He makes $750 a day 'Vibe Coding' Apps (using Replit, ChatGPT, Upwork)

this episode is unlike any of the other episodes because it is a really approachable way to make money on the internet vibe coding and teaching you how people are doing it by selling AI services on Upwork now there's people who are making 5 10 15 $20,000 a month solo devs or with a small team uh literally one prompting or a few prompting apps that they're finding on Upwork i thought this was really interesting not because I want to go and do Upwork stuff but because I think it's a really good way to sc you know scouring Upwork is a really really interesting way to see what are business problems and what are SAS businesses that you can create so I had uh my friend Billy uh come on the pod and this is a guy who's started and sold multiple Replet apps and just walk you through how to think about making money on the internet with Vibe Coding and how this whole Upwork ecosystem works enjoy the episode and uh I'll see you I'll see you at the end of it [Music] billy Howell on the pod uh Billy's been going viral recently because uh he's been vibe coding his way using Replet and selling those businesses not just one time but twice i I wanted to have him on on the show to explain well what what are we going to learn today i've got it laid out for you right here Greg uh we're gonna find out how to make profit with vibe coding okay there's a lot I got a lot of question marks but by the end of this episode what are some tangible skills that people are going to learn if if they make it all the way through sure uh the number one thing is how to sell an app that you've built uh that's what I get asked all the time is how to sell and how to market and the secret is not to market it's just to sell it so we're going to walk you through that okay so you've got me intrigued i'm I'm I'm That's where I'm at i'm in the intrigue phase let's Let's get right into it so where do we start okay so we start we need to find someone to buy to buy the app before we build it that's where most people get this wrong is they build something and then try to sell it or try to get users but I found it's a lot easier to just find something someone wants to buy and then build it and so if you don't know anyone that wants to buy an app you should go to Upwork that's where I sold my first Replet app is Upwork and so all I did was search for a technology that I knew I could replace with an app um so I use Air Table a lot with past projects and clients for my marketing agency and I was looking at these jobs uh to do Air Table consulting and I'm sitting at my desk looking at this one job description and I'm like I think I could do this better and cheaper than in Air Table because Air Table you have to pay $20 a seat right and so I made a prototype in Replet and then sent a loom to the guy and he's like I like your prototype in Replet better and that's how it went uh so we're going to try to look for something here real quick that we could we could mock up to someone okay yeah let's do it so by the way I must say I've never thought of doing this before building anything like I didn't even think this was an option but it makes so much sense right these are this is valid this is validating you want to build something that people have validated before right and you can you can go and you can search you know custom app development on Upwork or wherever uh but I think it's an easier cell when you're replacing something or doing something better than like existing software could do cool so what are we looking at okay so we're looking at Upwork listings here of people who've posted a job uh that has Air Table in the keywords so we're going to look for something that we can just make an MVP on daily Google ad calls no code expert need for MVP development this is one I see a lot on on Upwork um and it's basically going to be a bunch of integrations mvp that integrates Calendarly Zapier Air Table SMS functionalities not sure if I like that one what do you think i mean I I'm curious your like thought process so what about that one didn't you like well for the purposes of this podcast this is going to be a lot of digging around in uh developer docs to figure out how we get these integrations in a Replet app um Replet might not one-shot it but we could try it i mean just put another way you're trying to prioritize apps that are going to be easy to to push and deploy right you're not You don't want like complex apps yeah usually um we're looking for CRUD apps right something where it's just putting data in a database pulling it visualizing it um you can do all of this with vi with vibe coding uh for sure um it's just going to take you know more tinkering okay cool so yeah let's look uh let's look at maybe another one so this is case management um for a nonprofit so yeah I mean look at this this is perfect can you read it out loud we have a lot of like audio listeners who are not on YouTube so it's called air table base creation specialist needed uh the description is we are looking to have an air table base created for a nonprofit the system will function as a case management and assistance tracking system the database will keep track of clients their case notes any assistance given their household members and appointments there should be an interface design that will show an overview of the business number of active clients amount of utilities paid etc the interface should also allow workers to view client data and input data so they're willing to pay $125 fixed for this project when I look at that I'm kind of like is it worth it what's a good number how do you think about what the pricing of some of these are as an input for if I should build this or not yeah so this is definitely low um you can look at like how much money they've spent and kind of see so maybe these people are looking more for like overseas help um you can always submit a proposal and you know price over what they've put and hope that they bite um on Upwork you pay for like credits to apply like I'll show you down here like connects and you want to boost it so that you're the top one so they see yours it just bumps it up on the hiring side so it's first in their inbox i don't know these are probably like $2 $3 worth of credits um but yeah so for pricing this is too low um for something like this I mean the first one I did I did for 750 bucks uh I think you want to try to get like closer to that you know probably that was uh underpriced a little bit too but then like you can also get recurring revenue uh from clients who want you to add on features in the future charge them to pay for hosting so I charge clients to pay for hosting so it all kind of depends on are you trying to get your foot in the door you know what are you trying to make right yeah i think I guess the other piece of this is this person might be willing to pay $125 but someone else might be willing to pay $12,000 who knows right so I think maybe um people need to do their due diligence around like okay this might not be worth it for 125 but let me see on Upwork if there's other requests for a similar type of thing quick break in the pod to tell you a little bit about Startup Empire so Startup Empire is my private membership where it's a bunch of people like me like you who want to build out their startup ideas now they're looking for content to help accelerate that they're looking for potential co-founders they're looking for uh tutorials from people like me to come in and tell them how do you do email marketing how do you build an audience how do you go viral on Twitter all these different things that's exactly what Startup Empire is and it's for people who want to start a startup but are looking for ideas or it's for people who have a startup but just they're not seeing the traction uh that they need so you can check out the link to startupmpire. co in the description yeah and so you can just do fixed price and do higher automations is also a good keyword to look for because you can turn most automations into an app with a front end this one is talking about growing a tutoring business they need a custom workflow for parent parent inquiries yeah this could be good because they they they need they're pretty much trying to cobble together Google forms air table type form calendarly a bunch of stuff uh so you could make this into an app for sure and they want to pay 2500 bucks we'll take it we'll take that 2500 bucks is this Is this not too complicated or do you think we can do it we can definitely do at least you know the first part like the data entry on the call okay all right let's do it 2500 bucks let's do it yeah yeah and so at this point um if it's a simple enough app and it's 2500 bucks right uh you could go ahead and build the prototype and then in the proposal send it as a loom link of you like demoing it for them which I've done and it wor it worked for the for the first one I ever did otherwise I would just talk to them I would pull open like my air table my replet and be like I'm experienced with air table uh we could build it out in here or we could do it in replet and so replace air table with you know whatever SAS you're honing in on if it's air table or HubSpot or Google Sheets or whatever um so let's just try to prototype this real quick i'm going to take this whole description okay and we're going to cheat a little how are we going to cheat a little i like to feed it into chat GPT and just tell it like can you format these requirements as a PRD for an AI coding assistant be agnostic when it comes to tech stack but make sure to use shad CN okay so PRD is a product product requirements document is that right that is right yeah um it's basically a form that tells an engineer like here's all the features you're going to need uh here's like the the the data types and and everything lays it out so we're going to feed this into replet uh to see what we can get and then what was the other thing I said oh be a be agnostic about stack uh I'm pretty much saying like don't be opinionated don't say we need to use you know vite or you know this kind of web framework just let Replet kind of do its thing or whatever you use you know and you use it's you know you mentioned Replet a bunch of times like why are you using we've had Amjad on the the CEO of Replet on the pod before um you know why using Replet versus the the other platforms yeah yeah so uh I haven't tried Lovable uh but I've used Bolt i would say those are two comparable ones um I I like Replet because I think you can get from idea to MVP with the least amount of friction you don't have to worry about package management uh installing stuff does all of that and then you can deploy really easily and test um of course like I still use cursor and windsurf to edit if I have if I get stuck in a loop fixing something or if it's something really complex I'll pull the code down into Windsor for cursor and prompt there cool that's helpful all right so what are we looking at right here so this is the PRD that just failed to copy this is Yes so this is I guess the new chat GPT does like a split view for docs like this inquiry form which sends details to our Gmail inbox yeah so it's laying out all the features like we said just kind of translating it maybe so that an engineer or an AI engineer can understand it better so let's go into replet we're going to paste that in and I'm just going to say build me this app um use shad CN where you can for UI that's just the library I like of pre-built React components you don't have to say that and I think Replet might use it by default anyways um the other little hack is if you want a specific if you want to upgrade the UI like say you've got all the app logic done I go into VO and just describe what the page does um and then I pull that code so like an app screen for managing parent student inquiries you would obviously kind of lay it out more detailed than that but you can just pull the front end code from VO uh and it'll make you a mockup really quick uh this is really good especially if a client has sent you like if they sent you like a drawing like a wireframe you can just pop it into VO get that screen exactly um then in this step like you could just I'm not going to use this but you could also paste this in and tell replet I have the front end code for like this screen so don't worry about it and it works pretty well okay so let's start building and see what happens let's start building so this is Replet V2 it's going to go back and forth a little bit with you before it starts i think it's going to ask me some questions on features and it uses cloud 3. 7 uh what do you think of Cloud 3.

7 have you used it yeah I mean I use it frankly I use it more like less for coding and and more for writing basically um but I mean it's a step in the right direction you know we also used it to build a an a game um where uh that's pretty cool like one prompting a game so I guess I have you know I have used data it's like I don't know i'm just numb to everything now i'm like oh yeah I even forgot oh yeah one prompted a game that's crazy right like deployed game like fully deployed like insane if you would have told me that you know I went to CS school if you would have told me that you know how many years ago now 15 years ago that you'd be able to one prompt PRD design frontend code back you know backend deployed in like under a few minutes I would have told you you're crazy i think I would have said the same thing but because I wouldn't have understood any of those words because I studied public policy so you study So you're not even a developer by trade no just kind of a self-taught uh script kitty you know yeah for sure okay so it says Yep what is happening on the screen all right so the replet agents read the PRD and kind of thought about it build the initial prototype i don't know what would happen if you unchecked this uh but it asked do you want any of these additional features um we definitely want a Postgress database that's going to be the the where all the user data is stored um let's I mean this is kind of a wish list and I would say in the past with most AI this kind of feature where you have a wish list um distracts the AI uh but I think replet and cloud 3. 7 are so good now it helps it really helps kind of set up your app to bolt these things on later so I'm going to check the ones that we want um let's leave off Docu Sign maybe that was in the PRD though let's just see what happens and then I am not going to deal with payment processor right now uh because that could get messy so we're going to hit approve plan and start and this is why I mean tools tools like replet and you know the other ones like bolt and lovable are great because like you you know that after you do that prompt it's always going to make you something that runs so you always have a starting starting point so like you just get past that initial like writer's block um which is a crazy way to think about code and you know all the CS guys are probably cringing hearing me describe it like that but what can you say like it's vibe coding man that's exactly what you could say and watch them cringe as you say it yeah yeah i get a lot of people uh reply replying to my tweets that are just saying like like this app's going to break like this is AI slop this guy or you know he's just making this up is there any truth to that no no so these you know this product this app for example like you feel comfortable that this can go and see I mean we haven't finished it yet but that it could see the light of day that someone can use and it can be valuable in their business definitely yeah it it definitely takes I mean you know if you have no idea how code is working you probably shouldn't be putting out a production app right like it it does take some TLC to make sure to to look out for things that you know would break once you put it in production right like you you can anticipate what how users are going to break the app right um but so yeah you just kind of have to have the foresight for stuff like that um but otherwise yeah these are pretty solid apps they're just so simple they're crud apps you know right yeah um it's when you start adding more complex features that you could get in a little bit of trouble right i saw like the docy sign integration yep um you know payments that gets a you know a bit more complicated what else can get more complicated like what should people look out for yeah uh payments definitely docuign uh anything to do with a calendar AI in general gets kind of mixed up with date formatting in time zones okay so that's something to look out for for sure um but then on the other hand like I I made a I I've skimmed the HubSpot docs and kind of copied and pasted in what I thought we would need into Replet and it it set up OOTH with Replet like flawlessly like I was dumbruck um so it is pretty capable at at integrating other services um especially things that are just like kind of basic API like git calls um if you're doing anything with like posts where you're changing data and other apps or servers that could get really dicey you want to make sure you're like locked down there uh what I do now um that this is like a real agency is I kind of have uh a dev that specializes in this kind of stuff go through and do like the last 15% and kick the tires right um so that's and you can hire people straight off of Replet they have bounties and stuff to uh to do that kind of thing just like hourly or for a fee um so that's definitely a good idea while that's loading can you I I've never done a bounty on Replet heard I've heard about it can you just show show us where we'd find that and what that looks like it's just right here in your dash um let's see i actually haven't posted one yet okay so this is crazy so for like 500 bucks someone could get a custom tool an automated system $1,500 you're getting a web application but more interesting to me and I think a lot of people listening the get help button right like the fact that you can take it to 90% 80% vibe coding sometimes something's not working and it's really frustrating and you're sitting there talking to your girlfriend or whatever your wife your husband and you're like I'm so close i'm so close yet so far the get help button so what happens if you click get help okay you can give the bounty a name uh when you want it done by describe the problem and then use you have to buy credits I think yeah like it seems like they're called cycles which is like a coin or whatever yeah does Replet have its own crypto right like they're a memecoin um yeah so this is something I definitely need to be more disciplined about is just like 50 bucks to troubleshoot something is such a good use of money versus me banging my head against the wall burning $10 and in AI credits and six hours of my night you know totally and especially if you're useful like this is your business now right like Yeah so if you're going in you're you're going onto Upwork and you're trying to you know make this a business but by the way like how before we go back to the the replet screen like how much how much are people how much is the most you can money you can make on on uh Upwork like the top people of Upwork like the Michael Jordan the LeBron James oh dude of Upwork what are we talking about six figures easy i know I know one person that does over six figures he is but he has an agency team right to support him doing SEO um but you can go on his profile yep are we talking Oh you can go on his profile it doesn't say how much you make right it says how much dollars of jobs you've completed oh okay i'm curious if you know when you say six figures is it like a $100,000 is it $900,000 you know like if you go look as if I switched to client $200,000 earned $600,000 earned and it's like $12 an hour making $600,000 that's crazy right but not to I'm not picking on this guy but a lot of these are kind of like it's 12 $12 to get in the door and then maybe they upsell you or give it to their team right um I would say like solo dooo um I think you could easily clear 5 10k a month just without any outside help um definitely building apps but even just like if you consult on the right kind of pro projects surrounding like a specific SAS cool all right should we go back to see our replet let's see how we're doing here let's see what we've done building the initial prototype so real quick in Replet um this is the agent which kind of acts autonomously and right now he's going through and scaffolding out the whole app we can see it here in the files on the left um this is the client this is what the user sees he's building out components he I Isn't that weird that I say he i was going to call you out on that i didn't want to get you cancelled but is there a particular reason why you call called Reflet He no I don't know is that Yeah I don't know what I should say replet's building yeah we can move on from that the comment section is gonna is gonna is going to have a field day with that one uh yeah anyways you you can see all the files here the one thing is you got to make sure they initialize the database i've noticed lately sometimes the replet doesn't initialize the database right away um but he's still going so is he building a screen we can see yeah okay so we can see this dashboard so like a new thing in Replet V2 that they've added is it it it it scaffolds like a UI that doesn't do anything like line by line uh when you start building just to kind of wow you and like it blows my mind every time like it usually looks sick see if I can pop this out no not letting me okay and it looks like it's doing like a mobile app design which is interesting i'm curious why it's doing that so maybe it was in the PRD or maybe it's just taking some liberty i think maybe it is i mean I'm not against it i'm just It's just genuine curiosity yeah I've never had it do that interesting so you So that was probably my fault um I think usually I do specify web app and lately I've just been not doing that because usually it knows either way i mean we can interact with it in the web and you can see I've got so many so we can see where I started like four months ago i made a captain's log app that you just talk into and uses AI to like summarize your day and since then I made like 40 more it's crazy it's crazy and like Yeah and we'll talk at the end once this is live i want to talk about the process for selling right it doesn't have to be Upwork either um you know so my my best projects best projects are always referrals so people come in and just be like "Hey can you help me do XYZ with this client or fix this internal tooling? " So I think what you can do if you're like an entrepreneur trying to do this is like build like one one or two apps like probably more than that like get good at it like I made like 40 and then um just call business owners and be like what's the most annoying thing you do or like what's the most expensive SAS you use and then make them an app right and be like and try to sell it to them right and you know you decided to I mean we're talking about basically creating basically doing consulting this is essentially consulting you're doing consulting you're just using AI to do consulting and building but you're doing consulting um vibe consulting um vibe coding and you know but you've decided not to go and build a SAS and sell it to people right nightmare money pit so I want you to I want you to talk about um because people are I want you to talk about that because people are probably going to there's going to be a lot of people listening to this and be like this is this is small boy stuff you know this is a few hundred here a few thousand dollar there the real opportunity is building multi-million dollar a year SAS businesses what do you say to those people it's a nightmare and it's a money pit you're going to spend ton of money on marketing um I just don't I get that it's an asymmetric bet but like you got to have the juice to do that right yeah you have to have the you have to the reality is you have to know how to juice right like understanding distribution is a completely different skill set than for example building so if you're one of those people that you think you can build audiences you can build communities and you're willing to invest in that in yourself but also you know the time that it's going to take right building an audience is going to take time um right you're seeing this now yeah and the way my brain works too maybe yours is like this with doing kind of multiple different types of projects is I think if you're you know betting it all on one like app or use case you get so much tunnel vision and you lose out on the feeling of finishing other apps uh learning other libraries APIs like you're missing out on a lot of learning serendipity uh you know just surface area of luck so like the app that I just sold most recently uh I I sold to a consultant for a big company uh to the big company but she's like "This is amazing uh let's turn around and sell this to company B and C because they have the same exact problem they're in the same industry and I consult with them.

" Um so that's awesome you can do that if you find a problem find a I think this is a more direct approach to finding like a burning hot problem uh that you can then resolve to other people yeah i And and that's what excites me i got to be real with you Billy like what excites me about this whole approach around going to through Upwork finding problems vibe coding your way through it is it gives me ideas around building something that validated that I can go ahead and sell that to other people who have the same exact problem right um so that's that's what excites it for me cuz I'm I understand how to build distribution and and and how to sell frankly sometimes it's not even about distribution it's like how to how to take a product and figure out how to go to market um so yeah that's what excites me about this whole approach yeah yeah and I it's just it's a different lane B2B versus B toC you know i I find it's better to work with business owners i think they're more pleasant than you know a bunch of people who are going to be saying like "Your app broke and you need to do this and blah blah blah. " There's one source of truth if you're building apps for a business owner and it's the business owner right well it's a different So you can still build a SAS app and target business owners you know what I mean so you could still do I'm thinking like mobile viral app you're right totally viral don't do that and and you can you can build a a SAS app and literally only have 10 clients right but maybe it's 10 clients who pay you $1,000 a month yep um and and that's a that's an you know a whole other approach right so I think that when people think about building software they think about maybe it's you know the the popular example right now is like Calai um and I've had those guys on the podcast they're doing $30 million a year so they vibe coded this and basically you know great concept you know pancakes you take a picture of it how many calories right and then they Wait wait wait can I stop you great concept pancakes people like pancakes people don't like calories oh this is sick right so uh really smart and they've got like a few of these apps where they're you know again these aren't trained developers these are guys um who are like in in their they're teenagers and early 20s who kind of vibe coded their way into into this $30 million a year in revenue good yeah with a few of their apps now this to me is really cool yeah um but I think that this because it's such a distribution challenge is way harder than maybe building a SAS app that targets like the easy here okay think of it as like a framework right the easiest thing you can do is get paid for a service that's easy mode then like the medium thing you can do is build a software for B2B yep and then the hard mode is like building software like this for B TOC that's kind of like how I how I picture it i I think you're right uh I think it just depends on how you define like hard right you know how I defend how I define it is you know likelihood of success like that's how I you know and and likelihood of success with risk with risk right like it's way more risky um to go build a consumer social app that's why I like this approach that I'm doing because I can derisk by getting a price up front right yeah but you don't you derisk but you don't get the you you there's there's you know this business is worth a lot of money you know what I mean so um pros and cons with each and for everyone for the listener just think about what excites you does building the Calai excite you does building like a SAS app focused on uh a thousands thousands of customers excite you or 10 customers or does doing Upwork MVPs banging it out and doing you know between a hundred,000 and a million dollars a year of revenue um does that excite you with the small team so everyone needs to figure that out for themselves i I also just like unbundling existing SAS i mean I think that's going to be such a huge market soon uh I think you're going to be able to put in like you know hrefs. com and then have an AI build you an HRES clone using like a an API and it's going to be you know so much cheaper um I think that's an interesting way to find ideas is just looking at a list of SAS or like existing APIs and building apps for them yeah i mean so go to HS for a sec and then we'll go back to Replet so this is a business that does hundreds of millions of dollars a year I think in revenue and it's been around forever right forever oh yeah uh for years and years are you saying basically there's going to be an opportunity to essentially duplicate this um oh it sounds like what you're saying is there's two two opportunities one is you can duplicate this and make it cheaper or free and the second thing is when you say unbundle I think what you mean is target it for a specific niche is that correct so Hrefs for doctors just you know no no actually um I'm not taking a shot at at hrefs either uh which is struggling to to show me anything uh but Hrefs is a it's an SEO uh kind of tool and all their proprietary data is great uh but there's also public APIs that you can just pay to get SEO data from and so what I mean is like I think in the not so far future you'll be able to paste a SAS into like I don't know you could do it with chat GPT operator maybe um and have it look at the SAS and reverse engineer it and use a public API to to build it for you right got it okay cool does that make sense it does it does um should we should we head back to our our our baby yeah okay here we go so I'm going to preview this full screen wow wow wow wow yeah i mean look how good this Oh dude and it's it's mobile optimized that's why it was showing that oh oh man this is incredible like truly yeah ah so we've built this app okay we're seeing I mean this looks like it works like um but let's just double check what do you think an app like this is worth couple thousand bucks yeah definitely a and to kind of add on to what we were talking about earlier like the one of the advantages of of doing this is that you can charge for like this is your monster that you've made so then you can charge for ongoing support and feature requests and be kind of like a dev agency uh for your for your clients right i it's not like a novel concept but um like I don't know it worked it worked um here we go so we got a recent inquiry here scheduling wow whoa okay so it looks like it doesn't have Calendarly like it looks like this is a placeholder so this is a tutor this is going to be a user add in this is so funny this is so similar to one of my real life clients and they need this but the thing that is kind of stopping us from building this is data privacy laws for students we have to figure out how we can be like Uber compliant um otherwise everything is like Air Table and Google Sheets love it well this is this is awesome yeah and then what I would do is like I said if you got time to kill like if time isn't your constraining factor make it an MVP i I I kind of like I think you could just make like a It doesn't even have to do anything just make a mockup like landing page or dashboard you could even just do it in VO that would be even quicker uh as we saw here dude look how good this looks so this one was done in like thought for six seconds this is probably like a minute so what I would do is find the Upwork job make a mockup in VO record a one minute loom that says "I'm Billy i make apps i know you wanted it in Air Table but I made it custom flatter them by putting their like company name if you have it in the top left.