Learn to Write PRD’s like the AirBnB Founder

Welcome to the daily Parker Rex channel. We talk about AI news. We talk about things that you should care about if you code or if you market or if you do business in general. We talk about how to stay up toate on AI so you can get more leverage. We go through little questions, but whatever. Let's jump in. This guy pretty controversial. Some people hate him, but guess what? He's really good at calling shots. So, watch him for 30 seconds. This should be required for anyone. that [ __ ] is coming for you. Like it's I I do not understand how people don't get it. And so one of the reasons I want you to use social is if you're doing social heavy, it doesn't make the jump to AI feel as hard. When you do AI, it just, you know, if you're staying on the sidelines, it just gets it gets further and further away from you. It's hard to like go to the glasses when you're still on a flip phone, you know? And so we just need practice. It's I'll I'll use this analogy. I'm trying to get y'all to do [ __ ] every day like you're training for a marathon so you can then run the marathon. It's hard to go like eating chips and [ __ ] beef jerky on your couch and drinking beers and not doing anything physical for a year and then just getting up in the morning and running the [ __ ] New York Marathon. That's what a lot of you are doing with your business. By not doing all the steps every day, as the technology keeps moving, it gets harder for you and the gap gets bigger. You become more discouraged and it's [ __ ] game over. But it's shockingly easy even for the least discipline in the room when you realize how much pain is on the other side if you don't do it. And more importantly, how much once you look at it from a different angle, it becomes a lot easier. By the way, I lived this for the first 38 years of my life. Do going to the gym and eating well was the worst. Yeah. So, he talks a lot about in this talk, which I'll link below, that technology is undefeated. If you're on the sidelines, what are you doing? This is the same exact thing, but times a thousand as him when he called social media being a thing. He had a couple hundred grand saved up from the wine business and he wrote a check straight into Facebook buying the shares from Mark Zuckerberg's sister. Kind of crazy, but he's definitely worth listening to in this talk. Yeah, it fires me up. So, I just wanted to call that out and let's get into some content. So, what am I doing? I'm going to set up a new terminal because it's about to be awesome because cursor conflicts with other tools and I want to talk about augment. So, I'm going to jump into that as the bulk of the video. But one trend I noticed that's going on is this AI generated video is causing the internet to explode where they're like, is this fake? Is this not fake? And then now all these I guess he called them celebrity really just like influencer LA people are talking about it. But it ultimately led to a ton of traffic. And I think this could be a tactic where I have so much stuff of me just talking like this and I have a high inflection or low inflection and I just have a lot of information that I kind of want to make an AI video of myself saying stuff just to see. Controversial. So Benny Blanco hits me up, right? No way. Send me this clip. Yeah. Sends me this clip and it's the kid Loy speaking on a podcast. So I start watching it and just just check this out. A lot of people think me and him got beef or whatever just cuz he like kind of sounds like me, you know? So Leroyy's got beef like what do we Bro, it's AI, dude. No, it's not. Dude, what? Of course it's AI. Has zero followers. Watch it again. I actually heard same thing a couple days ago. I I couldn't believe he made the beat. He is AI. Thank you. So, basically what he does is he gets you to watch the video with the deep fakes and then you stick around because the music's actually [ __ ] fire, dude. It's genius. Well, then what are we doing, dude? Play a song. So, that's just a really interesting way where basically we're at a point now where people can't tell. That's crazy. And this whole influencer model is going to get flipped on its head where you've seen a progression from influencers are the way to trick people into buying products. Trick is such a tough word for it. But if we look at it as a progression first social comes onto the scene, right? Like 2008 Zuck, you know, everyone in startups, if you're watching this, you've seen the social network a thousand times, but it's like cool. Let's say like 2010, right? So we're here and then mobile pops off. So, we got the iPhone going. And then after that, you have like let's say 2015, you see every influencer chick ever just printing money, selling product, right? So, the money is just like flying towards them. And we've been on that run for quite a while, like strong through 2020, probably a whole generation of people sculpted by this idea of influencer. And so I'd say like you have all these but there weren't that many people doing it and then it gets saturated like any market does. So now you have a larger supply. The money is still coming in but it's getting split and then maybe you have the top players who were printing more because they gained leverage because they're early. Like anything, if you're early, you win. Think of chat GPT. But now, this was essentially a rental service. So, you could see a deal. I've been offered deals as a tiny YouTuber. And it's crazy how it works because now you see in coming into 2025 and beyond where you have the influx of influencers who are digital. So we're here and now Starbucks is doing it. So the second that Starbucks said that, second that they said, "Hey, we're going to do AI influencers was market validation at the enterprise scale, at the largest scale." And what that means for influencers is if you're not already above this, then what's the point? because there's such a massive influx of digital versions of people coming in where if you looked at it over time and green will be the humans. We'll go back to 2010 to 2030. I think it's going to look something like this. Maybe it'll flatten out. Maybe they'll bring people in house. I've gotten a job offer from this YouTube channel to come in house somewhere because they recognize that they can make more money. They can save more money and get more value by bringing you in house. So, I really think this is no longer going to be a trend. it's started to taper but then moving forward it doesn't really change the distribution mechanism for the product because at the end of the day people don't want to buy from a company they want to buy from the people associated they they they don't want to buy from a company they want to buy the feeling that they get and when I think about it it's really the folks that if you compare two side by side and you have this girl And another influencer. They're always girls for some reason. But this one is just pure influencer means they're good-looking. That's it. No, they're good-looking. They're consistent. They probably have someone that's managed them. But then you have the hybrid which would be influencer turned business person because at the end of the day they've captured enough attention to then leverage that into selling a product or a service. So then they say okay well my skew selection what am I going to sell? You never see good deals or long lasting products when something like this happens where you have the girl and then you have the guy over here from the business and he's got a briefcase and he comes in and he just says, "Hey, you go market this thing and you're going to have nothing to do with the product." That that typically doesn't work. What does work is when you match the niche. So let's say she's fitness, then she can go and she now runs a media company. So with that media company at the top, and this plays into something that I'm doing, my strategy, but with that media company, you can do all sorts of things and you have freedom because you've captured the attention. So you might have skew one, this might be a supplement, and supplements are the most commoditized things on the planet. But for all the people, let's say this is the audience and it has uh 1 million people in it. For all those people in there, they just like her. They don't they don't care what the creatine is. They like her. She takes it. It's part of her daily routine. I see this with my fiance. She'll buy anything from a person that she likes. And that's the same with you. It's probably if you're in my school, if you're in my community, if you're associated with me in any way, it's because we resonated on something. And so what they'll do with that audience is they lever it into other things. You'll typically see this is supposed to be a microphone, but it looks like that. It looks like a magnifying glass, but it would be a podcast and other media assets. And so I think this is really the only route to go for it to be a lasting business as things get tighter and tighter and tighter. So this big thing, bye-bye to this bottom third. I think if these folks can elevate and pivot into something else, that's great. But the fact that this is happening with Starbucks, the fact that this is happening in Hollywood right now, which is pretty much synonymous with culture, shows you that the writing's on the wall with AI generated stuff. And it's gotten a rebrand. They're talking about it. They didn't say deep fake. That was something that was going on for a while. And just like any generation turn or separation between generations. So let's use an example. You have millennials. I'm at the bottom. I'm the youngest millennial. If you're 32, you're the youngest. Then you have Gen Z. Then you have Gen Alpha. And there's distinct differences in our behaviors and the way that we see things because millennials were the last ones to go from no iPhone to iPhone. Gen Z was iPhone native, meaning they grow grew up with it in their hand. And the behavior shift there was a lot of millennials, they get anxious about posting online because it wasn't normal. We didn't have Instagram. We didn't have all these apps for a majority of our life. And I just see this across most millennials. It's a thing. Whereas Gen Z, they'll film themselves putting their socks on. And that's an extreme example, hyperbole for dramatic effect, yada yada yada. But how that ties into the deep fakes is I think that that same thing will happen. And this is just a pattern that I'm seeing where deep fake to us is weird. So new thing, deep fake is the bad derogatory label. We hate deep fake, but AI me some new term that gets coined. That's going to be totally cool. With Gen Alpha, it's going to be kind of weird. With Gen Z, it's going to be maybe more on the deep fake. What is that thing? And then we don't even have to talk about boomers or Gen X because they will not even understand it. And then these people will be horrified. So, it's coming. It's here. It's wild. Now, I want to go through I was going to do some terminal stuff because I'm excited about that, but it's either that or exploring I vibe with AI. Yeah, how about how about we [Music] do the I Vibe with AI stuff. So this came out of a problem that I'm having. So let's do this. And this is at a high level how to plan product. So I'll time stamp this. So this is for how to think of and plan a new product from scratch. So you always start with the problem and that way if it's yours you'll know exactly how you want to solve it. You'll know exactly what that that problem is more importantly and how you solve it actually is just by listing out potential solutions that solve the problem and then it constrains it back to the problem because you want to do it quickly. So in my case, I run a community and so community management is outdated. Now you look at this and you'd say that's not very helpful. So this is that exercise of writing a PRD where now I say okay great I know that it's outdated but what are the specifics? And because I faced this problem so many times, I actually have quite a few of them. So I can say the management of content is confusing and say hard to find stuff. That's for me and the members. It is hard to network with people because the messenger sucks because of all these different things. Now, luckily for people who want to join the community, we've solved this in the interim and it's a great solution. We use Discord. So, if you're thinking about joining, don't allow these problems to get in the way of that. In fact, you should join because it's going to be twice the price when we launch the solve for this. But I start in this space and as I go I get more and more specific. So just to speed it up, iOS app is not native. Therefore it feels off. iOS app crashes. Desktop admin tools are limited. No open or no API design aesthetic feels like Windows 2005. Reorganizing content takes 26 clicks to move something. Organization of threads is not intuitive. Company who builds said software does not ship 10x cost for 10x or onetenth the UX expected. That's just me saying for $100 a month as someone who runs a community that's nothing. That's fine. It's a great starting place. But I know from being technical and from being able to build all these things myself that this could be a green pasture because I know a lot of people are frustrated with school. They're frustrated with [ __ ] There's nothing in between. A lot of legacy companies. There's nothing that's AI first. And so I look at that and I say AI community all about being on the edge tech should use AI. So think of it that way. And I also think that customizability extremely limited. And again, all these things do not negate the fact that lots of people are learning lots of topics and moving forward and learning and enjoying by being on here. I just have really high expectations because I know what's possible. And that's why everyone I urge everyone to get more technical. I mean, 20 22 million people per month find this really helpful. 42 people within our community find it very helpful, but it's because of the content in here, not because the platform's so great. So, definitely join. You'll save money. But let's get back to it. So, I I've dialed in the problem a bit. It's building conviction around why I'd spend time on this, but I'm convinced that it's worth spending time on. And once I've gotten through these sort of problem statements, now I can start thinking about solutions. So, I want to start to narrow out a solution. And I have way more problems than this, by the way. They're not just for the members. Members have not complained. They're having fun. they like the Discord. But for me, I also have other things that I can bundle in. For instance, on acquisition, no discoverability. I have no detailed product analytics. I can't see how long someone's having a session. I can't run cohorts. There's no custom GPTs. And most importantly, I don't own the data. And that to me is frustrating because it's really valuable to know in order to serve customers better, you need information. And this isn't a I'm gonna take their credit card and steal it or whatever. That's on a payment processor. It's like Fort Knox. Like I'm not I don't care about that. What I care about is if someone has an issue and I don't know about it or they have an issue and I don't see the post for eight hours because they're in Sydney, Australia, they should have that question answered or have it really easy to be found. And again, no one is complaining about these problems. It's just I know it can be better and that comes through understanding what the interests of the members are, what their level of experiences in either coding or marketing and then matching that to the content that best fits. So on a technical perspective, it's just recommendations. It's vectorizing all the information, providing FAQs that can be answered via quick rag, providing both the personalized news feed of things that are relevant by using vector search, by using Google Trends data, by using anything on GCP and then piping it in there, providing recommendations around other members. There's just so much you can do because I'm looking at this not as a get-richqu scheme because that's literally the opposite. This is how do you provide the most value for people so you can reskill as many people as possible and selfishly build an awesome network of people who are like-minded because I live in a tech desert and I've already met a lot of really cool people. So, when I look at this, I say, "Okay, well, that's that's great and all, but how am I going to do it? And is there anything else in here that I'm having a problem with?" Well, I think we've we've probably got enough there. So, then I'd say, "All right, well, we probably Oh, wait. Just kidding. I have one more. no ability to store and share prompts as they are updated. So, this is a biggie because a lot of people care about prompts. And if your prompts get outdated or they're tucked away in some folder that's gathering dust, and I don't even know that it's gathering dust, then we're all just flying blind. And people can get frustrated because they don't even know it's there. And I'm going to be frustrated because I can't know if they saw it. So solution an AI first. What does AI first mean? We'll go over that. An AI first media and education platform for both web and iOS. Now, why not Android that will come later, but this is the start. And so, when I look at this, I'm like, okay, great. This will be a companion app. It just needs to do the basics. AI first means a lot of the tools that you make, people are just slapping AI in there and it doesn't take advantage of rethinking some of the foundations of how the application could work. So if I looked at [ __ ] for instance, I spent a bunch of time on there at like two in the morning last night trying to just figure it out and it made me feel dumb. It's cool, that's for sure. But for instance, I'm on here and it's almost information overwhelm. I had to ask a hundred questions last night because this is not it for my use case, but it is cool. You have, you know, some unique UI. You have apps you can build in it. They have an API school. I can't find anything personally like I can't f to me the reason I I gripe about this is having been a product manager and knowing how much you can deliver for your customers if you just know what's going on. It's just almost heartbreaking to just know that I'm so blind. And I think that's really why people don't know. they don't see that this is a problem. It's because they're not aware of what's possible. And that same thing can be said for building any product. That's the secret that you have that nobody in the world knows is your product insight. And so with this, there's this solution and then I'd spend more time breaking this out. So AI first to me it says well we have information for educating we have free products we have video we have all these different medias that we can do so we probably need some one aentic workflows for every step of the onboarding and we'll use agentic workflows within the context of what's called pirate metrics. Pirate metrics is an acronym. It is AR. I'll just show you pirate metrics. This is how you shape the product. So, we look at this. We'll just snag it. Put a G in there on accident. I was excited. So they called them pirate metrics because our but we build aentic workflows around each step in the funnel. So when I look at this I can start mapping out what would be possible. And what I like to do is I stole this from Brian the founder of Airbnb. It's called the 11star experience. And what you do is you have this thought exercise of what is the best possible thing that we could do. What's the 11st star experience? And he goes on to talk about it and how this was formative for Airbnb where well maybe a fivestar experience would be we went to India and we got a good ticket. Seven star would be we went to India, we flew first class, we had champagne and then we got there and it was beautiful. And then as it ratches it up by the end of it, it's, you know, 11 star is private helicopter from the plane to the front door. They ride an elephant in and then everybody's there with a party and a live band, all this stuff. And so we've also called that the Taj Mahal. So you think of it that way. And that exercise, I'm not going to do it on camera because it'll just take forever. But it's really fun because it fleshes out all these ideas. And that's the most important thing you can do is by doing this whole thing, you're flexing your product strategy muscle to get better at it. And then you'll get more leverage when you go and actually open a code editor. You notice I didn't talk about code in any of this. And that's on purpose because you're figuring out what to build and why to build it. I spoke to somebody today. It was a free consultation and he was telling me all about this amazing thing that he built for an investor and then at the end of it he realized that the investor just wanted something basic and so he got caught up in overgineering and spent a lot of time and I told him look this is still valuable because you learned something you learned some new skills but that wasn't what to build and there was no apply. So we're figuring out what to build and why and then how is later. That always comes later. So with the Gentic workflows, I'm going to the 11star experience. So with acquisition, I start listing out these workflows. So I'm thinking of a set of agentic workflows that do programmatic SEO. And this is not just dynamic routes where you know we have something like prompt and then we have some sort of dynamic thing that just spams content. But you'd start to think of what are all the interesting ways that we could do this. Well, maybe you'd have a set of growth agents that go to places um that people ask common FAQs. So you could have an FAQ agent that's my face that has me getting pinged to make sure that it answers that it's provided on the cursor forum or on anything in this space are answered properly. answering questions on other people's YouTubes, answering questions on other people's schools, answering questions anywhere, the Windsurfum, Reddits, subreddits, tons of places you could add that. And you just start layering these in. And because you'd have both the outbound agent and then the platform that you owned, you could see which ones converted because you could capture the name of the person and then associate it with the email as they go down through the funnel to get the value. The same thing goes with activation. So, do users have a great first experience? This is something that you be You will be tracking like a hawk. It's called ftx. What it stands for? First time user experience. That's different than obviously eux, which is existing user experience. And so this is the big one because you're trying to figure out how fast do they get through this funnel? And not just that, but how do they get to the aha moment? And an aha moment is that point of value in the product where they say, "That was a good call. I'm happy I paid for that." So an aha moment when you're buying a sandwich was, "Wow, that sandwich was good. I'm happy I got it." An aha moment for Twitter in their early growth days was getting to 13 follows. So following 13 other people was the metric that aha moment was triggered by because they knew if they had 13 people followed that they'd get value out of it. And so I'd go through and I just ideulate on all of these. I'd think about the retention. I have no idea what's going on. I have to manually go and fiddle through a list and and see what's going on. That's useless because you can't build product around that to serve customers. I'm doing it manually now. I see. Oh, someone wasn't active and oh, I got to say hi. What's wrong? How are they not getting value? Revenue. So, you can start thinking about, okay, we have a membership tier. Uh, what's tier one look like? Could we build an offer to allow more people in because I'm going to have significantly more information which means more leverage. That leverage is for the good of the customer. So, can we open up more customers to deliver more value? You could think about tier two. All these things are opened when you have the information. You're not flying blind. And then same thing goes with referrals. There's so many ways you can do this. it's bananas. And then you could have a system where give one get one or get a discount on Tuesdays or whatever it is. So that's the exercise in a nutshell. I realize it's been a half an hour, but this is how you make good stuff. And had I done this every time that I made something, I would have saved so much time and money, it would be bananas. And the next step of this is you actually outline and you start sketching things. So I've already done this in my head, right? But I wanted to demonstrate for education purposes. But then I'd actually start laying out some screens. Everyone says Figma is dead. They're wrong. I just lay out some basic screens. So I think that it'd be really wise to have a place that I find helpful. If I find it helpful, others will find it helpful because this is the cohort. They're AI enthusiasts and marketers and in media marketers and code. So having a place with images that are from different models so you can see kind of how they work and then you can maybe fork those, grab those, see where to go. I could have these broken out by the different models themselves. And then I'm really excited about this because this is something that I believe will be worth joining alone, which is most of the time when you're making a prompt, you're taking something from one version to the next. You're going from one medium to another medium. You have a long form, you want a short form. You have a video, you want a transcript. You have an image, you want a text prompt. You have all these from to, from two, from to from to over and over and over again. So why don't you just have that? Doesn't that seem obvious? And you could tweak this, sure, but this primitive seems right. So that's what I'm going to do, and I think it's great. So you'd have a place for all the prompts. You'd be able to store them. These would be on a separate page, but then you'd have the from and the to likely the ability to select the model. So, if I since I'm in here, I'll go ahead and I'll delete this and I'll grab my pen. I'll pencil. I'll come down here. And what I'd probably do is I'd have the from and the to and then I'd have models and I can just tick them off. And then based on the format, these are dynamic. And then as soon as I kick it off, I just start to see the outputs side by side. And on top of that, I started to build something a while ago called the Hemingway editor. Anyone's familiar? The Hemingway editor does highlights of the grade level of speech. So, let's grab this block. Let's pretend like this is a sales letter. And we have three of them. And I have let's say quad. And I have What's another good one? Huh? Gemini 25 and then I have open AAI mixing models and providers. How about it? Let's just do GPT40. Why not? And we'll do CL [Music] 37 thinking whatever. But then you have these and they've come back and you're, you know, so excited to read them that you just can't even wait. But you're gonna have to. But then you have a toggle, this magical little toggle up here. And if it's on, which we want, then you start to see the text highlights. And you start to see additional information in here where I can Oops. Grab a highlighter. And let me make a nice highlight color. Let's come in here. It'll highlight areas of the speech that you've written which you didn't write that you've basically the outputs. It'll grab those and it will start to find the areas that aren't so good and that allows you to speed up your cycles. I can't wait for this honestly because you would also get a grade grade level and you can manipulate that because the best communicators on the planet speak at a fifth to eighth grade level. Now, of course, depending on your audience, if you're selling to neuroscientists, then no. But for the broad base of people, you actually want a lower elementary level school. That is it for today's video. If you're excited about this this platform that I'm building, then drop a comment below with your thoughts on it. If you found it valuable for me to walk you through in this sort of longer format video. Also say make more long form video or you can comment rubber duck. This is a coding concept that you're always wrong. the computer's not 99% of the time. You're wrong. So, you can comment rubber duck, like the video for me. That's the way uh you know the elephant will want you to like it, right? Look at that elephant. And subscribe. And I will see you in That's kind of cool. I'll see you in the next video. Bye.