56:25Greg Isenberg · Interview
Screensharing Kevin Rose's AI Workflow
A Digg founder walks through the full pipeline of a personal Techmeme-clone he built alone — from RSS to vector clusters to an editorial gravity engine.
February 2ndThe argument in one line.
A niche AI mobile app can reach $10K a month without coding if the core feature explains itself in five seconds of someone else's video and the founder is genuinely passionate about the problem.
Read if. Skip if.
READ IF YOU ARE…
- You want to build your first consumer mobile app and have no coding background.
- You have launched an app with some downloads but cannot convert them to paying subscribers.
- You are trying to close influencer deals at rates under $2 CPM without an agency.
- You are skeptical that vibe-coded apps can generate sustainable recurring revenue.
- You are looking for a founder-tested system for running influencer outreach at scale.
SKIP IF…
- You are building B2B SaaS — every tactic here is specific to consumer mobile.
- You want a technical discussion of vibe-coding architecture or scalability.
- You are already past $10K MRR and need advanced growth frameworks.
The full version, fast.
George argues that idea quality is the multiplier most founders underestimate — his proof is running WrestleAI ($17K in month one) alongside a Rizz app that got 1.8 million views and made $35. His framework centers on a single gotcha feature so self-evident it explains the whole app in a five-second clip, then layers on a dual-purpose Instagram page (user funnel plus influencer credibility), a VA doing outreach at $800/month, and paid ads starting at $100/day from competitor creatives mined in Meta Ads Library. Three metrics decide whether to scale: conversion rate, a $2 ARPU floor, and monthly retention.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Create a free account →Who's talking.
01:10guestGeorge Lampropoulos
00:00hostGreg Isenberg
Where the time goes.
00:00 – 01:54
01 · Intro and Social Proof
Host frames George's story and sets up the $10K/month promise.
01:54 – 02:36
02 · George's Track Record
WrestleAI: 100K+ downloads, ~$200K revenue; two stealth apps tracking $10K-$15K/month.
02:36 – 06:29
03 · How AI Unlocks Fresh App Ideas
Three pillars of a good idea: simple, specific, sellable. AI democratized niche app creation.
06:29 – 16:16
04 · Reverse-Engineering a Viral Idea
Scroll your FYP as your ICP; ask three questions. The WrestleAI vs. Green experiment.
16:16 – 17:49
05 · Product Design and UX
Hand-it-to-your-mom test. 14-day core build, then onboarding, then RevenueCat.
17:49 – 21:25
06 · The Gotcha Feature
One feature that explains the whole app in 5 seconds. 90% of distribution effort goes here.
21:25 – 23:04
07 · Onboarding That Converts
Four functions: educate, social proof, personalize (sunk cost), FOMO before paywall.
23:04 – 28:55
08 · Actionable Plan to $10K/mo
Clean Instagram page as dual-purpose asset: user sales funnel plus influencer credibility signal.
28:55 – 33:35
09 · Outreach as a Numbers Game
Tailor FYP to ICP, DM at volume, hire VA at $800/month, close creators on calls.
33:35 – 36:20
10 · Paid Ads
5-15 creatives at $100/day, cut losers after a week, mine Meta Ads Library for formats.
36:20 – 38:30
11 · Reading Metrics
Conv rate after 100+ daily downloads, ARPU ($2+ target), retention at month end.
38:30 – 39:51
12 · Good Product Earns Inbound Creators
When the product is genuinely good, creators with 1.8M followers start reaching out unprompted.
39:51 – 43:35
13 · Vibe-Coding Skeptics
No AI slop — only underprompted. 14-day build cycle. Swift on Rork vs. React Native.
43:35 – 46:05
14 · Why Now Is the App-Building Boom
Comparison to 2015 dropshipping. 85% of the world does not know AI can build software.
46:05 – 48:09
15 · Closing Thoughts — Learn to Sell
Final sauce: social skills close influencers at $2 CPM. Door-to-door sales at 16 still paying dividends.
Lines worth screenshotting.
- A Rizz app with 1.8 million views made $35 in month one. A wrestling app with 1 million views made $17K. Idea quality is the multiplier, not view count.
- The gotcha feature is a product spec written backwards from a 5-second video clip — if it cannot be understood in that window, the feature is not ready.
- Influencer marketing is 90% sales. Every rate you pay is negotiated, not listed anywhere.
- A brand Instagram page with bought followers and creator collab posts does two jobs: warms up curious users and gives you negotiating leverage on creator calls.
- ARPU below $2 per download in month one is a signal — either traffic quality is low or the paywall is placed wrong.
- Onboarding's sunk-cost mechanic works by stacking 15 to 20 personalization questions before the paywall, not after.
- You only need one creator with high agency to run paid ads — have them copy competitor winning creatives 1-to-1 from Meta Ads Library.
- 85% of the world still does not know AI can build software. Consumer app markets are underpenetrated, not saturated.
- Churn dropped in half for WrestleAI after adding a calorie tracker. Viral features attract users; utility features retain them.
- Spending $800 per month on a VA for influencer outreach delivers asymmetric ROI at the early stage versus any other marketing spend.
- Social skills close influencers at $2 CPM instead of $10 CPM — rapport built on a call converts to rate discounts.
- There is no AI slop, only underprompted. Fourteen days of iterative prompting in Swift produces real products; one-shot generation does not.
The gotcha feature is the whole product spec.
WHAT TO LEARN
Build the app backwards from a 5-second video clip — if the core feature cannot be understood in the time a thumb hesitates over a scroll, the idea is not ready.
- A niche you are personally embedded in is a distribution shortcut — you already know the influencers, the content they make, and what their audience would pay for.
- The gotcha feature is a product design constraint, not a marketing exercise. If you cannot demonstrate core value in under 5 seconds of someone else's video, the feature is not clear enough.
- Onboarding should stack commitment before the paywall. Personalization questions create sunk-cost friction; a mock-analysis animation creates FOMO. Both increase conversion without changing the price.
- ARPU under $2 per download is a diagnostic signal. If influencer-driven traffic converts below that floor, the problem is the idea or the onboarding — not the influencer.
- A VA doing outreach at $800 per month can DM hundreds of creators a day and hand off responses immediately. This is asymmetric ROI at the early stage.
- Churn is a product problem, not a marketing problem. Adding adjacent utility features cut WrestleAI churn in half — viral features attract, utility features retain.
- Social skills close influencers at $2 CPM instead of $10 CPM. Learning to sell in any context compounds directly into cheaper distribution for years.
- Underprompted is the actual failure mode of vibe coding, not AI slop. Fourteen days of iterative prompting produces real products; one-shot generation does not.
Terms worth knowing.
- Gotcha feature
- The single app feature that explains the product's entire value proposition in under five seconds of screen time, designed to stop a scroll and drive downloads without a voiceover explanation.
- ARPU
- Average Revenue Per User — total monthly revenue divided by number of downloads, used as a quick signal of whether traffic quality and paywall placement are working.
- CPM
- Cost per thousand impressions, used here as the unit for negotiating influencer rates — a $2 CPM means paying $2 for every 1,000 views the creator's video receives.
- Vibe coding
- Building software through iterative AI prompting rather than writing code manually. The output quality depends on prompt depth and iteration count, not on whether the builder knows how to code.
- Rork
- An iOS app builder that generates Swift code via AI prompting. George credits it as the only vibe-coding platform that actually outputs native Swift rather than React Native wrappers.
- RevenueCat
- A third-party SDK that manages in-app purchases, subscriptions, and paywalls across iOS and Android, abstracting away the App Store billing complexity.
- FYP to ICP alignment
- Deliberately training your social feed algorithm to surface only your ideal customer profile's content, turning passive scrolling into a live lead-generation stream for influencer outreach.
Things they pointed at.
04:56productCal AI
14:14productManifest app (Stella)
Lines you could clip.
08:36
“Wrestle AI in the first month did $17K, and Green in the first month did $35. It was five weekly subscriptions. That's all we got from 1,800,000 views.”
19:14
“This feature needs to be so simple to understand that anyone who looks at this feature for five seconds instantly understands the entire thesis of your app.”
41:12
“There's no such thing as AI slop. There is, though, underprompted.”
45:52
“85% of the world still doesn't even use AI. They don't know AI can actually build software.”
07:33
“All your distribution is is you're distributing the idea and the mission of your product. So if you have a bad mission, it won't go far even if you get a lot of eyes on it.”
Where the conversation goes.
02:36 – 06:29denseIdea selection
06:29 – 16:16denseWrestleAI vs Green case study
16:16 – 21:25denseProduct and UX
21:25 – 33:35denseInfluencer marketing
33:35 – 36:20steadyPaid ads
36:20 – 39:51denseMetrics and iteration
39:51 – 46:05steadyVibe coding defense
43:35 – 46:05steadyMarket timing
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
00:00There's this 19 year old guy, George, and he used to work at TJ Maxx. And today, he quit TJ Maxx because he's making hundreds of thousands of dollars building mobile apps using AI.
00:11We go through how he does it, all the sauce, all the tactics, and he explains it in just a really simple way that by the end of this episode, you're gonna really understand how you can build a consumer mobile app that makes hundreds, sometimes thousands of dollars a day with AI. Now I'm not saying that everyone's gonna be able to do it, but I think that what you'll learn from this episode that it is that we are living through a time when it is possible now to come up with ideas, mobile app ideas that have AI integrated into it and build using AI that could actually generate real money.
00:46This is the sort of stuff that it could be a side hustle, it could be some passive income, and it could just be a cool little app that you build that doesn't make money but is just fun to build. So enjoy this episode with George.
00:58I thought it was really interesting, and I'll see you at the end of the episode.
01:10George is on the pod, and he's gonna tell us how to scale your mobile app to 10 k a month easy mode. You've got my curiosity piqued.
01:21George, how by the end of this episode, are are are people gonna learn how to do this? Are you are you being for real? Yeah.
01:27So the goal is to give you the basic framework
01:30on how to get to 10 a month as easy as possible. 10 a month, all it really is when you break it down is like $333 a day.
01:40So to get there, all you really need to do is do one to two things right, and you should be you should get there relatively quickly.
01:49Alright. Well, show me how, my friend.
01:53Yeah. So so first, a bit of, like, social proof for, like, why you should listen to me.
01:58I scaled up I built my first app or second app, technically, WrestleAI completely in public. It got over a 100,000, uh, downloads and made close to 200 k in revenue.
02:10And now I've been building two other projects in stealth, and both of them are on track to hit their first one of them is on track to hit 15 k this month. The other is on track to hit its first 10 k month this month as well.
02:22And I've been working cumulatively about three to four hours a week on each of them. So it's it's not always time in that gets you the results.
02:31It's it's it's by following a basic playbook that can get you there fast. Okay. So the three things that make, like, a successful app, in my opinion, are a good idea, a product that matches the idea, and then distributing that idea to other people.
02:48So the idea is really important, and my advice to everyone is always solve your own problems and create something you're passionate about. Because when you're sitting there and working for, like, ten hours on your your app, you know what I mean, you want it to be something that you actually genuinely resonate with.
03:04Otherwise, you won't have that same like, when when you're talking to influencers, when you're talking to people about your app, they can tell whether you truly believe in your own mission. And if you don't, it becomes harder to convince these people to kinda join you for cheaper rates, let's say. So I I I'm really big on creating a product that you're very passionate about.
03:23For me, was wrestling. I was I I wrestled for, like, all my high school career and my middle school career before it got cut short by COVID.
03:31I love the sport. The sport genuinely changed my life, so I wanted to build an app for myself when I was a freshman in high school.
03:38What would I have loved to have on my phone? That was the idea. And your idea, it should be simple, have one clear one clear problem, one clear solution.
03:47It should be specific, built for a defined audience. This way, once you know who your audience is, you could then reverse engineer how to find them.
03:56For example, like, I knew my audience was wrestlers, so I reverse engineered my for you page on my phone to be only wrestlers so I knew exactly who I had to sponsor to get in front of the right audience.
04:08And number three is it should be sellable. People should be willing to pay for this pay for this solution today.
04:15There's one there's one thing missing from that, in my opinion. So Yeah.
04:21You use the examples of Wrestle AI, Cal AI. I mean, the question we have to ask ourselves is what are ideas that now AI has unlocked?
04:33Right? Like, so, you know, people have been building mobile apps since 2008, and it became increasingly harder to to build something that people are gonna download and pay money for until two things happened, actually.
04:50AI came. Anyone could create something, but more even more than that, you can create these AI first experiences that just are, like, better than the existing experiences, like Cal AI.
05:01You can take a picture of your, you know, your your spaghetti and it'll tell you how many calories it has. And then the second thing is well, obviously, well, that's the main that's actually the main piece.
05:16It's just like you're creating more value Yeah. Exactly.
05:19For the customer. Right? So it's like, how how important do you think that is to the idea?
05:25I think I think that's huge. Like, I think AI
05:28has created so many creative ways to make engaging features. And another thing I think it is is the fact that to create an app in the past, you used to need, like, I don't know, like tons of developers.
05:42You either needed a lot of capital or stuff like this. But now that AI has kind of democratized creation, you can also get super niche in your apps.
05:54Like, when when I was telling people my idea, they were like, okay. But how many wrestlers are there actually? And it turns out there was a decent bit, but most people would have kind of scoffed at the idea of, like, a a wrestling app kind of having, like, any sort of success.
06:09You know what I mean? It allows, like, people who are more niche down into things to create good products for these, like, more niche communities without needing a ton of capital investment.
06:22Cool. Alright. Let's let's continue.
06:24Like, WrestleAI never needed to be a billion dollar company. Anyway, yeah. So here's kinda my process.
06:32If I wanted to if I was sitting there and I was like, I need an app idea today, how I would reverse engineer a viral app idea is I would literally just open up my Instagram or open up my TikTok. I would scroll for, like, fifteen minutes, and every time I watched a video, I would ask myself the question, who watches this type of video?
06:48What problem does this influencer's audience face? And how could an influencer promote this app seamlessly? So if you looked at, like, the wrestling content creators, I would say, how could they implement my app for five seconds within the first thirty seconds of their video and have it not feel like an ad?
07:08So I would actually target a lot of day in a life and day in a life wrestling influencers, and they did great for us because of that. So why is a good idea important?
07:17So many people, especially on Twitter, love talking about this idea of distribution versus, like, product, and I'm really big on having a solid idea because all your distribution is is you're distributing the idea and the mission of your product.
07:33So if you have a bad mission, it won't go far even if you get a lot of eyes on it. And my one kind of proof towards this idea is a real life experience I had.
07:44So in September, when I was launching Wrestle AI, I was also working on this other idea on the side. Now Wrestle AI, I'm a wrestler.
07:53I'm super passionate about wrestling, and I was really excited by the product I made. And this other app, I kind of hated, to be, like, to be frank, but I had an influencer who had 2,000,000 followers that was willing to partner with me.
08:07So this is where I kind of, like like, threw away my like, have a good idea and be passionate about the idea. And I was like, I'm just gonna make this, like, AI Riz style app, and it'll make a lot of money because, like, this guy gets millions of views.
08:23In that campaign, in the first month, WrestleAI got around a million views. And in the first month, this app, Green, which was like an AI Riz dating assistant, got 1,800,000 views.
08:36And Wrestle AI in the first month did 17, and Green in the first month did $35. It was five weekly subscriptions.
08:46That's all we got from 1,800,000 views. And I was sitting there looking looking at this app and just going, like, okay.
08:54Maybe it's because my marketing wasn't creative enough. Maybe it was because this and that. And then I really sat there and got true to myself, and I said, it's it's probably because the idea just sucks, and everyone's seen this before, and that's why we're not really driving conversions while WrestlingAI is very unique in this niche.
09:11So that's when I made the decision. I'm only gonna be working on ideas that I'm passionate about, and I have fun I have I have fun actively working on, so I just completely cut this app. I I regret ever building it, but it did teach me a valuable lesson to just really go after what you're passionate about.
09:28I have a question I have a question about the TikToker.
09:32What was his name? Young Cantaloupe? Mhmm.
09:34How are you able to convince someone with 2,100,000 followers?
09:40I mean, just like a year or two ago, were a cashier at TJ Maxx, and then all of a sudden, you're convincing this multimillion, you know, follower account to work on you to partner with you on an app that has no validation or anything.
09:57Yeah. I'm I'm a DM wizard. So I'll, like, I'll like, if you look at my Instagram account, I was sending, like, probably, like, hundreds of DMs a day to just anyone who's willing to build an app with me.
10:10And before WrestleAI and this app, Green, I also had this app called Fight AI. So what got me really into apps was I was in chemistry class or not chemistry.
10:21Physics. I was in physics. I was scrolling on my, like, TikTok for you page, and I saw an ad for Rourke, and it said build an app in minutes.
10:28And I didn't really believe it. Like, I I didn't know what vibe coding was. So that was my first step into vibe coding.
10:32I built this app called fight.ai, and it made me, like, two k in the first month just from me posting about it and, like, paying a meme page $50. So that gave me proof that, like, apps are the new way to make money.
10:44And from there, I was like, okay. Let me find, like, a creator to partner with fifty fifty. This way I don't need to manually grow from zero to one k.
10:53It wrote zero to 10 k. That was my first idea. So in that process, I talked to this guy, got on the phone with him.
11:01And you have to be a bit, like, outgoing. You have to be you have to be able to sell a bit, especially for influencer marketing. I always say influencer marketing is, like, 90% sales.
11:10You're like convincing people, like, to take cheaper rates than other people because you're selling them the idea of a long term partnership with a product that should be genuinely good. So with the way I convinced this guy is I was literally showing my Wrestle AI conversion rates.
11:27And I was like, if I could do this at a small scale, imagine what I could do if me and you partner up.
11:33And that's what that's what got him. That's a good pitch though. Right?
11:36Like, it's basically like, all I need is more fuel on this fire. I just need more top of the funnel people to come and use this thing because Exactly.
11:45I've basically built an incredible mousetrap. And when you're building products, like, that's ultimately how you wanna think about it.
11:52You wanna you don't wanna build a leaky ship. You don't wanna bring in a lot of people and they're not converting, they're not retaining, it's not valuable for them. You wanna build something that's a mousetrap, that keeps them in, that brings them in every day, that they see so much value that they can't get away with it.
12:08So you're basically going to this guy and you're like, hey, I built an incredible product for this niche. You have this big audience and I think a part of your audience, our people would be interested in this.
12:21Let's partner up. And what do you propose some you know, don't know if you're able to say this, but, like, what sort of deals can you propose? I usually do and a lot of people get on me for this.
12:30Like if I would partner with an influencer, which now I don't anymore typically
12:34because I feel like this strategy kinda it it only really worked for like Wrestle AI. The rest of the apps I tried it with never it kinda didn't work, and it usually had me manually having to grow it a lot more.
12:47But what I did basically was I I would be just fifty fifty.
12:53I would be like, look, you take half, I take half. Got it. See what happens.
12:58And like sometimes, like you could you could also vest equity. Like I know a lot of people do it in different ways where some people vest a person's equity. To be honest, I was I was in college.
13:07I I just needed my first win. I I would give them 50% up front and see what happens.
13:13Cool. That was kinda my idea with it. You you know, you're saying like for your new apps, you're you're not finding influencer marketing to be helpful.
13:22Is that is that what you're saying? No. No.
13:25Not that influencer marketing isn't helpful,
13:28but the draw you get from just one influencer isn't as big. It's easier it's easier to boot up a nice Instagram account and just run Instagram marketing yourself and sponsor just tons of creators if you have the capital.
13:44Right? For people who don't have the capital, I I always suggest because I didn't have the capital back then. So Right.
13:50I I really want a response with an influencer because it was an easier jump start. Right? But if you do have the capital to throw a bit of money at some influencers, I would suggest just bear like, we'll go deeper into it on how to set up an Instagram Okay.
14:04Get you like a decent amount of responses. But I would suggest you just do it like that without even tying up so much of your equity to an influencer. Sounds good.
14:12The right influencer is a game changer. Like if you look at you know, I think, you know the manifest app, Stella?
14:20Yep. Just her alone, I mean you can check it on censor towers, I'm not gonna be doxxing her.
14:26Just her alone brought in like 300 k a month to that app, right? Yeah. Like she's an absolute killer and she works with like a dev shop.
14:34Just her alone is driving like 99% of that business and now they're getting into net ads and maybe other influencers, right? So the right influencer could be game changing, but you you should probably have some system where you could maybe test out their content prior.
14:54I know a lot of people who do this where they test out the person's content, see if it converts high, and then they decide maybe they wanna bring them on as equity partners.
15:03But it's Yeah. Hard to one thing no ethos. So maybe they do those first and then do that.
15:08That's my I mean, one thing you one thing you can do is, like, literally just ask them what their ad rates are and be like, hey. I'll pay your ad rates, whatever it is, and then you test the content.
15:21And then based on that, you're like, okay. That worked really well. Worked better than expected or worked as expected or didn't work at all.
15:27If it didn't work at all, it's like, great. You just saved yourself whatever percent equity. Yeah.
15:31Exactly. You know, that's that's a good way to do it if you have the capital.
15:36Yeah. Or or if you could probably you could probably use that idea to negotiate cheaper rates with them as well. So you'd be like, look.
15:44I'll pay you. I know your typical rates are like 2,000 a video. I'll pay you 1,000, but if it goes well, you'll have the opportunity to get equity in this project.
15:53And as long as it goes back to this, as long as you're building something cool and you're building something you're passionate about, it should be very easy to convince people that your product's gonna be the next big thing. You know what I mean? Okay.
16:07That's at least been my experience when building products that I'm genuine that I genuinely resonate with. I was gonna say like what else what else do you wanna show? So design in UX is very important as well.
16:19You should be able to hand your phone to your mom, and she should understand how your app works without you having to tell her how to use it. This is kinda my litmus test. If your mom's technical, maybe don't do this.
16:29My mom still doesn't even know, like, how Siri works. Right? So so this is my test.
16:35I literally text my mom my app, and I go, can you figure this out? Right? Like, can you use it?
16:40Can you tell what the app's for? Can you tell how it how it, like, functions? And, uh, if she can't, which a bunch of times she's been like, oh, it's it's too, like, cluttered.
16:49You know what I mean? So I've been like, okay. That's, like, that's insanely valuable feedback.
16:55So if your mom can't use your app, you have to redesign it, and you have to make the the UI look a lot more simpler and easier to use. Uh, how to build?
17:06I personally don't know how to code. I've all my apps have been created but with AI. I built all my apps through Roark.
17:13And my framework for building is I dedicate, like, fourteen days to actually building out the main functionality of the app. What's that what like, core functionality and branding. What's the app gonna do?
17:22How's it gonna look? Then I spend a few days perfecting the onboarding. Right?
17:27So, like, the onboarding is really, like, uh, I go more into it later, but it's educating the user on what your app does, personalizing the experience, and then causing FOMO what they would do if they don't have your app. What would happen if they don't have your app?
17:42And then I would spend and then and then after, I add in RevenueCat and back end integrations if needed. Okay. So this is one of the biggest things your app needs, and this is one one of the features that you should spend probably 90% of your time on.
17:58So as I was saying earlier on how to make a good idea and a good product, you need to have a clear solution to a clear problem. Right? This is what I like to call the gotcha moment or the gotcha feature.
18:09Right? So CalAI, it was taking a picture of your food and getting the calories.
18:13For WrestleAI, it was throwing your wrestling match into the app, and it told you what you did right and what you did wrong in your match. Right?
18:20These were the only two features that we were pushing 90% of the time. And if we had other features we wanted to push, we would basically push it after we began saturating an influencer's audience.
18:33So if I took one of my influencers, let's say Corey Zaider. Right? The this kid was an influencer.
18:39He posted day in the life videos, and he would always subtly integrate our app into his content. If he if I saw that the return from his videos started lowering, I would then tell him, okay. Push this feature instead, right, to see if we could re get the uptick even if it's, like, uh, less than the original.
18:56Right? However, the biggest uptick from your app should get should come from marketing this feature.
19:03This feature is what should easily convince everyone to come join your app. And what you need to do with this feature is you need to this feature needs to be so simple to understand that anyone who looks at this feature for five seconds instantly understands the entire thesis of your app.
19:21Right? Cal AI, it's three seconds. Take a picture of your food, get the calories.
19:26Wrestle AI, it was throw throw your wrestling match, get the feedback. Right? And within three seconds of looking at the phone displaying that, users understood without it having to be said.
19:38Right? So this is a really important feature that you need to nail. This is this is 90% of distribution is is nailing the gotcha feature.
19:47Yeah. I mean, another way to think about that gotcha feature, which I agree, it's, like, so so important, is you're sort of, like, reverse engineering the TikTok or the Instagram reel.
19:56Yeah. Exactly. And you're building the product around that.
19:59So that's one, like, framework that I use when I'm trying to come up with startup ideas, especially in the consumer mobile app space. I'm like, okay.
20:08Here's my niche, and I've you know, there's a bunch of different primitives or product ideas I can go after, but you know?
20:17And sometimes I use AI to kinda help me brainstorm these ideas, and I say, like, what, you know, what would be a TikTok that would go viral?
20:26Like, what is that one feature that would show that it would show? And what's beautiful about these gotcha features, as you call it, is it's the best products you can explain in a sentence or less.
20:39Like, think about it. Right? Like, you know, what is Hinge or what is Tinder?
20:44Like, you swipe and you find, you know, someone to date. Or what is Uber? You press a button and then a car comes and picks you up.
20:53Or what is DoorDash? You know, you press a button and food comes to your door. Right?
20:57So it's like, these are the gotcha features that I think that we we you know, when you're creating mobile apps, a lot of us have this tendency to overcomplicate the product suite when it's just one core feature that is what's going to help, like, stop the scroll.
21:19100%. 100%. Yeah.
21:21I fully agree. That's exactly like, I I think about it very similarly.
21:26There we go. Alright. Onboarding.
21:28But, yeah, anyway, onboarding.
21:30So so many people overthink and overcomplicate the process. Onboarding has four functions. Right?
21:35Three main functions, one optional. Educate the user. What what is your product and why do they need it?
21:40Right? Social proof. This isn't a 100% needed, but if you do have social proof, it's nice to show.
21:47And then you wanna personalize the experience. The point of personalizing the experience is making the ad better and also increasing the time commitment they have to your app. So if I go and answer 20 questions in your app, making the best experience for me, and then I'm gonna get hit with a paywall, I now have like the sunken cost fallacy which is like I've already spent so much time on this app.
22:07Let me at least get the free trial and see what it is and maybe I'll cancel it. Right? And finally is you wanna institute FOMO right before the paywall.
22:15So let's say I have like WrestleAI. I'm uploading my wrestling match.
22:20It's going through this like analysis animation, right, where it's like targeting x y z, calibrating x y z. And then after it finishes, it then shows you the paywall and it says, like, unlock your rating now.
22:33Right? So now they have a ton of FOMO. What was my score?
22:36What was this? What was that? And it drives them more to convert rather than if I just personalize the experience like, okay.
22:43What what's your weight? What weight do you wrestle at? And then you get hit with a paywall.
22:47Right? That add that mock film analysis is super important to increasing conversions.
22:55But that's that's pretty much it. I also wrote a docs in docs.roark.com where I kind of go go over everything in this presentation in much more detail.
23:04So now the most important part, what everyone's here for is distribution. Here's my actionables plan to 10 a month, and I think it's relatively easy if you just nail a few things. Right?
23:15One of the more important things to nail is your Instagram page. Right? You need to set up your IG and you need to make it clean.
23:21So the way I go about this is this is blacked out because this is one of the projects I'm working on now that's kind of in stealth. Right? But I first bought the amount of followers.
23:31Right? You go find one of these sites that you could buy followers for. What this does is it gives you a ton of social proof.
23:37Right? And then if you can get it verified, I would pay for that. And then I would get three product demos right here.
23:45Try to make them as clean as possible. Pay a designer in Fiverr if you need to or just design it yourself. So have your three product demos right here and have like a really good bio with a CTA.
23:56Right? This is 90% of what you need to do. And then the other 10% is anytime you work with an influencer, have them like, in the beginning, every time you work with an influencer, have them collab post because now not only is your does your brand account look professional, but it also seems like you have a lot of motion and you're working with a lot of people.
24:18Right? This way so this this page serves two functions. Right?
24:23It's a sales funnel for your user because if I'm if I'm looking at WrestleAI, if I if I watch one of my favorite wrestling influencers, I see WrestleAI is tagged in the in the caption, and I'm like, woah.
24:35That seems like a pretty cool app. I then click on the tag in the bio in the in the caption. It gets me to my Wrestle AI page, let's say.
24:44I'll then watch four more Wrestle AI videos on my on the Instagram page. Right? And now we've went from a user who's slightly curious to very educated on the product.
24:54Right? Now they're primed to convert. Then they're gonna see the CTA in the bio.
24:58They're gonna click the App Store descript they're gonna click the App Store link, download the app. They're then gonna go through the onboarding, so an already pretty primed user is gonna get even more primed, and then that's what should give you really good conversion rates.
25:11Right? So your Instagram is both a sales funnel and social proof for other influencers to join. Right?
25:19So on the other end, if I'm an influencer and I see a DM from some brand, right, and I click on the page and I see that they have two followers and no posts, right, I'm gonna be like, okay. This is some, like, app that no one uses.
25:32Right? And, like, I I wouldn't even I I probably wouldn't even answer. And if I do answer, when we hop on a call, you won't have any, like, ethos or credibility in the conversation, so you lose all your authority.
25:44So you have to pay whatever insane rates he asks for. Right? However, if I have a brand account and I'm showcasing that I'm working with tons of influencers, right, now the influencer sees it.
25:56They go, okay. He's working with a ton of influencers, so he can't he can't be paying $0. He has to be paying something.
26:02Right? So I'm gonna then respond, yeah. What are what are your rates?
26:05Blah blah blah. The influencer will hop on the call with the person who has the, you know, high authority Instagram account. They're gonna get on the phone, and then as the brand owner, if I propose low rates, if I'm proposing, like, a two dollar CPM, right, which is really low rates or really good rates, um, if I'm proposing a $2 CPM to this guy, this if without the brand authority, he would originally think, okay.
26:29This guy is trying to, like, rip me off. With the brand authority, it's okay. This try this guy's kind of trying to rip me off, but everyone works for these rates, so maybe these are just the rates.
26:40Right? It's much easier to convince someone to do something if they see that 10 other people also did it. So it's really important to just nail this Instagram page for closing influencers and using it as a sales funnel for your other for your other consumers customers.
27:00The other I thing is
27:02say, like, you know, I think I think botting your account is against terms of service. I'm not I think I could be wrong, so do your own research.
27:13Yeah, yeah. Don't always listen to me. You can do what you want.
27:16This is what you've done. Yeah, exactly.
27:19Mean, WrestleAI is actually not botted. This account is blacked out, so I mean, don't need to bot it too. Can also just Yeah.
27:27You could also yeah. You could also do that. A 100%.
27:30Content that gets shared and
27:33and do it the good old fashioned way. It's just gonna it will take you longer. Right?
27:36Yeah. A 100%.
27:39And this is something I've only really been doing on newer projects. Yeah. Although, I will say it does have a slight high ROI.
27:48Another thing I would do is I would tailor your this is this is arguably even slightly more important than the Instagram page, is I would tailor your FYP to be your ICP.
27:59So what I mean by that is I would go on Instagram, and I would scroll like your ad like your ideal customer. You should have this detailed image of who your ideal customer is in your mind.
28:09Right? And you should think how would they act on Instagram. So I would literally just scroll, comment, like, share as my ideal customer.
28:17I would get in I would, like, try to, like, hypnotize myself to be my ideal customer. And for me, it was easier because I was a wrestler and I knew what I knew what type of content wrestlers watch. That's another reason you should build for what you're passionate about.
28:28So I'm scrolling through Instagram. I then make my entire for you page only wrestling content, only wrestling influencers. And then what you effectively just did is you now created a lead generation, like a like a lead, uh, generating thing.
28:42Right? So while I'm going through my for you page, I could scroll outreach, scroll outreach, scroll outreach. It makes the whole process 10 times easier than having to go through, like, Instagram search and finding the right perfect influencers.
28:55Okay. So for my next thing, outreach, outreach, outreach. Everything in, like, this this game is a numbers game.
29:01How many influencers can you, like, send messages? How many content can you make? How much content can you make?
29:06It's all numbers. So and, like, the more you do, the more you'll get out eventually.
29:12So now that your FYP is fully tailored to your ICP, you can literally just scroll and DM as many influencers as you possibly can. To be honest, I would replace yourself in this process though as soon as you possibly can. Right?
29:25If you had 2,000 to start this app, I would take, like, 800 and just pay a VA for a month.
29:35I think I think I would be willing to do that. I would be willing to spend almost half of my budget just on the VA because they're pretty integral to the process, and it also leaves you more mentally clearer. Um, so what you do is you basically now that you have a system of scrolling and outreaching, you can literally just hop on a call with a VA from Fiverr, Upwork, Founders Arm, whichever service you wanna use.
29:57Right? Hop on the phone with these people and just pay them whatever you want because they're not gonna ask for outrageous rates typically.
30:05Right? And you don't wanna be like you don't wanna be negotiating for cents on the dollar for you. You know what I mean?
30:10100%. If they ask for 800 a month, just pay it. They're gonna be working like you're gonna get way more than $800 in value.
30:19So when you hop on a call with them, show them the process of literally just be like, hey. When I scroll like, just go through my for you page, scroll. If the guy has over 25,000 views on average, outreach to him.
30:31That's that's literally what I tell him. Right? You could also make like, you could also give a really good s SOP for them, and that's pretty much the whole process.
30:41Then when a create when a creator responds, get get hold on.
30:46Sorry. This is, like, misspelled. When when a creator responds, aim to hop aim to hop on a call with them immediately.
30:54Right? So if let's say let's say I'm the brand.
30:59I text paid promo for x y z. And if you if you wanna DM scripts, again, docs.org.com, you could find, like, all of my DM scripts, my SOPs, basically everything I've done.
31:11So once a creator responds, try to get them on the phone immediately. Say, like, hey.
31:17Uh, are you around for an intro call with someone from our team? You know what I mean? From there, you could then once once face to face with them, it's much easier to sell them than over DMs.
31:28Because over DMs, if I'm an influencer, I could be like, my deliverables are $10,000 a video. It's much harder to say that to someone when they're right in front of you and they're basic like, you can basically show them on $10,000 a video for a creator that gets, like, 20,000 views a video is not, like, ideal, and you can't work long term.
31:47So, like, there's so many sales objections you could kind of use to get good rates. And as long as you're really passionate and it and they can see that you're not selling snake oil, they'll be much more unlikely to do cheaper rates for you.
32:01They'll give you, like, the the homie discount. You know what I mean? If you're just a charismatic, like, if you try to be, like, nice to them, try to be friends with them.
32:09And with with all my creators, I try to be as close as possible. Like, I'll pay attention on their Instagram if they have any important events coming up, and I'll text them, like like like, we have a lot of fighters, so people are going to fight. I'll text them, hey.
32:21Good luck. Like, I'll try to be very on top of having really good personal relationships with my creators.
32:28Then after you finish the call, try to get their WhatsApp. I I I like WhatsApp because it's easier to keep them separate. Right?
32:35So when I go on WhatsApp, know I'm gonna be dealing with influencers or my VA. But I know on text messages, I'm not gonna lose my mom's text when she texts me. You know what I mean?
32:44Because that was, like, a big problem when I started. I would, like, not be I was, like, never responding to, like, my family or my friends. And I was like Totally.
32:51I was like, this kinda sucks. Like, I wanna I wanna I I used to pride myself on my ability to respond to people fast and efficiently. And so I was like, okay, let me just separate the two so it's easier to have.
33:01Your mom could be giving you precious UX feedback and You'll you're not seeing never know.
33:09Plus, love my family. I would never wanna miss their quest. And then get them on a contract, and then the rest is kind of have them subtly integrate your app into their videos.
33:19Try to subtly integrate your app into their most viral content style. After that, it's pretty much smooth sailing. 90% like, the hardest part is basically getting on the phone with them and pitching them the idea.
33:31Uh, then your app your app, if your idea is good enough, should sell itself in their content. Uh, okay. Paid ads.
33:37To be honest, I've only just recently started running paid ads, but I'm starting to fall in love with them, and I'm starting to understand why everyone says that influencers aren't scalable past a certain point. Right? Influencers are great if you have not a crazy budget and you need to 10 x your rev like, you need to 10 x whatever money you have.
33:56But if you don't mind consistently two to three x ing your your money, right, paid ads are amazing. So once you get about five to 15 creatives, make sure they're quality creatives, throw them all in a testing budget, run them at a $100 a day, and then after the first week, I'll separate the winners from losers.
34:15Again, I'm not super optimal with it. Like, take everything I'm saying about paid ads with a grain of salt, but it's paid ads are pretty good.
34:24So another method is if, let's say, you're working with multiple multiple creators, you take five to 10 five to 15 creatives from them, throw them all in a testing campaign. Right? That's one end.
34:34But if you don't have five to 15 creatives from, let's say like, let's say you don't have five to 15 creators. Right? If you want, what you could do is you could pick one of your creators.
34:44You could pick one creator overall. Right? As long as he has high agency and is, like, cheap, have him basically not post content on his page, but have him just make paid ads for you.
34:57And the way you could find winning paid ads is you go to Meta Ads Library, look for people in adjacent niches or in your or, like, one of your competitors, if you have any. Right?
35:08Look at their ads. Understand why their ads work. So here's here's what Meta Ads library looks like.
35:14Right? So this is Strava. So let's say I'm making a running app.
35:17I would look up Strava. I would go through all of these paid ads. I'd be like, okay.
35:21Why is this paid ad? You could filter by high to low impressions too. Right?
35:26So it's like, okay. Why is this paid ad converting? Why does this have so many high impressions?
35:30Why are they spending basically so much money on this paid ad? They're probably getting a return. Right?
35:35Why does it work? How can I replicate it? Those are, like, the two questions you need to ask yourself.
35:40Then if you have a creator that has high agency, literally just send them the video and say, hey. Can you remake this video one to one with my app? And that like, I'm doing that.
35:49That's literally been my only growth channel with one of my new projects. We've only been running paid ads. I have one creator who has high agency, and he just makes all of our creatives for you.
36:01And you'll be shocked at the paid ads that, like, make you a lot of money. It'll be like literally a video of him saying, like, go download the app.
36:09Like, paid ads don't need to be this, like, super high quality content all the time.
36:16As long as it's real and, uh, authentic, it'll drive decent conversions. Um, now let's say after all of this, your first your first month in business is starting to end.
36:28That's when you'll probably wanna do this a little bit earlier actually, but that's when you're gonna wanna start looking at your conversion rate after you have some decent traffic. Right?
36:37So if you're consistently getting over, I'm gonna say, a 100 downloads a day, Now you have some rough metrics to look at your conversion rate and see where you could optimize your onboarding or paywall. Right?
36:48Maybe start AB testing pricing to see if you could boost your conversion rates. I would also look at your ARPU. This is also a really important statistic depending if you're depending how you market.
37:01So people who do, like, this really top of funnel UGC, like like girl on screen with, like, text. You know what I mean?
37:09Their ARPU is gonna be a lot lower because they're getting millions of views. Right? But the quality of the content isn't crazy.
37:16But if you're doing, like, influencer posts, your ARPU should be pretty high. You know what I mean? Because it should be kind of directly explaining what your app does, and it should be like high converting content.
37:26So look at your ARPU. Aim for a $2 ARPU at least in month one. And then ARPU basically being like how much revenue you're making per download.
37:36And then look at your retention. So now that the month's, like, starting to be, like, close to being over, see how many people unsubscribed. And then if your retention's really bad, it's probably because your app's not useful enough.
37:49So try to figure out how you can make it more useful. For WrestleAI, in the first month, our attention sucked.
37:55And what we did was we literally just added a calorie tracking feature because a lot of wrestlers are always managing their weight. They're always cutting weight. So once we added the calorie tracker, we saw our churn cut in half.
38:06And I was like, okay. Wow. That's insane.
38:09So people are coming for the viral feature, and then they're staying because the app's helping them keep on track with their weight and stuff like that. So we just added then after we after that, we started adding a ton more adjacent features that are just strictly useful and nonviral.
38:25And then what we saw was our our churn, like that they, like, solved our retention problem. And at the end of the day, the t like, the TLDR is good product and good creatives, and that makes your life so much easier.
38:38So my newest project, it's it's it's really fun working on it. I'm having a blast actually working on it.
38:44And what we've seen is we've seen people actually hit us up to work with us. So they've seen, like, one of our creators' videos.
38:51This guy, 20 k followers, hey. I wanna collab with you. This guy has 1,800,000 followers, and he he hit us up.
38:59We didn't even outreach to him. And it's solely because the product's actually, in my opinion, the product's actually cool and really useful.
39:07Um, we had a creator with, like, 50 k followers make videos for us for free just because he wanted to actually try the app. Like, he gets insane. Uh, we had another creator with 3,000,000 followers who averages, like, 400 k to a million views a video, and we have a four video deal with four stories for $2,500.
39:26That's like like, our CPM is gonna be maybe a dollar or something. Like, you only get these rates and this, like and people hitting you up if your product's actually cool.
39:37So if you can just make a good product and really focus in on that, it makes the whole other side of the distribution problem much easier. And that's basically the TLDR of how to make a successful app or a semi successful app. So a lot of people
39:53hear about these vibe coded mobile apps and they're like, there's no way that people can make thousands of dollars, tens of thousands dollars, hundreds of thousands of dollars, sometimes millions of dollars with vibe coded apps.
40:08But here you are, someone who's nontechnical.
40:11How old are you? Um, I'm 19, just turned.
40:15Exactly. Like, young guy, um, and, you know, no no real background in this, and here you are in the field, like,
40:28proving otherwise. So what do you say to people who who who who say that, like, it's not possible? Yeah.
40:34It's so funny because people, like, really hate on vibe coding. They call it AI slob, this and that. And, like, as as you see in the slideshow, like, I'm actually very big on product as a vibe coder, which is, like, funny enough.
40:45You know what I mean? And it's because at this point with AI models, Opus 4.8, right, especially if you're building apps in Swift too, I saw the difference between Swift and React Native, like, vibe coding as, like, night and day.
40:59And that's why that's why I really love Roark, and I talk very highly about them. I'm also, like, now working with them too because, like, I was, like, the first power user, so they brought me on.
41:10But but with with Swift and with these new AI models, there's no in my opinion, there's no such thing as AI slop.
41:19There is, though, underprompted. Right? If you keep prompting that's why I said fourteen days.
41:24I'm not, like, sitting here and, like you're not sitting there and one shotting these apps. Right? But if you consistently manually tweak each thing, at the end of the day, code is code.
41:34Right? So with enough prompts, you could really make these really cool products. Like, there's some apps built by some of these, like, Rourke users that, like, I downloaded just to, see what they were doing.
41:45And I actually use this one workout app, CoreOS. They they just launched this guy. I actually use this app every day because it's, like, one of the sickest health and wellness apps out there.
41:56It's like it tracks your sleep, tracks your workouts. Like, when you walk in the gym, it auto tracks how much time you spend in the gym. And I'm like, this is the one of the coolest apps I downloaded, and it's fully made by AI.
42:08It's, like, it's, like, insane.
42:12So I'm very big on a I'm very I'm very against AI slop and, like, this, like, doomer look at vibe coding because not only did it change my life, but you can genuinely build really great products with enough prompts.
42:27And, you know, with vibe coded products, like, have you ever encountered scalability issues?
42:35Like, you know, that's another thing that people say is that, like, okay, you're gonna use AI. Maybe it's good enough to get to a prototype with a few thousand users, but once you start scaling and actually getting hundreds of thousands or millions of dollars a year in revenue, you're just gonna have to rebuild the whole thing.
42:49So I haven't hit it yet. Maybe it's because I'm not big enough now, but but I personally haven't seen any scalability issues.
42:59It's also because I'm making a consumer app. So maybe if I was making a social media app, it would be different, and it would there would be, like, stuff like that.
43:08But, like, people always come at me like, oh, what about the security issues? Right? Yeah.
43:13I don't have access to any of your personal data. I'm not collecting that. I wouldn't I'm not putting myself in the risk of that happening to me because I'm not collecting any of the data to put you put my customers at a security risk.
43:26Right? Yeah. So I'm gonna say it honestly depends.
43:31Maybe if you're building a social media app, but personally, I haven't seen any of that.
43:35So in your in your opinion, is this the heyday of building mobile apps? Do you think, like, there's just so much opportunity right now, so many different niches, and, you know, it's a window of time before it just gets too competitive?
43:50Like, how do how do you see
43:52it playing out? A 100%. So this is this is like the in my opinion, this is the drop shipping boom.
43:57This is like the ecom boom. Right? Back in the day, meta ads, you used to boot them up for ecom.
44:02I I was never involved. I actually, like, failed at ecom and then went into apps. But but back in the day, like, 2015, people used to boot up these e comm sites, drop shipping, whatever, threw a little bit of money into med ads and then see, like, four to five x ROAS.
44:17Right? Today with apps, it's very similar. If you make an app and it's again, if it's cool, the idea needs to be cool.
44:24But if you make an app and just start posting about it yourself, this is what sold me on apps. I made an app, had no app experience with fight.ai. I made this app.
44:32I literally would just post myself talking to the camera. Guys, I'm making, like, the future of MMA. You could throw your, like, videos into this app, and it'll tell you what you're doing right.
44:41And at first, I just made it for fun because I wanted an internal tool. I was like, screw it. I'll try making a few TikToks about it.
44:46Right? The fact that I was able to make like almost 2,000 and by the way, it's still consistently to this month does a thousand every month, which is insane. Right?
44:55Because I haven't marketed it since July.
45:00So it's like I love seeing the subscriptions. It, like, makes my day out. Like, I'll see a guy who's been subscribed for, like, nine months, and he was, like, active, like, a week ago.
45:07I'm like, what could this what could this guy possibly be training for? Anyway, so the fact that you could sit there and still make five TikToks and get, like, 10 sales, right, shows that the industry is not saturated even though there are more people making apps than ever.
45:26Um, I don't believe it'll be saturated until three more years probably, and I I'm finding adoption is still pretty slow.
45:34Like, you you get caught up in this, like, Twitter mind share where, like, everyone's building an app. In actuality, no one's building an app. People still think you need to know how to code.
45:43Like, no one knows vibe coding exists. 15% of the world use still uses still hasn't even 15% of the world, like, only uses ChatGPT.
45:52Like, compared like, 85% of the world still doesn't even use AI. They don't know AI can actually build software. So, yeah, get started.
46:01Best time to get started though is today. Like, get on it, like, right away.
46:05And, you know, last question. You know, is there one tip, piece of sauce that you didn't share today that, you know, you wanna share that you think is gonna be valuable for anyone who's interested in in building a mobile app with AI?
46:23One tip.
46:25I mean, you shared you shared around paid ads. You shared around creators. You shared around choosing an idea.
46:31But is there anything else that you haven't shared today that, you know, you you wanna share?
46:35Yeah. Oh, yeah. Really focus on being a like, learning how to sell things.
46:43So all like, my the ability to close influencers came from when I was, like, 16, and I had to do, like, door to door sales for power washing.
46:54Mhmm. Or, like, I was running garage sales. So if if you have good social skills, it does make selling to influencers just so much easier because you could spend like, I'll give you an example of my last influencer call.
47:08The guy was Greek. Right? I spent 90% of the call talking about Greek like, Greece, and, like, me and him were just vibing about both being Greek.
47:16And then at the end I go, okay. So look, your content, your views vary a little bit.
47:22To not gyp you out, why don't we do like a $2 CPM? And since me and him were already kind of like buddies, he now doesn't wanna say no to me. You know what I mean?
47:30So it just makes it 10 times easier to close stuff. So social skills are very important Mhmm.
47:37Which is which is why vibe coding's great because the people who used to build apps didn't really have the craziest, like, social skills.
47:45George, thank you for coming on the podcast, sharing all the sauce with us. Appreciate it.
47:51Can't wait to see what you build next. I'll include links for where to follow George in the show notes in the description.
47:59I'll also include a link to the docs that he he had if people wanna check that out. And, George, thanks again for coming on.
48:07Yeah. Thank you so much.
The Hook
The bait, then the rug-pull.
George Lampropoulos was scanning price tags at TJ Maxx when he first saw the ad for Rork on his phone. A year later he walked out with a playbook that generated $200K in revenue from a wrestling app — and two more apps quietly tracking toward their first $10K months.
How they asked for the click.
MENTIONED ON CAMERA
FROM THE DESCRIPTION
PRIMARY CTAWhere the creator wants you to go next.
