David Burkowitz, leave him behind. High.
David: So yeah, anybody on this call could go right now to Jingle My Brand and create a song. I know Jim and Amy and others are going to be creating
**Jingle Features and Sharing**
(11:34) jingles, totally screwed the rest of their day. They're going to be canceling one and two and three o'clock meetings.
Leo: I love it. The back end for you. My back end looks a little different, but for you, you'll be able to play it. If you share it, it'll extend the song.
**Jingle My Brand: 100% Vibe Coded**
(11:52) David: Now you can embed a video version of it, as I just did during a talk yesterday, mind you.
Leo: Yeah. Guess what? This is 100% vibe coded.
David: Wow.
Leo: From front end to back end.
David: And tell us a little more in the weeds, what platforms did you use for vibe coding?
**The Cost-Effectiveness of Vibe Coding**
(12:11) Leo: Of course. Maybe I'll tell you a little about my philosophy. First of all, when people talk about this is expensive or that's expensive, it is all really inexpensive, right? To build this would have taken thousands of dollars. It wouldn't be possible really, but thousands of dollars and it would have taken months to build. Because I have a young child,
**Why Lovable is a Preferred Platform**
(12:33) a day job, startup, all that stuff, I decided to go with something more Lovable. I actually built this on Lovable. One of the reasons I love Lovable, and it might be a little more expensive than Cursor or even Claude Code, is that it helps me with all the backend stuff. When I put my son to sleep, I could come in
**Lovable's Role in API Connections**
(12:51) here and code, and I don't have to worry so much about the backends. I built this all, but it's not only that. I've looked into other LLMs and APIs that I connect to in order to generate the lyrics, in order to generate the songs themselves. Not everything actually happens within Lovable, but Lovable allows me to connect everything. I would say 98%
**Connecting Other Tools with Lovable**
(13:10) of it all is built within Lovable.
David: Yeah. Because the one point where I was getting a little stuck on things, and I still love Lovable, and we might actually have them present here soon, is
Leo: Yeah.
David: There are other things you typically need to connect once you start doing things a little more sophisticated, like what you're doing.
**Using AI for Meta-Layer Guidance**
(13:37) David: It's just a matter, do you keep it? Yeah. You just, and it's usually stuff that one of the things that's been so helpful for me for all of this stuff is having a running conversation with ChatGPT or Claude and being, "Okay, now I'm trying to do this, how can I do this in Lovable?" having that meta layer.
Leo: Yes.
David: So,
**Cursor vs. Lovable**
(13:59) Leo: I do that quite often. I actually prefer to use something like Cursor.
David: And if I, if it was just up to me, and I'm sorry, not up to me, it is up to.
Leo: But if it was up to my time and resources, I would actually use more Cursor or Claude, and do you use the terminal
**Recommendation: Pick One Platform**
(14:19) interface. I'm more comfortable with code. I'm more comfortable with connecting APIs on the back end, all that good thing. But using both in parallel will save you a lot of time and a lot of money and a lot of energy. So I definitely recommend that. Ultimately, what I recommend is pick one, try to learn it as best as you
**Cost vs. Value Proposition**
(14:35) possibly can, and just go from there. Cursor might be cheaper than Lovable, and by quite a bit if you compare them. But ultimately, this website would have cost thousands of thousands of dollars to build. So if I'm spending even 80% more, I'm okay with that.
David: Yeah. Base44 typically costs a little more than Lovable,
**Platform Cost Trade-offs**
(14:58) but it does more stuff inside it. Claude Code is only 20 bucks a month, and Netlify for instance, for hosting it, is free, or there's a great $9 a month plan. But you have to be a little more tech-savvy to use it and use a platform like Ghosty. So there are definitely all these kinds of tradeoffs, and I actually spent a long time talking
**Scaling and Gemini Guidance**
(15:22) to Gemini, "Okay, I'm not really comfortable doing this. Tell me for something I'm usually dealing with small B2B sites. Jingle My Brand versions of it could go huge. It could be owned by Hallmark.com. So you might need to, hopefully, you need to scale it one day."
Leo: Yeah. Yeah. Oh, that's a whole other
**Digital Sovereignty and "Jailbreaking" Platforms**
(15:43) conversation. Eventually, if Jingle My Brand gets so big, I will want to jailbreak it. I don't personally like that essentially a different corporation or another company is owning my technology or my platform or my IP, in a sense, Lovable.
David: It's behind their gate.
Leo: I'd like to break it out eventually. So while I am personally comfortable
**Building Quickly with Lovable**
(16:08) with coding and things like that, I look at it from a resourcing perspective. I don't want to spend time doing that. I just want to build and build quickly. So I would choose something like Lovable. I recommend the same for everybody too. One thing I'll say too, I'm connecting to different APIs and things like that. Just to create
**Social Lollipop: A Vibe Coded Startup**
(16:26) the song, that's one different API. To download the video, that's another API. This is all being connected in the back end. That's one of the things I love about the other thing is legitimately a startup. I'm a founder of this company called Social Lollipop, and I have a customer. I am building. I am iterating, and I 100% built it from Vibe Coding. So I'll show you some of the
**Free Lead Generation Tool: Safe Zones**
(16:52) tools that I have right now. The first tool, it's completely free, just between us. It's all lead generation right now. It's not the most fancy tool, but I use it as a marketing lead generation thing for people to use for free. If you're in social media, you know that being able to actually see what the safe zones are, where things are being blocked, is difficult. I go ahead
**Social Media Safe Zone Tool**
(17:11) and build something where you could go in here and let me find Mr. Bezos over here, and you can select the platform, mobile, desktop, Facebook, and go through the different types of posts. You'll actually see what the safe zones are, right? Super simple, but this is 100% vibe coded. The other thing that I
**Adberry: Paid Media Ad Insights**
(17:33) have here is something called Adberry, which I absolutely love. If you come in here, you're going to be able to see what paid media ads are being run by companies. You'll go ahead and, Nike, do a search. I'm showing you this because this is 100% vibe coded by me. I am not a developer. I just have ideas, and I know how to use talk sometimes.
**Adberry: API Connections and Report Generation**
(17:57) You're able to search, find your ads, different ads. I love this, generate reports, all that. This is just another example of things that you could build, but by also connecting different APIs. One connects to the Facebook API. Then if I generate a report for this, I am generating a report using ChatGPT or
**Using AI to Connect APIs**
(18:20) Gemini, or something that actually analyzes the text. I am in there typing out, "I want to connect to Facebook, how do I do it? I want to connect to ChatGPT, how do I do it?" And iterating and having conversations with Berky said, either I'm doing it within one of these tools, or I'm actually doing it on the side on Claude or Chach team. One other
**Speech Therapist Tools SAS Platform**
(18:44) thing before I let anybody ask. My wife and I have a business called Speech Therapist Tools. She's a speech language pathologist, and we build different tools that help speech language pathologists, and it's another SAS platform. You come in here, there's a yes no question generator. There's a chronological age generator. There's a tool that allows you to
**Developmental Goal Generator**
(19:04) type in your son's or daughters' or kids' favorite toy. Type in their age. You start playing, and then it'll give you different developmental goals and different developmental things and ideas that you could use specifically for that specific goal, excuse me, specific toy, and for your ultimate child's age. This is all 100% vibe coded.
**SAS Product Traction**
(19:26) Essentially what you've been seeing here are examples of SAS products. I've built SAS products. I ultimately think about how can I build a business around what I'm doing.
David: Do you have any traction on these now? How many users are using this?
Leo: Oh yeah. Speech Therapist Tools is probably the most popular one. We get thousands of visits a month.
**User Growth and Business Model**
(19:45) That's because it's based off of a different property that's existed forever. I'll show you, this is a WordPress site, but it's just a general business. We have a SAS platform that we've attached to it, but it's really resources for speech therapist tools. Social is more of an enterprise medium to size business tool, and I found it a
**Jingle My Brand User Base**
(20:04) few months ago. I have one client and a few free users. I say clients as people paying. When it comes to Jingle My Brand, I have over a hundred users, and I have probably, David, you are one of the top users, but I have at least five people using it every single day.
David: Wow.
Leo: This is a paid product.
**Free Jingle My Brand and Holiday Styles**
(20:22) David: This is free.
Leo: I am giving it away for free. Even during the holidays, David actually helped me create some of this as far as the music styles. Even created a version for creating Christmas music or Hanukkah music. You should still do this, even if the holidays have passed, but I think particularly, I'm not pandering, but David did such a great job
**Favorite Projects and Influences**
(20:43) of creating the Hanukkah styles.
David: I love every single song that comes out of it. It is one of my favorite projects I ever worked on. It's so funny as someone who grew up with all these references from the Beastie Boys and Adam Sandler, and folk music, to be able to contribute to this. It is one of my career highlights.
**Legal Disclaimer and Lovable Pro**
(21:11) David: And for legal purposes, they are in the style of Rob Rack Rob
Leo: Inspired by. Yes.
David: Yes.
Leo: You're not actually getting new Beastie Boy songs. No.
David: So, do you have the pro version of Lovable? 25 bucks a month? Is that
Leo: I pay more than that, but yes.
**Higher Lovable Costs Due to Usage**
(21:31) David: Why do you pay more?
Leo: Yeah. Oh, because I'm using it. I'm using it for all these different platforms. I'm using for different iterations. I'm not voluntarily giving them more money. I'm using a lot more credits.
David: More credits. Okay. That's
Leo: Yeah. Yeah. But as Burky said, I save money by having conversations in
**AI for Efficiency: Claude and ChatGPT**
(21:49) Claude or having conversations in ChatGPT, and then popping it over into Lovable and saying, "This is what I want to do." To reiterate, I love Lovable. Nothing against them, but because I'm a little more technical, I prefer to use Cursor or one of Google's tools or Claude Code. But it's just for time and resources for me. I'm happy to open it up and answer any
**Q&A and Tips**
(22:10) questions or David if you have any other questions, but I thought I'd give some ideas or some tips that I've learned. I love it. Happy to pause.
David: Yeah, let's see. Just other questions, other stuff you've shared so far. I know the URL for sociallollipop is going to make sure to put that in the chat.
Leo: Thank you. And it, yeah, it
**Jingle Length: Feature or Bug?**
(22:34) it is a funny thing that I noticed with Jingle My Brand that Zach just brought up, it tends to make longer jingles than you ask for.
Leo: Yeah. That is a known, whether it's not a bug, it's a feature. I have known about the bug, but no one's complained about it. People have mentioned it to me, but they're,
**Theme Songs vs. Jingles**
(22:56) "Oh, but I still like it." So I've never changed it. But I could easily change that. Jean was mentioning they're maybe more theme songs than jingles. What would you call them?
Leo: Oh, yeah. For sure. I think to make it a jingle, maybe I do have to shorten it and make sure that the time frame works. Yeah. Because I think
**The Fragility of Vibe Coding**
(23:13) the jingles are quick.
David: Yeah. Fit in a 50 or 30-second ad, right?
Leo: Yeah, for sure. Yeah. So, maybe that's what I'll do tonight.
David: There you go.
Leo: The funny thing is, I am sometimes scared to go in something because if you've vibe coded before, you could update one super simple thing, and
**Risk of Breaking Things**
(23:33) it just breaks everything. I'm always, "All right, I'm going to go in there, and I'm going to update this. I need to be okay that if everything else breaks, Zach, I saw you come off mute. Nice seeing your face, by the way."
Zach: I was going to say, to me, a jingle's eight seconds.
Leo: True, true. It's the hook that goes across every
**Voice Selection Feature Request**
(23:53) thing.
Leo: Yeah. I love it. It's your audio brand.
Zach: This is fun. I'm working on a retro brand. The 80s synth-pop is awesome. It doesn't, it's it's, does it let me select the audio voice?
Leo: Not the voice right now. We're
**Value of Feedback**
(24:13) recording this, right? Because this is all super valuable, and I love this feedback.
Leo: Oh yeah. I'm happy to build all these things out.
Zach: Yeah. I mean, as somebody who, from a recommendation from Burkowitz, has been spending a lot more time than probably needed inside of all the sweet things that Gemini's released recently.
**Gemini Voice Limitations**
(24:34) Zach: Yeah. Yeah. Even though they only have a limited number of voices they're letting you use.
Leo: Yeah.
Zach: It's nice to be able to switch between the voices to get it right. So that's just one recommendation.
Leo: Yeah, I love that.
**Pro Version Considerations**
(24:53) Leo: No, I love that. Let me just, because we're all marketers, let me give you my thinking behind this. I could have made this a lot more sophisticated, but I would have created a lot more friction. I am thinking about creating a pro version that I will charge for to recreate the song, remix the song, update the lyrics,
**Simplicity and Friction Reduction**
(25:11) and things like that. But the reason I made it so simple is because if I had to have you fill out more than this, or more Jingle My Brands, or generate your lyrics, you wouldn't do it, or very few people would do it.
Leo: Right now, I was, "Let me get it out there."
David: And then iterate. That's related to if I became a
**Financial Projections for Pro Users**
(25:36) pro user, are you able to run financial projections of how much tokens it uses if I were to become a pro user, right?
Leo: Yes, on my end. So, are you at this stage with this specific application able to pencil out how much money I make or lose you as a pro user?
**Monetization Strategy and Thresholds**
(26:01) Leo: For sure. I know how much I'm losing every time anybody uses it right now. But I'm comfortable. I've set aside a bunch of money saying, "I'm going to have this bucket of money for my own, not personally, it's by my business, but my corporation set out money." I have a little bucket, and I know how much people are charging. So once we hit that threshold, I will then revisit and be
**SAS Profitability and User Acquisition**
(26:23) like, "Am I going to start charging people or not?" But yeah, I need to, and this is the fun thing about, you run a SAS and you run a business, how many users do I need to pay myself back to make this profitable? I'm probably going to be losing money on the first 100 users, but after I get to 101, then I'm going to start
**Jingle My Brand's Potential**
(26:41) making money, and I can continue to grow. Those are all things what I'm talking about. The funny thing about Jingle My Brand is that I have people David, my wife, and other people saying, "You need to build this further. You need to build this further." But I essentially built it as a cool little tool. I'm, "It's definitely not a toy," but I'm, "It's
**Credit Consumption Transparency**
(27:00) just fun. Everyone's loving it. Everyone's having fun. Everyone calls it magic." I'm, "It's just there." But more conversations, I'm, "I need to jump on this more." And David, Zach brings up this other point that still confuses the heck out of me. I often have no idea how many credits I'm burning through, and then it's
**Credit Tracking Challenges**
(27:25) are any of the platforms really good at that?
Leo: No. I would say no. But that's why if you're worried about credits, and I still think this is all inexpensive, but 200 bucks can be a lot of money for someone, right? But ultimately it's 200 bucks versus spending thousands of dollars and months to build something, right? It's
**Firebase for Cost-Effective Building**
(27:47) all relative. If you were worried about money for whatever reason, I would use something like Firebase or even another one of Google's tools. Especially Firebase, it may not be the most powerful. It may not be the best, but it connects to a lot of Google's products in the back end, so things are easier, and sometimes
**Cursor for Comfortable Coders**
(28:08) cheaper. If you are a little more comfortable, use something like Cursor. I used to pay 25 bucks a month for Cursor, and I never hit my limit.
Zach: My question wasn't about, I think my question was to David, is it even possible right now with the way they're doing the tokens, and the way you consume the credits, to actually run a financial model?
**Financial Modeling for AI Agents**
(28:27) David: Oh, that's an interesting question. I look,
Zach: That's more what my push was. I got a buddy who runs a company here, and it's scaled, and they put in a bunch of AI agents to help them automate stuff, and they have no capability of modeling out their
**Is Small-Scale Financial Modeling Possible?**
(28:48) their input for what that translates on a monthly basis to their credits, and then the credits change.
Zach: Yeah. Just curious on the small end if that's even possible. David, do you want me to go?
David: Yeah. Why don't you go?
Leo: I think just anything else, it's a calculated risk. There's always going to be, unless you're fully verticalized,
**Calculated Risk in AI Costs**
(29:09) it's a danger that you're going to have to pay for, right? If I build widgets, and they're plastic widgets, where am I sourcing my plastic from? Are there going to be taxes? There's going to be that. So, I think it's always a bit of a calculated risk. But ultimately, a lot of this, because there's so much competition and the technology is moving so quickly, it is
**Inexpensive AI Generation**
(29:27) relatively inexpensive. I'm sharing this between us, I know it's being recorded, but us friends, if I just charge everybody here $20 a month, I'd be making money. It cost me so little to generate a song and generate a video that I'd be making money right off the bat. It could change tomorrow, right? But then just anybody else, that's how you plan.
**Diversification and Self-Hosting LLMs**
(29:48) I want to diversify. I want to have different tools. If it came to the place where I'm so worried about these tools, I could see about hosting my own LLM somewhere and self-hosted, which has its own other costs and stuff. But if the business grows, maybe that's something I could do with Amazon or Microsoft or something. It's an interesting time in the
**Sustainability of Free AI Tools**
(30:09) market because it is so competitive. At some point, the freebies are going to go away.
Zach: Are some of these things that people are building with vibe coding that are set aside as fun projects, if they did go to scale, is it something that's sustainable? That's, it's just a conversation I've had recently in the last couple weeks
**Digital Sovereignty and Platform Risk**
(30:33) with several people.
Leo: I love this conversation. I think we could talk for hours on it. The one thing I'll say, the sound bite of it all is, this is a big problem with digital sovereignty across the board. Look at Facebook. I come from the social media world where they built this garden. They told everyone they could come build
**The Facebook Garden Analogy**
(30:50) and get millions of followers on their page, and you post for free, and it's all great. Then one day they're, "Nope, we're going to build a gate and a wall, and it's not organic anymore. It's paid." Now you're, "Oh, but we need first-party data." Everyone's, "Oh, I need a website now." There was a trend when people, "Forget about a website. Don't
**The Continuous Risk of Digital Sovereignty**
(31:05) build websites." We live in this world where that is a continuous risk, and something that we all need to think about as users and business people, ultimately digital sovereignty. How does it exist? Does it exist with these big players? But I'll shut up now because I can talk about this forever.
David: You're the guest, so you don't have to
**Vibe Coding Focus and Frameworks**
(31:22) shut up.
Leo: No, no, no. But I want to focus on the vibe coding thing, especially if anybody has any questions, but I'm happy to jump into some tools or thoughts or frameworks that I wanted to share.
David: Yeah, and share away, right?
Leo: Awesome. First and foremost, plan as much as you possibly
**Planning with Whisper and Chat Tools**
(31:40) can. I actually use Whisper. I barely type. I just hit a button, and I talk directly into these tools. I open up a notepad, and I just talk. I'm, "I have an idea to build this tool, and the tool should do this, and I want to have this as the goal, and these are the type of audiences, and these are the type of users." I try to talk and iterate as much as possible in
**Start Ridiculously Small**
(32:00) ChatGPT, or one of the more conversational tools before I even put it into Lovable. So, when I'm feeding it into Lovable, I'm going in as planned and as thoughtful as I possibly can. If this is your first time building anything, I would say start ridiculously small. Build a calculator. Build a simple tool that you need help with, or maybe you don't even need
**Building Confidence with Small Projects**
(32:27) help with, literally an engagement calculator. Build something. The reason for that is I want you to build confidence. I want you to see the magic first before you're, "I'm going to try to build the best SAS in the world, or I'm going to replicate TikTok or something." It's going to break. You're going to fail, and you're going to be left feeling
**Importance of Backend Reporting**
(32:47) demoralized, "This doesn't work. I don't like this." So start very small. Build something ridiculously tiny and super simple, and then build from there so you can see the magic and gain the confidence. The other thing is I'm very big on building the back end. When I say that is reporting. For instance, yeah, Whisper Flow, I use that all day long. I
**Admin Sections and Data Collection**
(33:09) paid for it. I used the trial for two days, then I paid for it. "What am I doing here? Take my money, please." But build the back end. I have a backend admin section for all my tools. I know who's logging in, when they're logging in, when they sign the agreement, the songs that they're creating, things like that. I even make copies of certain things on the back
**Building Tools for Your Tools**
(33:28) end, not necessarily for Jingle My Brand, but I'm talking about across the whole website. Even when you use an email, it's collected in the back end. I could export it, right? So, think about that. The other thing, and this is going to be a huge money and time saver, build tools for your tools. For instance, I constantly build out. If I'm building
**Admin Tools for Content Management**
(33:49) this, let's say I'll give you the example for the Hanukkah one. See all these different styles. I don't constantly prompt Lovable or anything to update the different styles. I built a tool in the back end just for admins that says, "Allow me to build out prompts. Allow me to build out styles. Let me build out time frames," and things
**Gaining Control with Admin Tools**
(34:14) like that. Therefore, all I got to do is log into my website, go to my admin section, and constantly add different styles, and it's 100% for free versus going in there and, "Oh, update this. Now, change that." One thing I hate about that is I actually lose control if I'm just prompting. I don't know what it's building. I don't know what the prompts
**SEO for SAS Platforms**
(34:33) are. I don't know how the API connection looks. I don't know how many people are using it. So, I'm constantly building tools for my tools, and it saves me a ton of credits and a ton of money. The other thing I want to say is, when it comes to marketing and SEO, and I'm talking about SAS platforms more than if you're just building a landing page or some small marketing
**WordPress for Front-End SEO**
(34:58) project. What I do is, if you look at sociallollipop.com, it actually runs off of WordPress. So, I am able to control the website, the way it looks, build the blog, all that without using credits or anything. But if you start using the tools, then you go to app.sociallollipop, and that's when you're in the back end for
**Vibe Coding Tools and SEO**
(35:20) the vibe coding thing, because using WordPress or something else has better SEO. Right now, most of these tools suck at SEO. You could always build tools in the back end. I have for Jingle My Brand, I'm, "Allow me to have a tool on the back end that's for SEO, updating metadata and update." But I find that it's just easier and better to build something in the front
**Getting Started with Vibe Coding**
(35:38) end, and have the back end be built by the tools. Has anybody built anything, or is anybody currently building anything?
Earl: Leo, thank you so much for some of your tips. My question is, I've heard about the vibe coding, and I speak every day with my custom GTP bots on my AI team, but getting started with vibe coding, I'm, "Where do I even start?"
**Lovable.com: Your Starting Point**
(36:02) Literally, I get AI, but I'm, "Why do I even start in the beginning?"
Leo: I would tell you this. Go to lovable.com. Go to any one of these tools, Replit, or Bass, or Firebase. We could drop in the links in here, and it's super simple. Let me show you this again. Here's Lovable. It just says, "Start telling me
**Building a Simple Calculator**
(36:25) about the tool you want to build." Say again, start small. Say, "I want to build a calculator that measures engagement. I put in the interactions. I put the impressions. I put in the reach. And give me the answer." Hit go, and you're going to see the magic. You're going to have a tool that you could use. Especially if you start small, you're going to see something be built.
**Career Prospect Slot Machine Challenge**
(36:44) Earl: Awesome. I've always wanted to build a slot machine for career prospects. This is the industry, this is the company, this is the role. Do a slot machine and see what comes up based on my skill set.
Leo: Having spent enough time in these tools, I bet you can build that within the hour.
Earl: Sweet.
**Digital Marketplace Trend Space**
(37:02) Earl: Challenge accepted.
Leo: Go, go. If you need any help, reach out to me. I'm happy to talk.
David: I want to see it, Earl.
Earl: All right. Please.
David: Gauntlet thrown. I see you, David. Amy had a question about, "Curious if you know, Leo, the best tool to create a digital marketplace trend space where it constantly updates and creates
**Connecting to Trend APIs**
(37:22) charts to bring the insights to life. Can anything do that?"
Leo: Yes. Yeah. The answer is all of them could do it. The hardest part for you, and I could help you with this. I'm happy to go in deeper, but there's not enough time here to do it. It is then connecting to the APIs, right? Making sure that you have the connections to, let's say, TikTok
**Building the Frontend and API Connections**
(37:42) trends API, the YouTube Shorts trends API, and things like that. The front end is going to be pretty easy to build, and this is easy too. I'm talking about degrees, but you build a front end with the marketplace that's going to be Lovable or Replit. Then you need to find the API connections, get your API tools, get your app set up, and then make sure
**Data Scraping Complexity**
(38:01) that you collect all that information and feed it into Lovable or Claude or whatever, and you're going to be able to build that 100%. There are other tools that you could use to scrape data and have agents go out into the web, but that gets ridiculously complex. Essentially, any of these tools could do it.
Amy: Okay, you're definitely going beyond what I was even
**Learning Vibe Coding as a Newbie**
(38:22) imagining.
Leo: So, that's awesome.
Amy: Maybe I'll reach out to you on LinkedIn.
Leo: Yeah, please.
Amy: Okay, thank you.
Leo: Yeah, go ahead. Of course. Happy to do it.
Karen: I'm still learning what vibe coding is, a newbie in this space. I'll go as
**The Addictive Nature of Vibe Coding**
(38:42) Leo: Yeah. The best thing, being able to talk to one of these apps and have it build a site for you. It's, yeah. This thing, once you start doing it, it can get so addictive.
Karen: Yeah. It blows my mind. It really does. I'm going to keep going.
**Vibe Coding for Landing Pages**
(39:05) David: I literally do not give a talk anymore. I was prepping a talk that I'm giving tomorrow. Before I worked on the slides, I vibe coded the landing page. I was kind of more excited about the vibe coding than the slides. So yeah, but now it's totally messing up the order. I'm, "I can't not have that."
Leo: You build something.
**Building Tools for Day-to-Day Tasks**
(39:28) Leo: Yeah, I do the same if I need something at my day-to-day job. I'm, "Oh, this is hard to connect these two data points or mixing this." I'm, "How can I put it into whatever tool they give me to use and build it out?" One thing I also want to say with the short amount of time we have, Karen, you asked a good question. I'm sorry if I'm mispronouncing your name, but you asked a good question, how many people are actually using
**Marketing and Sales, Not Just Building**
(39:45) these things, right? Here's the thing with all these things, and I believe this across all AI tools. We're in the marketing game. We're in the sales game. This isn't necessarily on the building game. If you go right now, "Is there a tool that does this?" There probably is a tool. There's a tool that builds that you could probably build it.
**Winning in the AI Tool Market**
(40:07) It's the people that can get in front of clients, get social media attraction, get growth. Those are the people that are going to win because ultimately, if you want a tool that looks at social media analytics, I bet you if you go to any one of these directories, you're going to find 50, right? We're still in a weird place where can't ChatGPT just do
**Users Want Tools, Not Chatbots**
(40:29) this for you, right? If you're building a certain tool, a calculator, can I just go to ChatGPT? One thing I found, especially when it comes to speech language pathologists, they don't want the code start, "Here's a chat, ask it anything." They actually still want a tool. They just want a hammer, right? They don't want the Swiss Army knife that they have to
**Reverting to Simple Tools for User Engagement**
(40:47) figure out how to open and do that. People still want to say, "What's the age of this person?" and put in the information. I learned this the hard way. I had a bunch of tools built, and I'm, "I could just create a chatbot that does all this. Just ask it what the age is, and have it create a lesson plan." The user base went down. I reverted back to these
**Vibe Coding: Web Pages vs. Apps**
(41:09) simple little tools, and I got my traffic back up.
Earl: So for vibe coding, is it only really building web pages or things on web pages, or there other extensions of it?
Leo: Oh, that's super. No, I love this. Lovable and Replit and Firebase are a little more just for front end, SAS, server-side applications, right? But Cursor,
**Building Mac and Mobile Apps**
(41:39) Claude Code, and a lot of these other tools, you could build apps for your Mac, build mobile apps, and things like that. Yeah, this does get harder, right? Because then you need to understand how to install certain backend code stuff, Java on your Mac, and have it connect. A lot of times you're not seeing what you're building until it deploys. That gets harder, but
**The Challenge of Non-Web Interfaces**
(42:02) it's definitely possible. I've built some things on my backends, just as a Mac app that I wanted
Earl: Because I, you can just make web page after web page, but that's not necessarily how consumers are interfacing. They still are, there's all the rest in it.
Leo: One of the bigger problems that a lot of
**App Conversion for Vibe Coded Tools**
(42:21) these tools have right now, but they'll probably fix in the next few months, Lovable and some of these other tools, is that you can't actually create an app for it, or software for your Mac or Windows or something that's jailed within a website. I would love for Jingle My Brand to be an app on your phone, and it can be. There are tools
**Connecting Payment Systems**
(42:41) that you do a search, you'll find tools that take your vibe-coded app and make it into an Android app. You'll find those things, but it's not as easy as you would think.
David: Cool stuff that folks are sharing. Christopher, yeah.
Leo: So, what's funny is on Jingle My Brand, I already connected the payment system and all
**Stripe Integration Complexity**
(43:00) that, but it's super simple. A lot of times, Lovable itself will prompt you, "You need to create a Stripe account, and Stripe account needs the code." Then I'm, "Wait, where do I connect? Where do I go into the chat?" I'm, "Where do I get the code for Stripe?" Then it gives it to me, and then I put it into Lovable. So yeah, that gets super complex. The other
**Hitting Technical Walls**
(43:20) thing I'll say is for Social Lollipop, I've built the tools you see in some of the other things. There's going to be a time when I'm going to hit a wall, or I've hit walls before, where I'm, "I want to build this," and it's super complex. It doesn't exist out there. I need someone with a technical mind to go in there and
**Needing Technical Expertise**
(43:39) start coding and connecting things. So you will hit a wall as you start scaling, or I waste a lot of time asking questions that someone who knew better would have asked a better question. Christopher, have you built anything or are you building anything right now?
Christopher: Oh, he has. Yeah, he's been sharing some cool stuff.
**Wish List and Community Sharing**
(43:56) Leo: Oh, nice. I'd love to see it. Yeah, I'll make sure you get the chat too so you can see any of the links because we always have the funny, fun added conversation. Yeah. What other things is people trying to understand, or what's on your wish list?
Christopher: Yeah. I see Christopher. Lovable can get you a
**Prototypes for Technical Collaboration**
(44:26) prototype, which can help you sell the idea to a technical person, help you get to the finish line.
Leo: Yeah, that's exactly where I am with Social Lollipop. I have been going to events, startup events, startup pitches, and presenting it.
David: And this is something where, yeah, I also, for those of us who maybe have clients with their
**Explaining Ideas to Dev Teams**
(44:48) own development teams and things like that, where we're not going to ship a finished product to it. Then I've worked with others where I'm, "You can get this so much of the way there, and tell them the kinds of things you're looking for, instead of you trying to explain, 'Okay, I want this kind of site, and it looks Airbnb, but for booking
**Creating Visuals for Design Concepts**
(45:14) restaurants, and it should be this kind of thing, but it should be a much more old-fashioned design because these are more old-fashioned trends,' and trying to do all this stuff." You create something yourself that's so much more further along than I could never use Figma or all these tools. I'd spend forever, even in Canva,
**The Last Mile: Designers and Developers**
(45:35) trying to make something look right.
Leo: And it's, great. Then when they have a designer and developer, all that you can work with, they'll figure out that last mile and make it way better than you can imagine. Yes. Yeah. The other thing I think there's a pitfall, and this is we saw this, David, when we were in the SAS world,
**The Dangers of Self-Building**
(45:54) or even the agency world. Sometimes building stuff yourself could be dangerous. Things break. Things need to be updated. All that, unless you have someone dedicated to be looking at all that and changing it, especially if you're building something that's client-facing. If Jingle My Brand is low stakes, but if it broke at 3:00 in the
**Resource Planning for Tools**
(46:12) morning, who's fixing it? Who's updating it? Who's looking to see if it's done, right? If my business is marketing, and my business is, I don't know, selling widgets. I'm focused on the widgets, not necessarily on this one little tool that's a lead gen tool or something. So be careful to build things without properly resourcing it or properly understanding how much
**The Replet Trap: Unused Tools**
(46:30) time it's going to take.
Zach: Hey, thanks. This is great. I built a couple tools in Replit. The problem that I find is that I'll build something that has a limited use case for me, and then I'm just stuck paying monthly fees on something, I'm, "Oh, maybe I'll use it on this next thing or whatever." But it's
**Subscription Overload**
(46:52) just, it's a trap I find. Then I've got all these now. I'm looking at Claude Code, and I've used Lovable for sure, but I just end up with all these different apps, and with one thing or two things that I used for. How do you, what are your thoughts on that?
Leo: I'm going to give you more of a
**A Bigger Picture: Marketing Costs**
(47:15) bigger picture look at things. I try as a marketer, and money is money. Everyone has different financial situations from a business to a personal perspective. I get it, and I'm not minimizing that or doing anything to ignore that. So, there's that, but let's skip it for a second. I always say, "Look at how much marketing cost in the 90s." They cost so much money, and you
**Digital Marketing: Inexpensive Targeting**
(47:36) didn't even know how well the targeting was going, and where it was going, and the resources it took to build it was incredible. Now there's social media, and there's digital marketing, which is stupid inexpensive, right? Even as it gets more expensive, it's still really inexpensive. If I took a time machine, and I went back to the 90s or 80s, and I said, "There's a tool that
**The Value of $20 a Month**
(47:56) you could use to target almost anybody you want, and it's relatively inexpensive." You'd be, "Here's all my money. Forget about that billboard. Forget about the TV. Take all my money." Right? I'm, "This exists, and it's social media." So I look at that in the same world as the question you proposed. So this is costing me 20 bucks a month, $25 a month. You got to ask
**Assessing Tool Value**
(48:17) yourself, "Are these individual tools that I built actually worth that much? Is the value there for myself?" But ultimately, it's so inexpensive for you to have built that tool. Would it cost you thousands, maybe hundreds of thousands of dollars, right? Or taking you so much time. If it breaks, it's going to cost you hundreds to fix it. So, would I pay $20 a month to have
**Cost vs. Efficiency**
(48:37) a simple calculator that saves me two hours a day, two hours a month? I'd say yes. So, you got to look at, what have you built? What's the value it's providing, which is time, money, efficiency, and is that worth what you're spending? If you look at it from that calculator or that algorithm, that will answer your question.
Zach: But you haven't found there isn't
**The "Jailbreak" Dilemma**
(48:57) a way that you're, "Okay, this tool's done." I can't jailbreak it. We're probably never going to use it again, right? That's where I end up with on a lot of these things, right?
Leo: Yeah. I've run into that. Ideas I've had, I've built, or I've hit walls, and I can't build it any further. But because I'm building multiple tools,
**Managing Multiple Tools and Credits**
(49:16) and multiple things within the same platform, I say, "All right, that one's dead." But I'll use the credits for Social Lollipop or Jingle My Brand. So, I'm always building something. That's my own unique use case.
Zach: So, are you still just maintaining your subscriptions across a bunch of these things, and figure you're
**ChatGPT as an Affordable Partner**
(49:33) okay.
Leo: Yeah. Yeah. Exactly. But again, I look at it from the frame, "Shoot, ChatGPT is a partner that helps me think. Even just that's cheaper than a VA. I'm paying 25 bucks a month. That's free." Not everyone has $25 a month to spend, but what?
Zach: It's just, it's very easy to end up with
**The Challenge of Too Many Subscriptions**
(49:51) 50 tools that you're paying 30 bucks.
Leo: Yes. Yes. Yeah. Totally.
Zach: I, that's the challenge that I wrestle with a bit. It's, "Oh, I did this cool thing. Am I using it a lot? I don't know. I feel I'm going to use it."
Leo: There's only so many hours in the day. That's the, they're all fun. They're
**API Errors and Time Investment**
(50:07) just as you said. Some of it was headache for me building the app in Replit. I tried to do a few apps that just weren't possible. It just couldn't.
Zach: I just kept running into API errors and stuff, and I was finally, "All right, I've spent 25 hours trying to build this thing. How many more hours am I going to spend?" Right?
**Voice Cloner Success Story**
(50:24) Leo: Amen. Amen.
Zach: But I did build some simpler things that, to your point earlier, they're the calculator. I built this voice cloner that would take my clients' voices through Eleven Labs since it's a huge pain to go directly through there and do that. I was able to do that for this big video series that we did for them, and they weren't able to
**Subscription Philosophy**
(50:44) come back to set. So it was able to recreate that, and that was worth it for sure.
Leo: Yeah. No, I guess my philosophy, I look at it as this is just a problem with anything. I have a subscription to Wired. I have a subscription to Men's Health. I have a
**Valuing Subscriptions**
(51:02) subscription to this. I'm, "Just the other day I went through, I'm, 'I don't even read this. I don't have time to read that. Why am I paying for this?'" But I focus on the ones that are providing value. But I would love to subscribe and read all these things, right? Same thing for AI tools. If it's worth it for you, cool. If you
**Conclusion and Appreciation**
(51:17) could afford it, cool.
Leo: But if not, get rid of it.
David: Well, Leo, this is always better than expected. Appreciate you coming in on the fly and
Leo: Sharing all this stuff with us.
David: And in the community. Yeah. It's fun to get to see a lot of the V0.1 versions well before they launch. Really honored for
**Upcoming Speaker and Community Engagement**
(51:45) that. Thanks everyone for showing up. We got Peter Shankman next week talking about AI and comms, and he's also a dynamic speaker. You're in for a treat. Great to see everyone, and look forward to seeing what all you create with all these new things.
Leo: Yeah, please reach out to me. Happy to look in. Thanks for having me on all the
**Farewell and Future Invitation**
(52:08) great questions.
David: Yeah. You, I think you'll be welcome back here, so we know where to find you.
Leo: Awesome. Thank you.
(0:05) Hey everyone, welcome to another edition of AI Insiders with AI Marketers Guild by March Media. I am David Burkowitz, your usually friendly host. We've got a much more friendly guest today, Leo Morejon, a multiple-time colleague, multiple-time collaborator. He was with me from the start. He actually built the first Ad Markers Guild website.
**Leo Morejon's Entrepreneurial Journey**
(0:32) He's been with us in prehistoric days, and Leo has been more and more entrepreneurial, always doing cool stuff beyond these amazing hats he's worn on the agency and brand side. He can tell you all kinds of things about winning Guinness World Records for cookie brands and crazy stuff.
**Focusing on AI and Leo's Background**
(1:00) Today we'll focus a little more on the AI front and the stuff that Leo is able to ideate and come up with and market fast. It's cool. So why don't you share a little about who you are and anything I didn't quite cover yet?
No, of course. I appreciate being on. Thank you so much for having me. I appreciate everyone joining to hear me
**Leo's Career Philosophy: Gold vs. Shovel**
(1:23) speak and hopefully have a really fruitful educational conversation that really supports you and your work and everything you're doing. Berky, thank you for the kind words and for having me. I look at my career in two different ways: either I am looking for the gold or selling the shovel. I started in big ad companies from WPP, JWT to 360i at Densu.
**Career Shift: From Hunting Gold to Selling the Shovel**
(1:46) I look at that, I'm looking for the gold. I'm out there hunting. I'm out there digging. I'm out there shuffling, looking through what can I build and do that is amazing and impactful. Then I went into the technology world after that, still within the realm of social media and marketing. In the MarTech world, I look at that as selling
**Enabling Marketers: Entrepreneurship and Hormel**
(2:03) the shovel. I am giving the tools to other marketers so they can go and find the gold. From there, my experience goes into entrepreneurship to working at big brands. My day job currently is at Hormel, the makers of Spam, Applegate, Justin Peanut Butter. I'm running all social media, influencer marketing stuff. As Berkie mentioned, I
**The Best Time to Be a Builder**
(2:26) am always building. I'm always learning. I'm always testing. The last few months, I've shipped faster and built some of the coolest things and most successful things that I ever have in my whole life. I think right now is the best time to be alive for someone who's a creative, for someone who's a builder, or any marketer. If you're not getting up and
**Embracing AI Possibilities**
(2:48) getting excited every single day, maybe a little scared, you're really not testing and looking at all the possibilities that are out there. David: Why don't you dive in because we could ask why you're not so scared about AI taking your job, but I feel we all have our reasons and some of us might
**The AI "Aha!" Moment**
(3:10) even want AI to take some of our jobs. The stuff you are building, what's great with it is, was there even taking one step back, was there an aha moment for you where you're, "This is fun now, or I can do things I couldn't do before now?"
Leo: I remember it very distinctly. It was about my son's about to be 3 years
**Pre-ChatGPT Discoveries**
(3:38) old. It was right before ChatGPT came out. There was the playground so you could still play with GPTs. I remember going in there and building simple calculators or word replacement tools, and if this then that, just for text. I was profoundly amazed and felt I discovered gold. The only other time I ever felt that was
**Social Media and AI: World Changers**
(4:03) when I discovered social media in one way, before it was called social media, just doing some stuff on MySpace. I'm, "This is going to change the world." I saw the future. I'm, "This could do marketing. This could do coding. This could do essentially anything." I still remember that, and it coincides so well with my
**Personal Connection to AI's Emergence**
(4:23) son being born. There's a lot of positivity in my life looking at all this, and just being able to play with it is incredible.
David: So he's around the same age as ChatGPT, huh?
Leo: Yeah. Yeah.
David: That's super fun and just something that sticks out in my memory. But yeah, I remember working, and then
**Chat GPT's Accessibility Shift**
(4:42) when ChatGPT came out where you could actually talk to it, that was also incredible. This is going to change and make it a lot more accessible to everybody else, not just this playground that looks almost not a terminal, but just a little form that you work in.
David: And then you mentioned this new
**The Vibe Coding Phase Shift**
(5:02) phase shift or something to that effect happening. You said it several months ago, and what happened then? Why? What did it for you?
Leo: I think all these vibe coding platforms from Cursor, Lovable to Replit, to anything else that's out there. I've always built things, but I am not a coder. When I say I've built things, I probably should say more like I've
**Producing Ideas vs. Coding**
(5:23) produced things. I've always had a lot of ideas. I've hired. I've partnered with technical people. I've gone out there and tried to put scripts together, but it would take so long. It'd be so costly. The resources would take up so much energy and time that things would ultimately fail. Even if I had successful things, I actually had one of the first successful-ish
**"Prompt to Product" Tool**
(5:49) products when it comes to combining ChatGPT with some kind of marketing thing. It was called Prompt to Product. What I did was I allowed you to get a WordPress setup plugin where you could write a prompt but then turn that prompt into a feature. We actually used it for one of your earlier websites, Berky or David, we
**Translating Resumes and Recipes**
(6:10) looked at your job, right? Or your resume, and then we translated it to be more marketing. I had built one that is put in a recipe and then make it gluten-free or make it vegan or whatever. This is just on the back end, a few forms that you would fill out and say, "If someone inputs a recipe, make it
**Maintenance Challenges**
(6:30) gluten-free." You could put it into any WordPress site and make it a feature or a tool. It was super successful, but then it would break.
David: Have to pay someone to fix it.
Leo: Then they would change the API, I'd have to pay someone to fix it. I want the new features, I'd have to pay someone to fix it. While I was making money, I was
**The Rise of Vibe Coding**
(6:51) also losing money. Then just about the tail end, I was, "I can't do this anymore because I'm funding this and actually losing money."
David: Vibe coding really came to be. You could always vibe code and Chat-B, right?
Leo: Vibe coding as a vertical and a focus came to be, and I'm, "I could update this myself."
**Coding Independently**
(7:12) I can code this anytime something breaks or changes.
David: Yeah. This, by the way, one of the things, I've got this developer on Upwork for doing the kinds of things you're talking about, and he wrote me the other day. He's, "Did the servers change for your site and my High Caliber site?" I'm,
**Taking Control of Site Edits**
(7:36) "Oh, I forgot to tell you. I had given it to him to go, and then you put on WordPress to make it a little more stable or something." Then I said to him, "Yeah, I forgot to tell him. I want to go and make a bunch of edits myself and start screwing around with it more." So I took it back to Base44, one of these vibe coding platforms.
**Constant Iteration and Ownership**
(8:00) Now I'm constantly iterating myself. Even if WordPress can be more SEO friendly or some of these other benefits, this to me, I get to own this. They're great microsites all the time and things like that. It's fun.
**Sharing Vibe Coding Examples**
(8:22) Leo: It's so much fun.
David: Yeah, it's totally fun. Maybe I could show you.
Leo: You know what?
David: I was just going to say, why don't you share some examples of?
Leo: Yeah. And I can maybe vibe code, but do I know how
**Introducing Jingle My Brand**
(8:38) to use Zoom?
David: That is always one of the biggest tests of your technical acumen here.
Leo: Let's see here. Okay, we got Jingle My Brand. Yes. One of the most successful things I've ever done and one of the most fun things I've ever done. This was built essentially within a weekend. It took a couple of weeks, but
**Jingle My Brand: A Weekend Project**
(9:03) just as far as hours go, in a weekend.
David: And I will preface this and say that I spent eight hours in sessions training marketers. I built 250 slides of material, not including live demos and all this other stuff that I built for this massive training marathon on marketing stuff. This was their favorite thing. It was
**Everyone's Favorite AI Tool**
(9:33) in a way, the least important, right? It wasn't they're going, "Put this on TV."
Leo: And it was everyone's favorite, hands down. So now, David, you can explain it.
David: Yeah. No, of course.
Leo: All right. My wife and I always talked about jingles and how amazing jingles were, and it's so sad that after, I don't know, 1998, it just
**How Jingle My Brand Works**
(9:54) went away. We're, "It'd be cool if brands brought back the jingle." One weekend, I started coding this and developing it. All you got to do is go to the landing page jinglemybrand.com, type in your brand name, type in your product information. I'm just using a pug perfume, pug cafe. Type in all the benefits. Choose your style. You're going to choose 80s 90s pop rock,
**Free Jingle Generation**
(10:19) 80s new wave, classic jingle, the length. What's cool right now, you can generate your lyrics on the fly based on that information. If you want to use this, this is completely free right now. I am funding this all with my own money. Anytime you create something and use it, I am happy to pay for it for you. You don't have to do anything. It's
**David's High Caliber AI Jingle**
(10:38) just free for you to use. Then within a second, you generate music. I already have one loaded up. I loaded up David's here. Here we go. This should work. Ready?
David: Yes. Yeah.
Jingle: Smart and true. High AI here for you. From strategy to making it fly, they'll help your business reach the sky. High
**Create Your Own Jingle**
(11:05) caliber AI. It's the real deal with David Burkowitz. You know how good it will feel. He's an author and exec with a brilliant mind. The other David Burkowitz, leave him behind. High.
David: So yeah, anybody on this call could go right now to Jingle My Brand and create a song. I know Jim and Amy and others are going to be creating
**Jingle Features and Sharing**
(11:34) jingles, totally screwed the rest of their day. They're going to be canceling one and two and three o'clock meetings.
Leo: I love it. The back end for you. My back end looks a little different, but for you, you'll be able to play it. If you share it, it'll extend the song.
**Jingle My Brand: 100% Vibe Coded**
(11:52) David: Now you can embed a video version of it, as I just did during a talk yesterday, mind you.
Leo: Yeah. Guess what? This is 100% vibe coded.
David: Wow.
Leo: From front end to back end.
David: And tell us a little more in the weeds, what platforms did you use for vibe coding?
**The Cost-Effectiveness of Vibe Coding**
(12:11) Leo: Of course. Maybe I'll tell you a little about my philosophy. First of all, when people talk about this is expensive or that's expensive, it is all really inexpensive, right? To build this would have taken thousands of dollars. It wouldn't be possible really, but thousands of dollars and it would have taken months to build. Because I have a young child,
**Why Lovable is a Preferred Platform**
(12:33) a day job, startup, all that stuff, I decided to go with something more Lovable. I actually built this on Lovable. One of the reasons I love Lovable, and it might be a little more expensive than Cursor or even Claude Code, is that it helps me with all the backend stuff. When I put my son to sleep, I could come in
**Lovable's Role in API Connections**
(12:51) here and code, and I don't have to worry so much about the backends. I built this all, but it's not only that. I've looked into other LLMs and APIs that I connect to in order to generate the lyrics, in order to generate the songs themselves. Not everything actually happens within Lovable, but Lovable allows me to connect everything. I would say 98%
**Connecting Other Tools with Lovable**
(13:10) of it all is built within Lovable.
David: Yeah. Because the one point where I was getting a little stuck on things, and I still love Lovable, and we might actually have them present here soon, is
Leo: Yeah.
David: There are other things you typically need to connect once you start doing things a little more sophisticated, like what you're doing.
**Using AI for Meta-Layer Guidance**
(13:37) David: It's just a matter, do you keep it? Yeah. You just, and it's usually stuff that one of the things that's been so helpful for me for all of this stuff is having a running conversation with ChatGPT or Claude and being, "Okay, now I'm trying to do this, how can I do this in Lovable?" having that meta layer.
Leo: Yes.
David: So,
**Cursor vs. Lovable**
(13:59) Leo: I do that quite often. I actually prefer to use something like Cursor.
David: And if I, if it was just up to me, and I'm sorry, not up to me, it is up to.
Leo: But if it was up to my time and resources, I would actually use more Cursor or Claude, and do you use the terminal
**Recommendation: Pick One Platform**
(14:19) interface. I'm more comfortable with code. I'm more comfortable with connecting APIs on the back end, all that good thing. But using both in parallel will save you a lot of time and a lot of money and a lot of energy. So I definitely recommend that. Ultimately, what I recommend is pick one, try to learn it as best as you
**Cost vs. Value Proposition**
(14:35) possibly can, and just go from there. Cursor might be cheaper than Lovable, and by quite a bit if you compare them. But ultimately, this website would have cost thousands of thousands of dollars to build. So if I'm spending even 80% more, I'm okay with that.
David: Yeah. Base44 typically costs a little more than Lovable,
**Platform Cost Trade-offs**
(14:58) but it does more stuff inside it. Claude Code is only 20 bucks a month, and Netlify for instance, for hosting it, is free, or there's a great $9 a month plan. But you have to be a little more tech-savvy to use it and use a platform like Ghosty. So there are definitely all these kinds of tradeoffs, and I actually spent a long time talking
**Scaling and Gemini Guidance**
(15:22) to Gemini, "Okay, I'm not really comfortable doing this. Tell me for something I'm usually dealing with small B2B sites. Jingle My Brand versions of it could go huge. It could be owned by Hallmark.com. So you might need to, hopefully, you need to scale it one day."
Leo: Yeah. Yeah. Oh, that's a whole other
**Digital Sovereignty and "Jailbreaking" Platforms**
(15:43) conversation. Eventually, if Jingle My Brand gets so big, I will want to jailbreak it. I don't personally like that essentially a different corporation or another company is owning my technology or my platform or my IP, in a sense, Lovable.
David: It's behind their gate.
Leo: I'd like to break it out eventually. So while I am personally comfortable
**Building Quickly with Lovable**
(16:08) with coding and things like that, I look at it from a resourcing perspective. I don't want to spend time doing that. I just want to build and build quickly. So I would choose something like Lovable. I recommend the same for everybody too. One thing I'll say too, I'm connecting to different APIs and things like that. Just to create
**Social Lollipop: A Vibe Coded Startup**
(16:26) the song, that's one different API. To download the video, that's another API. This is all being connected in the back end. That's one of the things I love about the other thing is legitimately a startup. I'm a founder of this company called Social Lollipop, and I have a customer. I am building. I am iterating, and I 100% built it from Vibe Coding. So I'll show you some of the
**Free Lead Generation Tool: Safe Zones**
(16:52) tools that I have right now. The first tool, it's completely free, just between us. It's all lead generation right now. It's not the most fancy tool, but I use it as a marketing lead generation thing for people to use for free. If you're in social media, you know that being able to actually see what the safe zones are, where things are being blocked, is difficult. I go ahead
**Social Media Safe Zone Tool**
(17:11) and build something where you could go in here and let me find Mr. Bezos over here, and you can select the platform, mobile, desktop, Facebook, and go through the different types of posts. You'll actually see what the safe zones are, right? Super simple, but this is 100% vibe coded. The other thing that I
**Adberry: Paid Media Ad Insights**
(17:33) have here is something called Adberry, which I absolutely love. If you come in here, you're going to be able to see what paid media ads are being run by companies. You'll go ahead and, Nike, do a search. I'm showing you this because this is 100% vibe coded by me. I am not a developer. I just have ideas, and I know how to use talk sometimes.
**Adberry: API Connections and Report Generation**
(17:57) You're able to search, find your ads, different ads. I love this, generate reports, all that. This is just another example of things that you could build, but by also connecting different APIs. One connects to the Facebook API. Then if I generate a report for this, I am generating a report using ChatGPT or
**Using AI to Connect APIs**
(18:20) Gemini, or something that actually analyzes the text. I am in there typing out, "I want to connect to Facebook, how do I do it? I want to connect to ChatGPT, how do I do it?" And iterating and having conversations with Berky said, either I'm doing it within one of these tools, or I'm actually doing it on the side on Claude or Chach team. One other
**Speech Therapist Tools SAS Platform**
(18:44) thing before I let anybody ask. My wife and I have a business called Speech Therapist Tools. She's a speech language pathologist, and we build different tools that help speech language pathologists, and it's another SAS platform. You come in here, there's a yes no question generator. There's a chronological age generator. There's a tool that allows you to
**Developmental Goal Generator**
(19:04) type in your son's or daughters' or kids' favorite toy. Type in their age. You start playing, and then it'll give you different developmental goals and different developmental things and ideas that you could use specifically for that specific goal, excuse me, specific toy, and for your ultimate child's age. This is all 100% vibe coded.
**SAS Product Traction**
(19:26) Essentially what you've been seeing here are examples of SAS products. I've built SAS products. I ultimately think about how can I build a business around what I'm doing.
David: Do you have any traction on these now? How many users are using this?
Leo: Oh yeah. Speech Therapist Tools is probably the most popular one. We get thousands of visits a month.
**User Growth and Business Model**
(19:45) That's because it's based off of a different property that's existed forever. I'll show you, this is a WordPress site, but it's just a general business. We have a SAS platform that we've attached to it, but it's really resources for speech therapist tools. Social is more of an enterprise medium to size business tool, and I found it a
**Jingle My Brand User Base**
(20:04) few months ago. I have one client and a few free users. I say clients as people paying. When it comes to Jingle My Brand, I have over a hundred users, and I have probably, David, you are one of the top users, but I have at least five people using it every single day.
David: Wow.
Leo: This is a paid product.
**Free Jingle My Brand and Holiday Styles**
(20:22) David: This is free.
Leo: I am giving it away for free. Even during the holidays, David actually helped me create some of this as far as the music styles. Even created a version for creating Christmas music or Hanukkah music. You should still do this, even if the holidays have passed, but I think particularly, I'm not pandering, but David did such a great job
**Favorite Projects and Influences**
(20:43) of creating the Hanukkah styles.
David: I love every single song that comes out of it. It is one of my favorite projects I ever worked on. It's so funny as someone who grew up with all these references from the Beastie Boys and Adam Sandler, and folk music, to be able to contribute to this. It is one of my career highlights.
**Legal Disclaimer and Lovable Pro**
(21:11) David: And for legal purposes, they are in the style of Rob Rack Rob
Leo: Inspired by. Yes.
David: Yes.
Leo: You're not actually getting new Beastie Boy songs. No.
David: So, do you have the pro version of Lovable? 25 bucks a month? Is that
Leo: I pay more than that, but yes.
**Higher Lovable Costs Due to Usage**
(21:31) David: Why do you pay more?
Leo: Yeah. Oh, because I'm using it. I'm using it for all these different platforms. I'm using for different iterations. I'm not voluntarily giving them more money. I'm using a lot more credits.
David: More credits. Okay. That's
Leo: Yeah. Yeah. But as Burky said, I save money by having conversations in
**AI for Efficiency: Claude and ChatGPT**
(21:49) Claude or having conversations in ChatGPT, and then popping it over into Lovable and saying, "This is what I want to do." To reiterate, I love Lovable. Nothing against them, but because I'm a little more technical, I prefer to use Cursor or one of Google's tools or Claude Code. But it's just for time and resources for me. I'm happy to open it up and answer any
**Q&A and Tips**
(22:10) questions or David if you have any other questions, but I thought I'd give some ideas or some tips that I've learned. I love it. Happy to pause.
David: Yeah, let's see. Just other questions, other stuff you've shared so far. I know the URL for sociallollipop is going to make sure to put that in the chat.
Leo: Thank you. And it, yeah, it
**Jingle Length: Feature or Bug?**
(22:34) it is a funny thing that I noticed with Jingle My Brand that Zach just brought up, it tends to make longer jingles than you ask for.
Leo: Yeah. That is a known, whether it's not a bug, it's a feature. I have known about the bug, but no one's complained about it. People have mentioned it to me, but they're,
**Theme Songs vs. Jingles**
(22:56) "Oh, but I still like it." So I've never changed it. But I could easily change that. Jean was mentioning they're maybe more theme songs than jingles. What would you call them?
Leo: Oh, yeah. For sure. I think to make it a jingle, maybe I do have to shorten it and make sure that the time frame works. Yeah. Because I think
**The Fragility of Vibe Coding**
(23:13) the jingles are quick.
David: Yeah. Fit in a 50 or 30-second ad, right?
Leo: Yeah, for sure. Yeah. So, maybe that's what I'll do tonight.
David: There you go.
Leo: The funny thing is, I am sometimes scared to go in something because if you've vibe coded before, you could update one super simple thing, and
**Risk of Breaking Things**
(23:33) it just breaks everything. I'm always, "All right, I'm going to go in there, and I'm going to update this. I need to be okay that if everything else breaks, Zach, I saw you come off mute. Nice seeing your face, by the way."
Zach: I was going to say, to me, a jingle's eight seconds.
Leo: True, true. It's the hook that goes across every
**Voice Selection Feature Request**
(23:53) thing.
Leo: Yeah. I love it. It's your audio brand.
Zach: This is fun. I'm working on a retro brand. The 80s synth-pop is awesome. It doesn't, it's it's, does it let me select the audio voice?
Leo: Not the voice right now. We're
**Value of Feedback**
(24:13) recording this, right? Because this is all super valuable, and I love this feedback.
Leo: Oh yeah. I'm happy to build all these things out.
Zach: Yeah. I mean, as somebody who, from a recommendation from Burkowitz, has been spending a lot more time than probably needed inside of all the sweet things that Gemini's released recently.
**Gemini Voice Limitations**
(24:34) Zach: Yeah. Yeah. Even though they only have a limited number of voices they're letting you use.
Leo: Yeah.
Zach: It's nice to be able to switch between the voices to get it right. So that's just one recommendation.
Leo: Yeah, I love that.
**Pro Version Considerations**
(24:53) Leo: No, I love that. Let me just, because we're all marketers, let me give you my thinking behind this. I could have made this a lot more sophisticated, but I would have created a lot more friction. I am thinking about creating a pro version that I will charge for to recreate the song, remix the song, update the lyrics,
**Simplicity and Friction Reduction**
(25:11) and things like that. But the reason I made it so simple is because if I had to have you fill out more than this, or more Jingle My Brands, or generate your lyrics, you wouldn't do it, or very few people would do it.
Leo: Right now, I was, "Let me get it out there."
David: And then iterate. That's related to if I became a
**Financial Projections for Pro Users**
(25:36) pro user, are you able to run financial projections of how much tokens it uses if I were to become a pro user, right?
Leo: Yes, on my end. So, are you at this stage with this specific application able to pencil out how much money I make or lose you as a pro user?
**Monetization Strategy and Thresholds**
(26:01) Leo: For sure. I know how much I'm losing every time anybody uses it right now. But I'm comfortable. I've set aside a bunch of money saying, "I'm going to have this bucket of money for my own, not personally, it's by my business, but my corporation set out money." I have a little bucket, and I know how much people are charging. So once we hit that threshold, I will then revisit and be
**SAS Profitability and User Acquisition**
(26:23) like, "Am I going to start charging people or not?" But yeah, I need to, and this is the fun thing about, you run a SAS and you run a business, how many users do I need to pay myself back to make this profitable? I'm probably going to be losing money on the first 100 users, but after I get to 101, then I'm going to start
**Jingle My Brand's Potential**
(26:41) making money, and I can continue to grow. Those are all things what I'm talking about. The funny thing about Jingle My Brand is that I have people David, my wife, and other people saying, "You need to build this further. You need to build this further." But I essentially built it as a cool little tool. I'm, "It's definitely not a toy," but I'm, "It's
**Credit Consumption Transparency**
(27:00) just fun. Everyone's loving it. Everyone's having fun. Everyone calls it magic." I'm, "It's just there." But more conversations, I'm, "I need to jump on this more." And David, Zach brings up this other point that still confuses the heck out of me. I often have no idea how many credits I'm burning through, and then it's
**Credit Tracking Challenges**
(27:25) are any of the platforms really good at that?
Leo: No. I would say no. But that's why if you're worried about credits, and I still think this is all inexpensive, but 200 bucks can be a lot of money for someone, right? But ultimately it's 200 bucks versus spending thousands of dollars and months to build something, right? It's
**Firebase for Cost-Effective Building**
(27:47) all relative. If you were worried about money for whatever reason, I would use something like Firebase or even another one of Google's tools. Especially Firebase, it may not be the most powerful. It may not be the best, but it connects to a lot of Google's products in the back end, so things are easier, and sometimes
**Cursor for Comfortable Coders**
(28:08) cheaper. If you are a little more comfortable, use something like Cursor. I used to pay 25 bucks a month for Cursor, and I never hit my limit.
Zach: My question wasn't about, I think my question was to David, is it even possible right now with the way they're doing the tokens, and the way you consume the credits, to actually run a financial model?
**Financial Modeling for AI Agents**
(28:27) David: Oh, that's an interesting question. I look,
Zach: That's more what my push was. I got a buddy who runs a company here, and it's scaled, and they put in a bunch of AI agents to help them automate stuff, and they have no capability of modeling out their
**Is Small-Scale Financial Modeling Possible?**
(28:48) their input for what that translates on a monthly basis to their credits, and then the credits change.
Zach: Yeah. Just curious on the small end if that's even possible. David, do you want me to go?
David: Yeah. Why don't you go?
Leo: I think just anything else, it's a calculated risk. There's always going to be, unless you're fully verticalized,
**Calculated Risk in AI Costs**
(29:09) it's a danger that you're going to have to pay for, right? If I build widgets, and they're plastic widgets, where am I sourcing my plastic from? Are there going to be taxes? There's going to be that. So, I think it's always a bit of a calculated risk. But ultimately, a lot of this, because there's so much competition and the technology is moving so quickly, it is
**Inexpensive AI Generation**
(29:27) relatively inexpensive. I'm sharing this between us, I know it's being recorded, but us friends, if I just charge everybody here $20 a month, I'd be making money. It cost me so little to generate a song and generate a video that I'd be making money right off the bat. It could change tomorrow, right? But then just anybody else, that's how you plan.
**Diversification and Self-Hosting LLMs**
(29:48) I want to diversify. I want to have different tools. If it came to the place where I'm so worried about these tools, I could see about hosting my own LLM somewhere and self-hosted, which has its own other costs and stuff. But if the business grows, maybe that's something I could do with Amazon or Microsoft or something. It's an interesting time in the
**Sustainability of Free AI Tools**
(30:09) market because it is so competitive. At some point, the freebies are going to go away.
Zach: Are some of these things that people are building with vibe coding that are set aside as fun projects, if they did go to scale, is it something that's sustainable? That's, it's just a conversation I've had recently in the last couple weeks
**Digital Sovereignty and Platform Risk**
(30:33) with several people.
Leo: I love this conversation. I think we could talk for hours on it. The one thing I'll say, the sound bite of it all is, this is a big problem with digital sovereignty across the board. Look at Facebook. I come from the social media world where they built this garden. They told everyone they could come build
**The Facebook Garden Analogy**
(30:50) and get millions of followers on their page, and you post for free, and it's all great. Then one day they're, "Nope, we're going to build a gate and a wall, and it's not organic anymore. It's paid." Now you're, "Oh, but we need first-party data." Everyone's, "Oh, I need a website now." There was a trend when people, "Forget about a website. Don't
**The Continuous Risk of Digital Sovereignty**
(31:05) build websites." We live in this world where that is a continuous risk, and something that we all need to think about as users and business people, ultimately digital sovereignty. How does it exist? Does it exist with these big players? But I'll shut up now because I can talk about this forever.
David: You're the guest, so you don't have to
**Vibe Coding Focus and Frameworks**
(31:22) shut up.
Leo: No, no, no. But I want to focus on the vibe coding thing, especially if anybody has any questions, but I'm happy to jump into some tools or thoughts or frameworks that I wanted to share.
David: Yeah, and share away, right?
Leo: Awesome. First and foremost, plan as much as you possibly
**Planning with Whisper and Chat Tools**
(31:40) can. I actually use Whisper. I barely type. I just hit a button, and I talk directly into these tools. I open up a notepad, and I just talk. I'm, "I have an idea to build this tool, and the tool should do this, and I want to have this as the goal, and these are the type of audiences, and these are the type of users." I try to talk and iterate as much as possible in
**Start Ridiculously Small**
(32:00) ChatGPT, or one of the more conversational tools before I even put it into Lovable. So, when I'm feeding it into Lovable, I'm going in as planned and as thoughtful as I possibly can. If this is your first time building anything, I would say start ridiculously small. Build a calculator. Build a simple tool that you need help with, or maybe you don't even need
**Building Confidence with Small Projects**
(32:27) help with, literally an engagement calculator. Build something. The reason for that is I want you to build confidence. I want you to see the magic first before you're, "I'm going to try to build the best SAS in the world, or I'm going to replicate TikTok or something." It's going to break. You're going to fail, and you're going to be left feeling
**Importance of Backend Reporting**
(32:47) demoralized, "This doesn't work. I don't like this." So start very small. Build something ridiculously tiny and super simple, and then build from there so you can see the magic and gain the confidence. The other thing is I'm very big on building the back end. When I say that is reporting. For instance, yeah, Whisper Flow, I use that all day long. I
**Admin Sections and Data Collection**
(33:09) paid for it. I used the trial for two days, then I paid for it. "What am I doing here? Take my money, please." But build the back end. I have a backend admin section for all my tools. I know who's logging in, when they're logging in, when they sign the agreement, the songs that they're creating, things like that. I even make copies of certain things on the back
**Building Tools for Your Tools**
(33:28) end, not necessarily for Jingle My Brand, but I'm talking about across the whole website. Even when you use an email, it's collected in the back end. I could export it, right? So, think about that. The other thing, and this is going to be a huge money and time saver, build tools for your tools. For instance, I constantly build out. If I'm building
**Admin Tools for Content Management**
(33:49) this, let's say I'll give you the example for the Hanukkah one. See all these different styles. I don't constantly prompt Lovable or anything to update the different styles. I built a tool in the back end just for admins that says, "Allow me to build out prompts. Allow me to build out styles. Let me build out time frames," and things
**Gaining Control with Admin Tools**
(34:14) like that. Therefore, all I got to do is log into my website, go to my admin section, and constantly add different styles, and it's 100% for free versus going in there and, "Oh, update this. Now, change that." One thing I hate about that is I actually lose control if I'm just prompting. I don't know what it's building. I don't know what the prompts
**SEO for SAS Platforms**
(34:33) are. I don't know how the API connection looks. I don't know how many people are using it. So, I'm constantly building tools for my tools, and it saves me a ton of credits and a ton of money. The other thing I want to say is, when it comes to marketing and SEO, and I'm talking about SAS platforms more than if you're just building a landing page or some small marketing
**WordPress for Front-End SEO**
(34:58) project. What I do is, if you look at sociallollipop.com, it actually runs off of WordPress. So, I am able to control the website, the way it looks, build the blog, all that without using credits or anything. But if you start using the tools, then you go to app.sociallollipop, and that's when you're in the back end for
**Vibe Coding Tools and SEO**
(35:20) the vibe coding thing, because using WordPress or something else has better SEO. Right now, most of these tools suck at SEO. You could always build tools in the back end. I have for Jingle My Brand, I'm, "Allow me to have a tool on the back end that's for SEO, updating metadata and update." But I find that it's just easier and better to build something in the front
**Getting Started with Vibe Coding**
(35:38) end, and have the back end be built by the tools. Has anybody built anything, or is anybody currently building anything?
Earl: Leo, thank you so much for some of your tips. My question is, I've heard about the vibe coding, and I speak every day with my custom GTP bots on my AI team, but getting started with vibe coding, I'm, "Where do I even start?"
**Lovable.com: Your Starting Point**
(36:02) Literally, I get AI, but I'm, "Why do I even start in the beginning?"
Leo: I would tell you this. Go to lovable.com. Go to any one of these tools, Replit, or Bass, or Firebase. We could drop in the links in here, and it's super simple. Let me show you this again. Here's Lovable. It just says, "Start telling me
**Building a Simple Calculator**
(36:25) about the tool you want to build." Say again, start small. Say, "I want to build a calculator that measures engagement. I put in the interactions. I put the impressions. I put in the reach. And give me the answer." Hit go, and you're going to see the magic. You're going to have a tool that you could use. Especially if you start small, you're going to see something be built.
**Career Prospect Slot Machine Challenge**
(36:44) Earl: Awesome. I've always wanted to build a slot machine for career prospects. This is the industry, this is the company, this is the role. Do a slot machine and see what comes up based on my skill set.
Leo: Having spent enough time in these tools, I bet you can build that within the hour.
Earl: Sweet.
**Digital Marketplace Trend Space**
(37:02) Earl: Challenge accepted.
Leo: Go, go. If you need any help, reach out to me. I'm happy to talk.
David: I want to see it, Earl.
Earl: All right. Please.
David: Gauntlet thrown. I see you, David. Amy had a question about, "Curious if you know, Leo, the best tool to create a digital marketplace trend space where it constantly updates and creates
**Connecting to Trend APIs**
(37:22) charts to bring the insights to life. Can anything do that?"
Leo: Yes. Yeah. The answer is all of them could do it. The hardest part for you, and I could help you with this. I'm happy to go in deeper, but there's not enough time here to do it. It is then connecting to the APIs, right? Making sure that you have the connections to, let's say, TikTok
**Building the Frontend and API Connections**
(37:42) trends API, the YouTube Shorts trends API, and things like that. The front end is going to be pretty easy to build, and this is easy too. I'm talking about degrees, but you build a front end with the marketplace that's going to be Lovable or Replit. Then you need to find the API connections, get your API tools, get your app set up, and then make sure
**Data Scraping Complexity**
(38:01) that you collect all that information and feed it into Lovable or Claude or whatever, and you're going to be able to build that 100%. There are other tools that you could use to scrape data and have agents go out into the web, but that gets ridiculously complex. Essentially, any of these tools could do it.
Amy: Okay, you're definitely going beyond what I was even
**Learning Vibe Coding as a Newbie**
(38:22) imagining.
Leo: So, that's awesome.
Amy: Maybe I'll reach out to you on LinkedIn.
Leo: Yeah, please.
Amy: Okay, thank you.
Leo: Yeah, go ahead. Of course. Happy to do it.
Karen: I'm still learning what vibe coding is, a newbie in this space. I'll go as
**The Addictive Nature of Vibe Coding**
(38:42) Leo: Yeah. The best thing, being able to talk to one of these apps and have it build a site for you. It's, yeah. This thing, once you start doing it, it can get so addictive.
Karen: Yeah. It blows my mind. It really does. I'm going to keep going.
**Vibe Coding for Landing Pages**
(39:05) David: I literally do not give a talk anymore. I was prepping a talk that I'm giving tomorrow. Before I worked on the slides, I vibe coded the landing page. I was kind of more excited about the vibe coding than the slides. So yeah, but now it's totally messing up the order. I'm, "I can't not have that."
Leo: You build something.
**Building Tools for Day-to-Day Tasks**
(39:28) Leo: Yeah, I do the same if I need something at my day-to-day job. I'm, "Oh, this is hard to connect these two data points or mixing this." I'm, "How can I put it into whatever tool they give me to use and build it out?" One thing I also want to say with the short amount of time we have, Karen, you asked a good question. I'm sorry if I'm mispronouncing your name, but you asked a good question, how many people are actually using
**Marketing and Sales, Not Just Building**
(39:45) these things, right? Here's the thing with all these things, and I believe this across all AI tools. We're in the marketing game. We're in the sales game. This isn't necessarily on the building game. If you go right now, "Is there a tool that does this?" There probably is a tool. There's a tool that builds that you could probably build it.
**Winning in the AI Tool Market**
(40:07) It's the people that can get in front of clients, get social media attraction, get growth. Those are the people that are going to win because ultimately, if you want a tool that looks at social media analytics, I bet you if you go to any one of these directories, you're going to find 50, right? We're still in a weird place where can't ChatGPT just do
**Users Want Tools, Not Chatbots**
(40:29) this for you, right? If you're building a certain tool, a calculator, can I just go to ChatGPT? One thing I found, especially when it comes to speech language pathologists, they don't want the code start, "Here's a chat, ask it anything." They actually still want a tool. They just want a hammer, right? They don't want the Swiss Army knife that they have to
**Reverting to Simple Tools for User Engagement**
(40:47) figure out how to open and do that. People still want to say, "What's the age of this person?" and put in the information. I learned this the hard way. I had a bunch of tools built, and I'm, "I could just create a chatbot that does all this. Just ask it what the age is, and have it create a lesson plan." The user base went down. I reverted back to these
**Vibe Coding: Web Pages vs. Apps**
(41:09) simple little tools, and I got my traffic back up.
Earl: So for vibe coding, is it only really building web pages or things on web pages, or there other extensions of it?
Leo: Oh, that's super. No, I love this. Lovable and Replit and Firebase are a little more just for front end, SAS, server-side applications, right? But Cursor,
**Building Mac and Mobile Apps**
(41:39) Claude Code, and a lot of these other tools, you could build apps for your Mac, build mobile apps, and things like that. Yeah, this does get harder, right? Because then you need to understand how to install certain backend code stuff, Java on your Mac, and have it connect. A lot of times you're not seeing what you're building until it deploys. That gets harder, but
**The Challenge of Non-Web Interfaces**
(42:02) it's definitely possible. I've built some things on my backends, just as a Mac app that I wanted
Earl: Because I, you can just make web page after web page, but that's not necessarily how consumers are interfacing. They still are, there's all the rest in it.
Leo: One of the bigger problems that a lot of
**App Conversion for Vibe Coded Tools**
(42:21) these tools have right now, but they'll probably fix in the next few months, Lovable and some of these other tools, is that you can't actually create an app for it, or software for your Mac or Windows or something that's jailed within a website. I would love for Jingle My Brand to be an app on your phone, and it can be. There are tools
**Connecting Payment Systems**
(42:41) that you do a search, you'll find tools that take your vibe-coded app and make it into an Android app. You'll find those things, but it's not as easy as you would think.
David: Cool stuff that folks are sharing. Christopher, yeah.
Leo: So, what's funny is on Jingle My Brand, I already connected the payment system and all
**Stripe Integration Complexity**
(43:00) that, but it's super simple. A lot of times, Lovable itself will prompt you, "You need to create a Stripe account, and Stripe account needs the code." Then I'm, "Wait, where do I connect? Where do I go into the chat?" I'm, "Where do I get the code for Stripe?" Then it gives it to me, and then I put it into Lovable. So yeah, that gets super complex. The other
**Hitting Technical Walls**
(43:20) thing I'll say is for Social Lollipop, I've built the tools you see in some of the other things. There's going to be a time when I'm going to hit a wall, or I've hit walls before, where I'm, "I want to build this," and it's super complex. It doesn't exist out there. I need someone with a technical mind to go in there and
**Needing Technical Expertise**
(43:39) start coding and connecting things. So you will hit a wall as you start scaling, or I waste a lot of time asking questions that someone who knew better would have asked a better question. Christopher, have you built anything or are you building anything right now?
Christopher: Oh, he has. Yeah, he's been sharing some cool stuff.
**Wish List and Community Sharing**
(43:56) Leo: Oh, nice. I'd love to see it. Yeah, I'll make sure you get the chat too so you can see any of the links because we always have the funny, fun added conversation. Yeah. What other things is people trying to understand, or what's on your wish list?
Christopher: Yeah. I see Christopher. Lovable can get you a
**Prototypes for Technical Collaboration**
(44:26) prototype, which can help you sell the idea to a technical person, help you get to the finish line.
Leo: Yeah, that's exactly where I am with Social Lollipop. I have been going to events, startup events, startup pitches, and presenting it.
David: And this is something where, yeah, I also, for those of us who maybe have clients with their
**Explaining Ideas to Dev Teams**
(44:48) own development teams and things like that, where we're not going to ship a finished product to it. Then I've worked with others where I'm, "You can get this so much of the way there, and tell them the kinds of things you're looking for, instead of you trying to explain, 'Okay, I want this kind of site, and it looks Airbnb, but for booking
**Creating Visuals for Design Concepts**
(45:14) restaurants, and it should be this kind of thing, but it should be a much more old-fashioned design because these are more old-fashioned trends,' and trying to do all this stuff." You create something yourself that's so much more further along than I could never use Figma or all these tools. I'd spend forever, even in Canva,
**The Last Mile: Designers and Developers**
(45:35) trying to make something look right.
Leo: And it's, great. Then when they have a designer and developer, all that you can work with, they'll figure out that last mile and make it way better than you can imagine. Yes. Yeah. The other thing I think there's a pitfall, and this is we saw this, David, when we were in the SAS world,
**The Dangers of Self-Building**
(45:54) or even the agency world. Sometimes building stuff yourself could be dangerous. Things break. Things need to be updated. All that, unless you have someone dedicated to be looking at all that and changing it, especially if you're building something that's client-facing. If Jingle My Brand is low stakes, but if it broke at 3:00 in the
**Resource Planning for Tools**
(46:12) morning, who's fixing it? Who's updating it? Who's looking to see if it's done, right? If my business is marketing, and my business is, I don't know, selling widgets. I'm focused on the widgets, not necessarily on this one little tool that's a lead gen tool or something. So be careful to build things without properly resourcing it or properly understanding how much
**The Replet Trap: Unused Tools**
(46:30) time it's going to take.
Zach: Hey, thanks. This is great. I built a couple tools in Replit. The problem that I find is that I'll build something that has a limited use case for me, and then I'm just stuck paying monthly fees on something, I'm, "Oh, maybe I'll use it on this next thing or whatever." But it's
**Subscription Overload**
(46:52) just, it's a trap I find. Then I've got all these now. I'm looking at Claude Code, and I've used Lovable for sure, but I just end up with all these different apps, and with one thing or two things that I used for. How do you, what are your thoughts on that?
Leo: I'm going to give you more of a
**A Bigger Picture: Marketing Costs**
(47:15) bigger picture look at things. I try as a marketer, and money is money. Everyone has different financial situations from a business to a personal perspective. I get it, and I'm not minimizing that or doing anything to ignore that. So, there's that, but let's skip it for a second. I always say, "Look at how much marketing cost in the 90s." They cost so much money, and you
**Digital Marketing: Inexpensive Targeting**
(47:36) didn't even know how well the targeting was going, and where it was going, and the resources it took to build it was incredible. Now there's social media, and there's digital marketing, which is stupid inexpensive, right? Even as it gets more expensive, it's still really inexpensive. If I took a time machine, and I went back to the 90s or 80s, and I said, "There's a tool that
**The Value of $20 a Month**
(47:56) you could use to target almost anybody you want, and it's relatively inexpensive." You'd be, "Here's all my money. Forget about that billboard. Forget about the TV. Take all my money." Right? I'm, "This exists, and it's social media." So I look at that in the same world as the question you proposed. So this is costing me 20 bucks a month, $25 a month. You got to ask
**Assessing Tool Value**
(48:17) yourself, "Are these individual tools that I built actually worth that much? Is the value there for myself?" But ultimately, it's so inexpensive for you to have built that tool. Would it cost you thousands, maybe hundreds of thousands of dollars, right? Or taking you so much time. If it breaks, it's going to cost you hundreds to fix it. So, would I pay $20 a month to have
**Cost vs. Efficiency**
(48:37) a simple calculator that saves me two hours a day, two hours a month? I'd say yes. So, you got to look at, what have you built? What's the value it's providing, which is time, money, efficiency, and is that worth what you're spending? If you look at it from that calculator or that algorithm, that will answer your question.
Zach: But you haven't found there isn't
**The "Jailbreak" Dilemma**
(48:57) a way that you're, "Okay, this tool's done." I can't jailbreak it. We're probably never going to use it again, right? That's where I end up with on a lot of these things, right?
Leo: Yeah. I've run into that. Ideas I've had, I've built, or I've hit walls, and I can't build it any further. But because I'm building multiple tools,
**Managing Multiple Tools and Credits**
(49:16) and multiple things within the same platform, I say, "All right, that one's dead." But I'll use the credits for Social Lollipop or Jingle My Brand. So, I'm always building something. That's my own unique use case.
Zach: So, are you still just maintaining your subscriptions across a bunch of these things, and figure you're
**ChatGPT as an Affordable Partner**
(49:33) okay.
Leo: Yeah. Yeah. Exactly. But again, I look at it from the frame, "Shoot, ChatGPT is a partner that helps me think. Even just that's cheaper than a VA. I'm paying 25 bucks a month. That's free." Not everyone has $25 a month to spend, but what?
Zach: It's just, it's very easy to end up with
**The Challenge of Too Many Subscriptions**
(49:51) 50 tools that you're paying 30 bucks.
Leo: Yes. Yes. Yeah. Totally.
Zach: I, that's the challenge that I wrestle with a bit. It's, "Oh, I did this cool thing. Am I using it a lot? I don't know. I feel I'm going to use it."
Leo: There's only so many hours in the day. That's the, they're all fun. They're
**API Errors and Time Investment**
(50:07) just as you said. Some of it was headache for me building the app in Replit. I tried to do a few apps that just weren't possible. It just couldn't.
Zach: I just kept running into API errors and stuff, and I was finally, "All right, I've spent 25 hours trying to build this thing. How many more hours am I going to spend?" Right?
**Voice Cloner Success Story**
(50:24) Leo: Amen. Amen.
Zach: But I did build some simpler things that, to your point earlier, they're the calculator. I built this voice cloner that would take my clients' voices through Eleven Labs since it's a huge pain to go directly through there and do that. I was able to do that for this big video series that we did for them, and they weren't able to
**Subscription Philosophy**
(50:44) come back to set. So it was able to recreate that, and that was worth it for sure.
Leo: Yeah. No, I guess my philosophy, I look at it as this is just a problem with anything. I have a subscription to Wired. I have a subscription to Men's Health. I have a
**Valuing Subscriptions**
(51:02) subscription to this. I'm, "Just the other day I went through, I'm, 'I don't even read this. I don't have time to read that. Why am I paying for this?'" But I focus on the ones that are providing value. But I would love to subscribe and read all these things, right? Same thing for AI tools. If it's worth it for you, cool. If you
**Conclusion and Appreciation**
(51:17) could afford it, cool.
Leo: But if not, get rid of it.
David: Well, Leo, this is always better than expected. Appreciate you coming in on the fly and
Leo: Sharing all this stuff with us.
David: And in the community. Yeah. It's fun to get to see a lot of the V0.1 versions well before they launch. Really honored for
**Upcoming Speaker and Community Engagement**
(51:45) that. Thanks everyone for showing up. We got Peter Shankman next week talking about AI and comms, and he's also a dynamic speaker. You're in for a treat. Great to see everyone, and look forward to seeing what all you create with all these new things.
Leo: Yeah, please reach out to me. Happy to look in. Thanks for having me on all the
**Farewell and Future Invitation**
(52:08) great questions.
David: Yeah. You, I think you'll be welcome back here, so we know where to find you.
Leo: Awesome. Thank you.