VIDEO_ID: 8gV6xsof964
URL: https://www.youtube.com/watch?v=8gV6xsof964
FETCHED: 2026-02-23
CHARS: 65009
================================================================================
We just launched an AI agent that hit a million dollars in just a few weeks. But the crazy part is that it only [music] took us 14 days to build. And of course, we recorded the entire process from start to finish. So starting first on day one when my co-founder Vitali and I had absolutely no idea how to build an AI agent, what idea we were going to even build, all the way through to building that AI agent without writing a single line of code. So we vioded the whole thing and then launching that to our first customers without having to spend a single dollar. And so my hope is that by watching this video, you can get the entire playbook of how to launch your business from [music] scratch and get to a million dollars in revenue. But you'll also see all the ups and downs that we go through through this entrepreneur process and the [music] doubts and the emotional toils and the eventual wins that I think will really prepare your mindset to [music] make sure that you win. So steal this playbook because we made every mistake under the sun. And I really feel like if you take all of the tactics and learnings that we share in this video, you will be so much more successful than we were on a shorter time period with way less toil. So, let's get into it. >> Let's [ __ ] go. >> Let's talk about our game plan because we have none. But what game plan? Think about this. Just 14 [snorts] days prior, we were just two bumbling idiots who showed up in London and were like, "We have this idea. We think that there is a multi-billion dollar opportunity out there if we can help [music] anyone actually create more content for themselves so they can build their own brand, build their own audience, build their own business, like basically go out there and build their own dream. But the challenge of that is is how do you build an AI agent that can actually help you [music] do that? So that's what we're going to try to figure out over the next two weeks here. Let's talk about game plan for the next two weeks. No matter how good Vital and I might might be at hacking something away, we just fundamentally can't build something that complex with that many nuances and that many different customer segments all in just two weeks. And so the first thing we need to do is figure out what is a 2 week MVP scope that we can build that will validate whether or not we have the right hypothesis of a [music] multi-billion dollar product. You actually break it down into a bunch of different subsets of problems. So whether it's hey do you need help with content ideation or do you need help with the actual content creation of making the carousels for you and autogenerating those. So we could understand out of this whole greater vision we have which subset of the problem should we actually go after first before we go out and horizontally attack the rest of them. And so in order to achieve this huge vision we have, we've actually decided to scope down the problem into a subset of a problem that we believe is a really solid microcosm test to validate whether or not we can actually do the bigger picture, which is to start with something that's going to seem really initially unsexy. And that fatality is to start on what platform? >> On LinkedIn with content ideation. >> Sure, it has incredibly cringy content on it, but the content itself is really simple, right? It's just text and maybe some photos and carousels. [music] That's it. Whereas, if we were to try to understand and analyze a 52cond Instagram reel or a 45minute YouTube video, that would essentially be an impossible task in just 2 weeks. And so, now that we have an idea and a vision established, we now need to actually build that idea. And Vital and I have already told you guys that we've literally never built an AI agent before. So, before we even get to the build stage, we first have to figure out how to build an AI agent, which is going to take us how many days with Taly? unknown. [laughter] >> Yeah, we're going to figure it out. >> But no more than five. >> Yeah, no more than 5 days or else you're screwed. So, let's get started. >> All right, so Vital, you've made some good progress. >> I figured out how to scrape the entire history of creator post since the beginning of time. I was testing [music] it on myself. My first post on LinkedIn was in 2016, and in total, I posted 303 times. [music] So, that piece is working. The most productive way for us to to work through this [music] is to send an email to a creator that is going to impress them. So, we talked about before that that email is going to be like a very comprehensive in-depth analysis of their five best posts that will help a creator understand [music] why those post did so well at a deeper level than they've realized themselves. [music] And what's clever about this game plan is that it allows you to tackle two different challenges at once. It allows you to both build [music] your product and acquire customers at the same time. Because if you think about it, the person who is reaching out right now to our customer prospects is Stanley. [music] And in order for Stanley to be able to do that and send those emails, you have to build and train a model that actually makes good content ideas. And then [music] from there, even though we're sending a really simple basic HTML email, we're able to bring people in to try the full product where we can then observe how they use Stanley, talk to them to get their feedback, and then use that feedback to improve our baseline Stanley so that the next time he sends out another wave of code emails, his hit rate and his conversion rate will be even better. Because the important thing for you to note here is that you should not wait for perfection. That's actually one of the biggest mistakes first- time entrepreneurs make is they wait to have something that's absolutely [music] perfect in their eyes rather than iterating really quickly on these 24-hour cycles where we would build a feature, get it live into a customer's hands as soon as possible, get their feedback, and then iterate and ship out an even better feature the next day. So, that night, we got to work. All right, turning on the lights here. 11:55 on day two. How did we do? What did we accomplish? >> I feel so sleepy right now. I don't know. [laughter] >> But I think we built an AI agent and we made a promise. That's for sure. >> All right, [gasps] day three. How are you feeling? >> I'm feeling pumped. I'm going to get Chad to a place where today is tomorrow. My hope is that we don't need to touch it anymore. So, we solved the problem of giving you really exceptional content ideas, but somehow in solving that problem, we somehow made it really freaking bad at normal conversation. Everything just started to become interpreted through like this is your content idea, like go post it now. And the crazy part is that Vitalia had built all of this so far without writing a single line of code. He had vibe coded it. >> But what am I doing other than eating a sandwich? This little toasty as the Brits call. I got to figure out what the product feature is. So, it's going to drive product growth where people are going to be so delighted or surprised by how quality or valuable this bot is that they screenshot it. So, like what is the screenshot? It's going to create the product led growth motion. All right. So, we made great progress today where before when we were chatting with Stanley when we were asked it at post ideas oftent times it would say things like post number one share a journey story or share a story about how you started Stan. All day we were showing with our AI being so generic just saying like post this broad idea or post that broad thing that's really really boring and unhelpful and uninsightful. But after grinding all day, we finally had a breakthrough where now instead of suggesting [music] broad generic ideas like share a customer story or share your entrepreneurial journey, [music] instead it's saying things like, "John, share your tactical playbook of how you scaled Stan from 0 to $10 million [music] in just a year and a half." Where if you see the difference, this side is so much more visceral and so much more valuable and unique and powerful. It just like pulls you in so much more than this highle broad generic fluff. What our agent now does is it pulls from your unique story and everything it knows about you to give you really actionable ideas that you want to post immediately. I'm going to take this small little win we've had today because I've been frustrated all day with not making enough progress. Uh but here we are on step three of a million of getting to the journey of making a great AI agent. [laughter] >> Yes, step three. >> Sweet. Thanks for telling. >> Sorry. Right back at you guys. >> But of course that success and progress is always shortlived in the world of entrepreneurship which is on day four. We hit another wall. And so Vital and I end up going on a bus ride through London on one of the iconic London buses. All right, let's talk about all the problems and frustrations [music] we're facing right now. We built an agent that is doing a decent job. What it doesn't do a good job at is keeping track of the conversation and [music] answering questions that are actually being asked. What we realized was that we were so myopically focused on just one specific small part of the problem that we didn't zoom out enough and really think [music] about how every problem you solve is actually an interconnected and related chain. And so what we needed [music] to do was actually to zoom out and think about Stanley on a more systemsbased level and think about what were the foundational prompts that we give Stanley at this existential metahuman level that would then trickle down into how he specifically gave both content ideas but also specifically how he had conversation with you about your content strategy or even how your day is going rather than if you're talking about the rather him randomly forcing a completely unrelated content idea down your throat. You know what's super interesting is I'm currently using AI to improve itself. It was really struggling if you see here to figure out how to actually give the right answer that I was looking for. [music] And so I just started asking it to self-examine itself and is super self-aware [music] in a really meta way. And it's like I'm sorry for the oversight. You're right. Actually to improve the backend prompt that you provided me. I actually then ask it to give me the tactical prompt [music] that will actually improve this prompt going forward for it. it so we can actually give our users the most valuable response possible. And out comes Oh, wow. This is way better. So, the first time around, you'll see from the old post that it was just saying a couple recommendations that were pretty broad. But this time around now, it's actually analyzing my recent post and my top posts of all time and then giving me insights from it. And I already see like two or three ways I can already improve this prompt even more. This feels like the agent has done way more work on my behalf and therefore I respect it more, which is an interesting insight as you guys build your agents as well. So, I'm going to keep refining this and we're gonna see [music] if we can get this to be even better. So, just to be clear, we're eating Korean food with forks and all this Asian food with forks last few days because this place doesn't have chopsticks. I personally feel offended. Oh [ __ ] It's not that we don't know to use chopsticks. I want to be exceptionally clear. Stoked. With that being said, day four. We're finally feeling like we're making real progress. So, basically, we shipped something that looks like decent. Matali told me earlier today that he has not written a single line of code himself building this whole thing. He's voded the entire thing, which is a crazy mental model shift. And then now that it looks kind of pretty, now we actually have to make it actually worthwhile to use. And so we basically the part that I was owning more [music] was around making sure that the conversation itself which is the product in this case right is the actual is it's crazy thing is the conversation itself >> conversation the product also not >> yeah right it's it's a crazy mental model shift and hopefully if like there's one takeaway you guys have from this it's like how do I uple the conversation with my agent building AI agents requires a very different mental model of what you describe [music] as valuable versus a conventional SAS or software product in the sense that in the age of software, often times you're thinking about specific features or a specific visual UI that's going to blow someone's socks off. Whereas with an AI agent or a conversationbased interface, the innovation isn't so much in delivering the feature or the value, which you obviously should, the outcome, but it's also in the human experience of working with the agent. Like what is the quality of that conversation? How do those words on a page [music] or the words that are spoken to you make you feel? And this mental model shift, I think, is going to be really key to exceptional product design in the future of AI. We've started to get there in terms of making the conversation useful to people enough where they'd want to potentially come back. And so that means, drum roll please, what is our goal tomorrow? It's a pretty big deal. >> Well, we're going to get some people, real people, to use it. >> Yeah, we're going to test out our first five beta testers. We have a couple more things that we got to work on. um like I got to make the conversation a little bit more interesting in a couple different ways, but tomorrow you're going to see how we very scrappily growth hack our way to our first five customers. You're going to realize that building an MVP in a startup in the beginning is truly like Wizard of Oz. Like you've got the core value prop that you're testing and then you're going to see everything that we've got planned to like make it work in the meantime. >> John was going to masquerade Stanley. >> Yes. >> Carry out a few conversations like hey why is your agent so slow? I was like draw fast. [laughter] >> But it's to test like the core value prop. The key learning that you should have as you succeed over time is like hey I can actually build anything. It's just about what I actually [music] focus my time and energy on is the true test of I think someone who's able to get really strategic and really successful long term. All right. So let's talk about what we're doing for acquisition strategies because there's two specific things we're doing that people are always like, "Oh, John, how do I grow my business?" These are two specific ways that are really scrappy that don't require any money. Just requires your hustle and your ingenuity. So, I want to share with everyone what we've actually been working on today around acquisition strategies. >> Well, the one and the most important one is building public. >> Yes, >> people got to know what you're doing. >> Before we even knew exactly what we were building, we started sharing our journey. In all of our posts where we shared both the ups and the downs, we started getting more and more comments from people saying, "This looks awesome. I'm super bought into this journey. I would love to be one of your first customers when you launch." And through this process, you start to realize that you can flip the old conventional outbound model where you're begging people to use your product and believe in you into having dozens or hundreds or thousands of people like truly come knock on the door to say, "Hey, I want to build with you. I'm super passionate about what you guys are doing and I'd love to contribute in [music] some way." And this here alludes to one of the reasons why personal brand is so valuable in today's day and age. And one of the key reasons why we started building Stanley because we've recognized ourselves in building a personal brand, how valuable it's been for our business as a differentiated distribution channel. And if we can build an agent that can give that to as many people as possible to democratize that access, well, mission successful. To be clear, it's a vulnerable act to do that as well, right? To be like, "Hey guys, I want to do this thing, but I have no idea what I'm doing, but I'm figuring out I have this broad direction." And then to refine that over time. I mean, it's really hard also to build the self-awareness around, okay, I'm trying to figure this thing out. Also, I have to make content on top of it. So, it's a lot to take on, but if you can master each one of these fundamental building blocks, you're basically seeing how we're able to while building our product also start to generate demand. >> And another thing that I'm thinking about is like, can you imagine the fear of like failing in public? >> Oh, yeah. >> If you do this whole content series and you tell this entire story and then you're on the on the other end of the like, guys, [laughter] you know what? We can't do this. That's going to be devastating. >> No, it's it's it's an accountability thing, right? Like we have two forms of accountability and making sure we actually get this done. One is building in public. Like we're just not going to fundamentally fail given that public commitment. And the other piece that I think is more important to us is like our team is up back there cranking on like OG stand. Our promise to them was we were going to come back with really clear insights on how we're going to bring AI back into our [music] company and turn our company into an AI agent first business uh in order to really help our customers get to the [music] next level. So, all that being said, that's actually just acquisition strategy number one. The second strategy actually builds off of something, if you guys have been following my content for a while, which is my fundamental belief that if you learn how to send a really good cold email, that skill can truly change your life. And I really want to highlight this because a good cold email will get you into rooms that you have absolutely no business being in. Whether it was Mark Cuban responding to me, it's helped me land top investors and it also helped me get my first 100 customers. So, with that being said, these are two really scrappy acquisition ideas that we'll see whether or not play out over the next few days. >> I'm really honestly with my kick lately. I get it now, dude. I get why people loved this band back in the day. That particular guitar solo shakes me. All right, so we have officially sent out our first five cold emails slash product hooks to our five potential customers using our officially shipped MVP. >> I thought you said officially shipped. >> And so basically what our clever hook idea is to see if we get any responses or anyone to engage is we picked five of the biggest LinkedIn creators out there. Justin Welsh, Larara Costa, [music] Chase Diamond, Brenda Gahan, and then Steven Bartlett. And what we've done basically is we've taken out the outputs of Stanley here and we sent them a cold email where we analyze their top five LinkedIn posts and then write this whole email here that analyze not only why those posted well, then summarize those into key patterns, which then we then translated into actually sharing the top three specific post ideas that they could do in the future based off of all that prior data. And so our thesis is, and we'll see if anyone responds, is that if we can provide meaningful data in a way that people will actually look forward to. So I would argue seeing a reminder of your top five post is actually going to be something that is uh really, really valuable. And if it is, can we translate that into something that's even more valuable, which is future content, cuz you're only ever as good as your last post, unfortunately, as a creator. And then from there, if it is actually that interesting, will they either respond to us? So they will they give us feedback, a 1 to 10 score on how how helpful this was? Uh because if they do that, then we can actually calibrate, well, why was that response actually good? And then even better yet, because we're trying to drive traffic [music] to the actual Stanley web agent, we say, "Want to prompt me more? Ask me for five more content ideas." And so we actually just put up a tool [music] to capture user sessions and we'll see if anyone actually clicks through and actually messages their specific Stanley agent to actually get more ideas. That is everything we're going to do today to capture customers and test the value prop of Stanley [music] if it's even good enough or resonates at all. It's a Sunday morning though and so we are actually going to take a fun little break and do two very British things. [music] All right, Vitali, you just said something pretty impactful, profound. What if we figure out managed to figure out how to do this? >> Oh, we're going to do the hardest thing that is there in building companies. You already done one of that to build something remarkable. 30 million business like is not a small thing but building something from scratch is one thing changing something you've built once you're there with your two people is completely different basically in order to build this broader agent we in some ways have to let go of how tightly we held on to OG stands and instead pivot everything to AI and so what we have to do is keep our core business alive because it's literally [music] what pays the bills and use that to fund the replacement of that core business where we have this $30 million business today stands store that is the simplest and easiest way for anyone to start their own online business. But we also recognize that there is an even larger opportunity to help our current customers even more as well as help even more people specifically by helping them build their personal brand and make content which often times is the [music] first thing you need to do in order to even be able to monetize on the internet in the first place. And so we have this ultimate vision where we can build one all-in-one agent that just delivers you the outcome of entrepreneurial success where he helps you make content, builds your personal brand, and helps you build your business all in one place. But here's the hard part. To actually do that and go out and build that broader vision, we have to go back to our team and completely restructure how we think and how we do things. Because right now, 100% of our company is focused just on core stance. right now. We have to allocate so many of our team members and so much of our resources and our capital to investing in these new places while also keeping our core business alive. >> Um, okay. So guys, we uh we're starting to make a little magic. >> So this is the day that we finally cracked it. >> I'm trying to confuse it on purpose and see where it breaks. So first um I'm asking what does it know about me? And it gives a comprehensive summary of where I'm at. >> You're the co-founder Stan. You have all this, here's the data, but then it gives you a full analysis of you as a person and what you're good at in your content and also the patterns that lead to performance and how specifically you should be growing, which is pretty sick. And it quotes a bunch of numbers that is actually not something that I've said, but it's like aggregated from all the posts. Like look at this thing. You're on a 14-day sprint in London to build a million-doll agent, a bold [music] public experiment that's generating significant engagement, 170 to 500 reactions per post, which is like sick. >> It did its own analysis. So, but then the like I want to confuse it. I don't want it to continue the conversation. I want to kind of break the flow like, "Hey, tell me about cars." And then it says like, "Hey, I can certainly talk about cars. However, I'm specifically designed to help you succeed as a content creator on LinkedIn." So, it knows what's up and it's not kind of getting getting distracted by this. I was like, "Okay, I'm going to persist with the cars, generate post ideas about cars that's relevant to my audience." And check out this idea. This is pretty neat. >> They're very they're very solid. >> Yeah. The first one is the Tesla approach to building my AI agent. Just like Elen shipped a barely working roadster first, we are launching our alpha with a duct tape and dreams, >> which is like a really good premise, right? It's really solid. >> It's really solid and it's like it knows what I'm up to. It knows what the world is up to and it finds a way to blend these two together and there's four more ideas and then I'm saying okay. >> But also even it calls back, right? So we were just laughing about this. So, Vitali made a post years ago about his fivestar Uber driver that had 5,000 fivestar ratings and no like nothing lower. And that's crazy that it pulled that back up out of the memory and then brought that back as a theme. I mean, that's insane. >> Yeah. Like it really linked it to to the cars. Like now that I'm thinking about like Uber is about cars, right? So, it just [music] really context matches so well. >> The $30 million ARR engine which is it's it's our business specifically. We're making $30 million of ARR and then knowing Formula 1 is something [music] that's hot and trending right now. It's like so good. It's This is so quality. But the best part is this part here, which is we're starting to crack the real nut around what truly truly matters that no one else has really cracked, which is, drum roll please, Vitali. Okay, so check this out. The hook, the Tesla approach to building my AI agent. Day eight of 14, building a million-doll AI agent in London. Just realized I'm following Elen's playbook, >> which is fundamentally a great hook. So, first of all, it's noticing that Vitali's done seven out of 14 posts up until this moment. And so, it just decided to write the next post in the series and then the hook of just realized I'm following Elon's playbook like Elon is so controversial as a as a concept and a person like and then playbook like that is actually genuinely a good hook. >> I'm like lowkey I I might post it tomorrow. >> You might you should post it. I I genuinely think you should post it tomorrow and then the next day you post how actually yesterday's post that was one of my ideally highest performing was actually completely written by our AI agent. >> Yeah, that would be the best the best market. >> I mean that's a true test of how much we believe in it. >> But the the coolest thing about this that I'm absolutely loving is that it drafted this post and it closed it. Check this out. Check this out. It closed it with PS. John just asked why I'm writing about cars at 3:00 a.m. Fair question. Mexico is so good. Insane. Insane. Like guys, we've been like working on this prompt for now like six or seven days, trying to understand how to build the agent in the right way. And like this, this is the first good thing that I've seen that I've been impressed by. As in, we're finally starting to nail the really good content ideas. And now to have it write something truly in a way that's actually good content with good wording that doesn't sound cringe and like forced and AI. Like, dude, this is pretty [ __ ] good. Woo! >> Safe to say we were happy about the breakthrough. All right, so it is currently 1:30 in the morning and I am realizing that everything is actually just about your prompts. We had [music] essentially been agonizing over just a single page of instructions for now a week straight. Literally just a combination of a bunch of sentences and words. I realized that it's like how the Coca-Cola recipe is locked in a vault somewhere where the Coca-Cola recipe is theoretically just a couple words, but on top of those exact words and in that exact order and that exact recipe, a multi-billion dollar business is burnt. And so the innovation isn't in the code. It's specifically in the crafting of the exact instructions and words [music] you give to the agent. It's in finding the perfect words to make the magic happen. And this is what intellectual property or IP looks like in the AI age. But of course, even with that insight out of the way, we hit another roadblock. All right, so let's talk about the last few days. A couple days ago, we realized that all of the prompts that we had built were trying too hard to force Stanley, our AI agent, to do certain things at really specific levels of outcomes. So, let me explain. The more instructions you give to your AI agent, the more likely it is to break. Because if you tell your AI agent to both be friendly and data driven and casual and chill, but also super smart and intelligent and professional and all of these things, it has no idea how to compute that in a way that is consistent. Just like a human being that is being birthed for the first time, if you give it all of these instructions, it gets confused and fails. And so we basically a couple days ago had to blow up our entire agent. We also shifted models from chat GBD to cloud, which took a day or two. And so we had this really frustrating feeling where like we basically felt like we had to take a bunch of steps back, which when you're on a 14-day time crunch, 2 days is like beyond gold for you, right? It's terrifying [music] and starts to induce a lot of stress cuz you're like, I just wasted the last 2 days of sleepless nights. I'm already exhausted. I already felt like we probably weren't going to hit our goal cuz by the way, we have no real revenue at this point. And your whole team is counting on you to come back from this trip with something amazing. And so the weight of realizing that we had to take a massive step back and [music] start from first principles and square one again was absolutely crushing. But then something [music] interesting happened the next day. Um we actually spent the whole time at Steven Bartlett's office. So Steven Barllet, the host of the Diary of a CEO podcast, got really excited about one [music] specific feature that we were building. And that feature was our outlier content engine where basically we find the highest performing posts in your niche and then we use AI to rewrite that post in your specific voice taking what works as a formula and making it authentically yours. And that excitement [music] at a point where we were already exceptionally low gave us massive conviction. And we rode that wave hard the next day, managed to ship a ton and realized, okay, we actually need a break cuz we are exhausted. and we finally went to go see London. And that basically brings [music] us to today where we're going to start cranking again because we actually have a really big customer interview with uh one of the biggest creators on LinkedIn, Laura Costa, who she's going to give us a ton of feedback on what she thinks of Stanley and whether or not she would actually be down to use it herself. So, we got to get cranking, but I am excited to see what Laura has to say. >> You're our first customer interview. We We want you to be brutal. >> Okay. I like it. I like the vibes. I think that the language could be more so we've done this personalized [music] to you. So I just don't know cuz that's cool cuz if I know it's personalized to me and my niche >> personalized to you >> and I didn't know what this was but it's like hey do you need more inspiration to write today here are the best three posts that we found specifically for your niche or like hey I spent the last 3 days analyzing your competitors or like your peers. M >> if you're struggling for some inspiration to write content, here's >> the three best use of content I found for you >> in the name. And I think using the word personalized, so like we spend time doing this for you specifically. >> Yeah. >> I think that's [music] that would be what would get me reading this >> excitedly. But I like it. I love this. This is what I need every single day. >> So after this interview, we ended up going to a huge creator event with Lara where our game plan at the time was to go out and meet multiple more beta customers. But reality hit us like a truck where in that moment we were surrounded by hundreds of people and we were so exhausted from 10 days straight of hustling and working and coding that we went fully nonverbal and had to peace early and we're just like whatever [ __ ] it we desperately need some sleep. [music] I can tell you that I'm absolutely gassed right now. Just truly like the car ride home the and I were just so silent, so tired cuz there was so much like stimulation. There were so many people. But we did get good feedback from Lara today. So what's our game plan starting tomorrow? >> I think my mind is now at a state where I already forgot what you just asked. Point being we are just going to wake up, show up and do the best we can just like we've always done and [music] we're going to find the way in the process. >> Yeah. And I think um we have enough to iterate on over in the next few days where I would set our goal to land our first few customers within 2 or 3 days. That would be a stretch, but let's try. Look us. [music] So the next sales plan was to land more customers or die trying. >> All right. So I teed up one of our first ever customer calls here. And so what y'all are going to get to see me do here is how I think about customer discovery. So asking questions, trying to figure out what the pain points are and then how do I channel that potentially into [music] positioning our product and what we built. So selling the product and really through that process trying to get as much feedback possible on how we can improve the product for them and then also if it actually meets the bar of being a product that actually adds value to someone's life. So let's see how I do. What's up Eric? We appreciate you taking some time. We basically we've been cranking away the last two weeks on trying to create something that actually adds value to your content creation flow. And so we'll walk you through that in a little bit, but I think the best way to benchmark is to understand better from you what are the current pain points in your content creation process when it comes to LinkedIn specifically. >> It's the wild west when it comes down to LinkedIn. It's not what it used to be in terms of a job board where employers and employees are just making connections. It's not quite creator le either like Instagram is. One of the main issues is for creators knowing how to build something which is structurally sound on one platform and then how to scale it properly because most of them they don't know how to build the funnel. They don't know how to push people from the platform to their freebies from these freebies to their newsletter from these newsletters to build something sustainable. And thus I feel like these creators are going to burn out quite quite quickly. >> Yeah. Yeah. Yeah, that makes sense Harry. And so if we were to just scope down the problem to ideation, even if you think about your ideation process and content creation, I mean what are what are the pain points you have kind of in ideation? >> I look at content creation in phases as well. So I start with the idea which is then broken down into hook and delivery mechanism. So how do I make it sexy and then how do I provide value? >> Totally. >> Then from that point onwards we move to copyrightiting. I personally built out a custom GPT. I'd love to hear more about the custom GBT you built like what what did you do to build that and and yeah I'm curious how you use it and what you like about it. >> So I have a J2 project which is based on my language. I have files that are essentially combination of my best carousels that I have fed it and then I have styling guide that I have also fed that is essentially the foundation of the GPG. I have a prompt one for the different formats that I work with. So I have one prompt which is for captions. that have one prompt which is for carousels. >> In terms of those GPDs, what do you wish they could do too? Cuz there's obviously limitation to custom GP. [clears throat] >> What I feel like is missing right now is some a tool that does the whole process. There is no tool that can help you synthesize your language, build the foundation which is essentially a clone of yours because if people are not familiar with AI, they cannot create a custom GPG that sounds like and it sounds general. They also lack the taste that is necessary for you to actually be able to align it properly. So I don't want just a tool that can give me ideas or just a tool that can analyze v viral posts. I want something that can take that and make sure that the context is relevant for me based on the foundation. >> That's actually exactly what we've built. The product basically that we want to run by you is is twofold. One is it's a thought partner and writer. So the way that you built a custom GPT, I would argue that ours might over time perform better cuz not only do you need to update your GPT constantly with new posts, but it doesn't have the information of what else is performing in the market. Yeah. >> And so I I'd be keen to get your feedback if you want to share your screen, Harry. >> [music] >> So, I'm curious for you. If you hadn't built your custom GPT and you saw this, how much would you assign valuewise to this on a monthly basis? >> Well, Harry, you said that this would cost you a couple hundred a month. You know what we're offering this for? >> What? >> $29 a month? >> Yo, that's crazy. >> Do you want to be one of our first beta customers? >> Yes, sir. I would love to do that for you guys. >> Heck yeah, dude. >> Absolutely. Thank you so much once again. Hope you have a phenomenal rest of your day, >> dude. You too, Harry. Great to see you. >> Thank you, Harry. >> Likewise. >> Bye. >> Bye. >> LET'S [ __ ] GO. [screaming] THANK YOU, HARRY. You're the [ __ ] man. All right, we got to crank now. You ready? >> Yep. That's what I work to do. >> See you guys. All right, so let's talk game plan because Vitalia is beat from grinding all week. It's 8:00 on a Friday. We're having a crazy night out. We have 20 different cold email targets. So people have no idea what Stanley is. You've never heard of it. You've never heard of us. And we're going to have Stanley write personalized endofweek analytics reports on how these creators posts have been performing because what we're trying to test basically is can we generate so much value and use this cold email as a hook that they start to play instantly. And so over the course of an hour I crank out and set up all these emails to send so that we [music] can wake up tomorrow and see who bites. >> She's still using it. >> She's still using it. This is sick. Guys, we're watching a live user use our product. She's using it right now. >> She's still there. >> Oh no. Something went wrong. PLEASE TRY AGAIN. NO. >> SO SHE GIVES US A SECOND CHANCE. >> OH, she's still gone. She's still going. She's still going. [screaming] >> So she might actually use this. She's copying this. She might actually use this. >> She might actually use this. Dude, >> you should give a customer of her life from this. >> She's copy. She's copying it. Let's go. She's scoping. She's scoping the whole thing. She's like, "This is my good. It's actually [ __ ] good." >> And so this moment was so wild. Like I can still feel the energy and emotion of that moment because you were holding on for dear life, like clinching your butthole, for lack of a better term, watching this person use this and you already know how clunky and shitty [music] your product is because it's an MVP product, but then there's that like ray of hope and it's like, "Holy [ __ ] they're using it. They're using, they're using, oh my god, they're clearly getting value." like they're copying and pasting and they're about to post your [music] idea that your AI agent actually generated for them because it's actually good and they liked it. And then in that very next moment and it's this is truly just a metaphor for the broader entrepreneur [music] journey. You were thrown into the trough of despair because your product just breaks on them the next second. >> Another customer. >> We got another customer 3 minutes ago. >> Oh my god. >> Don't tell me. >> Don't broke. I'm gonna cry, dude. I'm going to [ __ ] cry. >> No. >> No. [screaming] All right. So, what are we going to do from here? We got two customers actively trying out sand in front of our eyes. >> And they're not churning. They're coming back. >> They're not turning. They're coming back. They're finding some sort of value. So, one of the biggest issues with Stanley's user experience is that Stanley [music] takes a really long time to think of a really good prompt and post for you. It can take up to 20 plus seconds cuz he's [music] really cranking his brain. And so for our users, they're just seeing a little loading dot and just sitting there being like, I have no idea what's going on and I'm probably going to get bored and I'm going to exit from this product because I don't want to wait for this. And so as you think about giving your user feedback of what is occurring in your product, it's really important for us to show what Stanley is doing behind the scenes. Aka, he's currently analyzing [music] your best posts. He's also then analyzing all the top performing posts in the market right now to make sure and then crunch the numbers and [music] really think through how to get you to the best performing post possible. And so we're going to turn that boring loading dot into an experience that keeps users excited. And so now you're going to get a live example of me and Vitali vibe coding the feature. So if you're looking to learn how to vibe [music] code or understand what it looks like, this is your moment. >> Okay, first let's establish the the starting point. Let's look at the dot and count how long it takes. 2 3 4 13 14 15 16 So it took 16 seconds until we could actually do everything that we need to do. So we got to be entertaining for that amount of time. So first we have we have the the black blinking dot. All I'm going to say is make it more fun and playful by changing [snorts] the pulsing black dot. Let's see what we can Let's see what we come up with. And so what it's now done as an interpretation of playful is it's written a bunch of lines of code, but it's actually brought the Stanley icon in to be like the bot thinking. And its interpretation of of of playful is it has a sun and and a moon emoji now circling Stanley as it's thinking. So our next step here is I'd give it a prompt of two things. One is well remove the sun and sun and moon emoji. And then also we need to have it give a prompt of what it's currently doing. So it's like currently retrieving your post, analyzing your post, all that kind of stuff. So [music] now we should create a prompt here that's actually going to do both of those things. thinking, reading your recent post, analyzing your best performing posts, looking for inspiration from other top posts, crafting a thoughtful response. I think this is actually Yeah, that's okay. Uh, what what else should we throw in? >> Crunching numbers. Crunching the numbers. So now, theoretically, it should update the code base to actually have next to Stanley. Oh, there we go. And so now it's rotating through these things reading. And so we got to get rid of this sun emoji. Animal rotating heat logo and rotating statements. >> Okay, so now we're good. And so now you guys are seeing how this actually [music] works, which is it takes a couple iterations to direct them properly. So through live v coding, you get better management of being really really specific around someone and like a real person for example and giving them really really good directions. Now we're basically done on that and we can chip that as a change. And that guys is your lesson on vibe coding. All right. So today is day 12. So the question is what are we doing today? where at this point we basically have a semblance of product market fit where we sent an email a cold email out to 10 people of which five or six people sent us feedback back about how much they loved it which was a really positive sign for our first ever alpha ship and a really scrappy MVP ship but obviously a super small sample size and so at this point we've got to do two things [music] number one is we've got to get more data on usage aka we got to bring in more customers that's going to be my focus [music] today number two is we have a general sense of why people are enjoying it, but we haven't really like really gotten deep enough on understanding the problem around this and how we can double down on it in a way that that's where we need to focus for the next few days is like how do we truly build a product that is so valuable that really just like hits that one like satisfying point in your brain of like holy [ __ ] this is super helpful and I want to come back every single day cuz this is so so so valuable to [music] my content creation flow. And so those are two key things we got to focus on through day 12. So, we noticed something both really interesting and [music] really promising that day. There were three users in particular who seemed to really love Stanley. Specifically, we had one user who spent over 45 [music] minutes in a single session with Stanley and two other users who were coming back to use Stanley for 10-minute sessions multiple times a day. [music] And that kind of retention and usage where people are willingly going out of their day to spend up to an hour of their time with you is incredibly exciting and promising. where our job from there was to watch their sessions and understand how were they using the product, why were they getting value, why were they coming back and then double down and iterate on those specific features. [music] All right, time to ramp up the customer calls for today. Mike. Hey Sophia. Sir, I would love to spend this time understanding from you. What are your current pain points and goals with debt? And that's where [music] Stanley comes in. Well, there's two pieces specifically that that going to try to help you do. Number one, let's I'm actually curious. Let's tell to write that post that you had around money. >> We're only 2 weeks into this, so you know how much we're going to improve it. >> Content is a power logging. So, the top 1% of posts get 80% of the returns with it because I think what we started to build here is is going to be likely helpful. I'll be dead. And if you're happy to meet [music] with that, then yeah, for sure. Love it. It's good to meet you, man. Yeah, appreciate that. >> All righty. Just cost $200 made. Big [ __ ] deal. How are you feeling? >> Feeling great. This is how his tribe [music] dashboards look like. Up and to the right. >> To the right, baby. >> That's actually infinite growth over last week. So, some could say we're pretty good. >> To the moon. >> To the moon. Doesn't look anything like our standout, but it will one day. Couple million. Couple million. >> Yes. The beginning. >> Maybe a little billion. Maybe a little couple hundred million to a billion. We'll see. All right. So, today is day 13. >> I feel so stressed. >> Really? I was >> so sorry. You go to shift so much. So, it's pretty clear to us at this point we have a pretty strong semblance of product market fit where of our small sample size of customers, what's our daily activation usage so far? Pretty much everyone that we showed Stanley to, like what we're seeing, it's blowing my mind. They're coming back every day and they're spending at least 30 minutes talking to Stanley about their content without us asking them. and also most importantly to know on a Saturday and a Sunday. >> So reflecting on that second day of sales calls, I started to hit my stride and really understand what were the key pain points that people cared the most about and how specifically our product Shanley could help them the most in that. And what was really exciting, [music] really promising more than anything was that after people say, "Holy [ __ ] this is actually really good. You guys, this is only 7 days in. Like I can only imagine what this looks like 6 months from now, 12 months from now. It's already like really helping me make better content." And we'd [music] have people write in a day later saying, "Holy [ __ ] John, my first post with Stanley already went super viral." What was even cooler than that is that [music] people kept coming back to use the product for 30, 45 minutes. Even on weekends, like on Saturdays and Sundays, would [music] come back to Stanley and use it for 30, 45 minutes to write LinkedIn posts. Like on the weekend, things were really starting to feel like they were coming together. And so at this point now what we're thinking about is okay [music] for the next two days how do we wrap this up in a way that we can productize this experience fully. And so I kind of thought about how what I described today. Today is the dwizard of ozification of the product where there's just so many places where like you can't just like lease something out in the world for it to grow cuz right now you can't even for example [music] set up for an account yourself. You have to like manually provision you one because that's something that as you get scrappy with an MVP it's like do I need to build that or should I just test it first [music] before I even build that? And so the next few days why Vital is feeling more stressed than than me is cuz he's responsible for a lot of that stuff. Okay guys, there was 14 days for [music] us to build and we figured out that building a registration flow is optional. Okay, we're not going to do that. So we focused on building that the user experience [music] itself and we hacked our way with the links to actually identify customers. Uh but the way to build like fast and actually [music] quality is to be really rigorous about what is it that moves the needle that helps you validate your ideas versus what's unnecessary. And for us, registering users was not a good use of time. So we focused on other things. But you got to catch up on all of that. So So now is time for us to invest in that. Yes. The best way to put it is most ideas of what you need to build are probably net good ideas. It's much more important to think about the sequencing of all of those potential ideas. So with that being said, we're going to create the rest today and then we actually have a very important dinner tonight with probably if you were to pick one person to be the most important creator to have buyin and excitement about our product on, it would probably be him. I am so pumped. >> Yeah. All right, see you guys there. >> All right, we got a little workout in. Now we got our customer calls for the day. Let's go. Woo! Back to back. So Matala's head is covered, but he's smiling cuz we got a customer. >> Well, great job. I'm loving this already. Oh, that makes us so happy. Thank you so much. Talk soon. See you. >> Yeah. Bye >> bye. So, >> well, those are pretty much 100% record. >> Yeah. Got a couple more customers. All right, let's head to dinner. All right, we are now headed to dinner with Justin Welsh, which is a crazy full circle moment. So, to give you context, Justin Welsh is the Michael Jordan of LinkedIn. Like, he is the OG LinkedIn guy. Like every single creator on LinkedIn who loves LinkedIn and thinks about this specific market, this [music] specific environment looks up to Justin Walsh. So whatever Justin Welsh says and whatever he believes in as a product, you could think about all the adoption can rain down from what Justin recommends. And so in many ways, all of our hard work was built up for this specific moment, which is could we deliver Justin a product and experience that he was like, "Holy [ __ ] this is awesome. I want to use it." So let's see how it goes. >> Hey there, sir. Good to see you. How are you? >> So just the first question. >> My very first question just for you to to meet Stanley. I think you should ask him what does he know about you? It's kind of a recommended question. So just get that. >> Oh wow. Okay. Interesting. It knows that I have 745,282 followers, that I've got a private network of 800 entrepreneurs, even though that's brand new, just 4 weeks in. Very cool. Your content philosophy philosophy is refreshingly counterintuitive. This is very cool. [music] I mean it's really interesting because if if I type in what do you know about me on chat GBT or [music] claude it can pull from like my website which is yeah is you know a of a static point in time when it was created I don't know a year and a half ago right and has a lot of my stuff around solarreneurship and things like that but it doesn't have any of the the stuff [music] from my posts your writing style is distinctive and powerful which is true >> sounds about right [laughter] >> the subs are true for sure the short punchy sentences that sounds like Dustin Walsh to Yep. Yep. Lead with bold contrarian statements to [music] stop the scroll. Use personal stories and vulnerability. Always enter the clear CTA to your Substack. Unsubscribe. Yeah. Nice. This is very recent. Content performs exceptionally well. 1 to 12,000 reactions. You help entrepreneurs escape the performance of entrepreneurship and focus on actually building. You're particularly passionate about helping people design businesses around their lives and not the other way around. That's like my core sort of mantra. this knows me a lot better than, you know, Claude or Chat GPTt does simply because it's pulling from the stuff that I actually write versus what it's carving off of a website, you know, that's got an about me section, which I think is really interesting. >> And the second question that now that you're working with an an advisor partner compiler that knows this much about you, let's see the quality of the ideas that will come up [music] for you. So, the next prompt that I would encourage you to to try is give me five post ideas. Like if it was to to shorten your ideation cycle and suggest for you something to post, what would it be like? So, let's let's see. Oh, cool. >> I'm curious if these Ooh. Oh, the second bullet Jen's going to love. >> Cool. [laughter] Uh, I track three numbers that matter more than revenue. Hook into the metrics obsession. Flip it on its head. Everyone tracks MR, CAC, LTV. I track days I close my laptop at 3 p.m. uninterrupted dinners with my wife. Books read with zero business value. It's really interesting. I like that. >> That's good. >> Yeah, that is [music] good. The way that I look at this is I would I would I would edit this to to be more in my tone, right? So, it's it's writing the idea more so than writing my tone. >> Um, but this captures how I think. It captures the stuff I want to write about. What I do like about this is [music] that it's not a full post. It's more of a le. It's almost like a trailer. So I could take what the trailer has here and turn it into something more [music] meaningful that that's in my tone of voice in a full post inside the unsubscribed three members who broke the rules in one. This [music] is cool. This is like one thing I like is that it does have recency bias so that like it's talking about things that I'm talking about today. I like that. One of the things that I spend the most time on each week is ideation. Um once I have an idea, writing is actually the easy part for me. Once I get like three bullet points or a couple of sentences on a page, I can flow. I can write and this is this is something that helps with that problem. It would be interesting if it didn't have as many anecdotes that were not true, right? Because I have to kind of throw that one away or or or think about a different story or was able to pull somehow from anecdotes that in fact were true. If I write about that on Substack someday, if I write about that on a different platform, like if it could pull that information, I think would be really really useful. >> Why don't you ask that? What are my best anecdotes that drive the most engagement? What are my best anecdotes that drive the most engagement? That's a good question. Based on your top performing posts, here are the anecdotes and storytelling patterns that drive the most engagement. The critics were wrong narrative. Yes, this is one of my recent most popular posts. You can't build a one-person business. You can't do this. You can't do that. So, start with what they said. Reveal the truth about critics. They're scared, not wise. And end with, I did it anyway. You should too. Underdog story. Very cool. I like that. Build the company you want to work for. Uses a simple but powerful reframe. Instead of finding the perfect job, create it. Acknowledge universal pain points. Offer empowering alternative. Admit it's hard but worth it. This is cool because a challenge that I struggle with and I've talked to a lot of people building tools to try and solve this problem. uh but I haven't found a tool that would solve the problem [music] yet is I write so much content and actually finding stuff that I wrote once and never repurposed is a thing that's really challenging. [music] >> So this surfaces up some of the stuff that I I forgot I actually wrote. >> We have a weekly email actually. It's called the easy repackaged idea email. It's just one post you haven't posted about or topic in 6 months and then three ways to easily repackage it. >> That would be a tool I would pay for. >> Yeah. And then on and of itself >> repackage it. Let's just try asking if we haven't tested it, B. Like, what should I post again that I haven't posted for a while? >> If this pulls up stuff that I haven't repurposed, I'll be very impressed. You haven't done this in 2 months and it consistently crushes 189. That's cool. The building company you want to work for. That one I haven't done in a while. Fear is fuel. I know this this post very well. Um, and I haven't posted that one in a while. >> Nice. >> It's cutting down that prep time [music] that we talked about, that 3 to 3 and 1 half hours that I spend every week creating stuff. So that's super useful. And I also want to share this. We only traded on your last 500 post. It actually hasn't read the entire thing because it's like we had a lot of we only had so much time to actually prepare. >> By the time you upload everything since the beginning of time, just imagine how much intelligence it would have to actually pull the insights from all of those times. >> Yeah, that's super interesting. >> Love it. >> And this is just like the the version of this doesn't have scheduling or publishing capabilities. It's just a it's a it's a creator creator partner, >> right? It's creative partner. just a thought partner for you to just [music] go have an easier time getting the most difficult part of it. >> I love it. >> All right, Stanley is now officially Justin Walsh approved. Let's go. We're uh going to go to bed now, I bet, and then prepare for a big final last day. And so with Justin's approval, we were heading into launch day. All right, [sighs and gasps] so it's day 14. >> It's day 14. It's the afternoon and we're a few hours away from actually completing the one of the most exciting projects we've ever done. >> So, Nutella just put up a really cool launch post. This is the time we'll look back on and say, "Holy [ __ ] we [ __ ] did it. >> We [ __ ] did it." And that was the moment, the pivotal moment where we finally figured out how to make Stan Aai first. The things that we've discovered, the changes in thinking, the knowing what's possible by going through this experience, invaluable. Can't read it. can't hear about it, can't store it anywhere. You got to walk the path and then you feel it and then you know that the future is going to look very different. >> This is the moment where we drive our first $200,000 in revenue in just a single day, which is where we've been building hype through building in public every single [music] day. So, 14 days straight of posting content about how we're progressing. And we've also finally built a scrappy little MVP that people aren't willing to pay for, which is crazy to [music] think about in of itself. I think this was like this was truly was a very pivotal moment for us and we built an incredible agent along with it and um I was just telling one of our new customers who signed up just today at the end of the day what's going to make me the most hyped about this is seeing our creators and our customers use this to actually outperform the market to make incredible content that they're really proud of and really build their businesses. And so um I think we're we're doing our best to try to thread all the different needles and end up still being commercially successful at the same time. So, we're going to launch tomorrow, see how it goes. >> We had over 2,000 people comment on our launch post that Stanley was going live that they would be interested in using the [music] product. And through that post alone, we generated over $200,000 of revenue. Like, think about $200,000 of revenue just from that post alone. Just a reminder, 14 days prior, we had absolutely [music] no idea what we were doing. We had to struggle through the entire process and all the ups and downs. with so many doubts across the whole journey. And we had that moment there. It was like, "Holy [ __ ] we've done [music] something that is genuinely valuable to people and then already is generating real business outcomes for real. They're literally growing their followings and getting two or 3x better engagement on their posts in just 14 days. Like imagine what we can do as we [music] take this product further." And so the question now to ask is, how do we grow our agent from 200,000 AR to a million dollars plus in just a [music] few more weeks? This is the strategy that I want to make sure that you take down for your business, which is we didn't have to spend money on paid ads. We didn't have to do any crazy strategies. All we ever do and all we ever think about is our organic flywheel. I want you to recognize that your customers are your best salespeople. If you're delivering a product that delivers actual genuine value ad to their lives that [music] they love, they will do a much better job than you ever will running a ton of ads, what have you, trying to convince someone to use it versus them telling someone very authentically and very honestly like, "Hey, this product has delivered me immense value and you should use it, too." And so, out of our beta court, we had so many people posting about their experiences where they 2x their engagement. They cut how much time they needed to actually post content out by 2/3. And we even had one early beta customer tell me that he could already attribute $200,000 of contract lead value to using Stanley. And so you can imagine that if you're delivering that kind of outcome to people and they're sharing that with the world, if you just double down really intentionally on that flywheel where you know, hey, if I can deliver people's success and do everything I can to help them be successful and then they'll share their journey, their success story with the world and with their peers and friends and then you just rigorously execute on improving the percentage of people who actually succeed and then improving the number of people [music] who actively share their stories very very quickly. You can see how this snowball or this boulder we've been trying our best for 14 days to push up the mountain. Once you kind of get it over the edge and you find that product market fit and you find those one or two things that you need to improve in the product and you know how to double down and then you start pushing it down the mountain, you start to see how this thing compounds and it gets easier and [music] easier and easier. And I think if there's any lesson I want you to take out of this entire experience is that this entrepreneurial journey will have you staring into the void so many times, like so many times more than you ever think you should. It's in those moments [music] where you have to zoom out and trust the process. Trust that as long as you are hyperfocused on truly and genuinely delivering value to people, solving a problem for them, and persisting through that, and trying to figure out every creative way possible that you can possibly [music] break through, you will eventually make it.