Transcription
This is exactly how I'm using Claw in my businesses, and my businesses make more than a million dollars a month. This is an important business video for the claw-using entrepreneur. Because just like with programming, a kid who graduates with a computer science degree can tell you all about what code can do, but they don't have the decades of experience to tell you what the code should do. And would you know it, that my two decades of being a business owner, not using things like Claude, I'm finding immediate applications and I'm finding immediate real-world drawbacks that I want to tell you about in this business. So, this video, while I'm training one of my Claude bots, let me explain to you what I'm doing.
And this first one is, I'm using Claudebot to help me get my next business to a 9-figure valuation that fast. I've got a company called Raande. This is an Airbnb property of mine, by the way. And what we do is a full-funnel management of marketing, the algorithm, pricing strategy for Airbnb hosts and their listings. And that growth multiple will be very good for our valuation. But there's a lot of work onboarding a host. So, what I've done is I've created a bot that emulates human behavior because Airbnb is so good at catching robots. I have a device like this with a head, but it's a Mac Mini. And what I have it do is, I have it listen to an email account waiting for a co-host invite. And then once that co-host invite comes in from an Airbnb host, it goes and does all of the setup process, bringing the Airbnb listing in, bringing in the pricing software, putting it together, making sure all the names match, updating our customer internal data, all of that. And then it starts running our last-minute pricing strategy. There's still so much that a human has to do. And so, in tandem, I'm also training it on everything that I know with revenue management so that one day it can do all the hard stuff, too. But the deeper parts of running a revenue management strategy for an Airbnb listing are still beyond Claude.
But what I do love, and I think a lot of you will like this too, if you're in a space like mine, not just Airbnb, but there's plenty of businesses where business owners thinking that adopting Claude is going to give you some edge, what they're going to do is they're going to run some sort of Claude questionnaire and ask for answers to questions that Claude's not qualified to give, and it's going to give owners the wrong answer, and they're going to run their businesses wrong. I've actually tested Claude and ChatGBT in this space, and the answers they give on revenue strategy is childish and often dangerous for the host. I really like as an Airbnb host that there's no replacing human excellence yet in the space.
Now, the second part of the Airbnb business that I'm using Claude in is kind of like a Claude siege. I have it accessing a channel manager, which connects to Airbnb and all these other places that these properties can sell. We already have automated messages built into this channel manager, but they're stuff that has to occur conditionally. And I would say that these are when a guest is being persistent. For example, I've got a property downtown Dallas that has old-school keys, vintage doors, really beautiful loft-style property, but people get hard keys and sometimes they lose the keys, or sometimes they lock the key inside. Now, at 3:00 AM, somebody could say, "Hey, I've locked myself out." We don't have enough units of this nature to hire a VA to stay up all night. In my personal opinion, we've got very few buildings like this, even with the hundreds of doors that I have. What we can do now is my Claudebot, certain hours of the night, can wait for someone to say that they've lost their key or that they've locked themselves out. And my Clawbot can then give, only in those windows of time, automatic instructions to get the backup key that we have. Or we can also make it persistent where a guest messages us multiple times for assistance, or they say that they've tried something else and it's not working during the day, that the Clawbot will then give them this super secret stuff. Essentially, what I'm trying to say is a Clawbot, instead of just having this auto-message kind of thing or auto-response kind of thing that's very service-level, you can teach it to understand duress and sentiment and persistence and then condition it to respond to a certain amount of elapsed time or a certain number of messages sent where the guest isn't dropping a topic. We can build those persistence gateways. I'm thinking there's going to be a lot of applications for this.
And next, my Cloudbot's really good at conditional conversations. So, we have automated messages sent out where we ask people how they liked something when a guest checks out to try to then get a five-star review. But normally, a human would take over that conversation. Well, that automated message where someone can respond can now be run by this bot, where it then has a sentiment-driven conversation. It listens to their problems. It thanks them for their stay, whatever. It can then reach out to me and say, "Hey, we've got a DevCon 5 guest, probably deserves a refund, whatever." And then we can try to get a five-star review.
Now, this isn't the most special stuff in the world because there are companies that do chatbots and AI agents, but let me circle back to what I said about pricing. I love that people are using stuff and not knowing how to use them. In our space, and in your spaces probably too, a lot of the companies that create these AI as a service, they don't have a lot of industry experience. So, they build stuff based on their understanding of chatbots. They build their stuff based on the understanding of how the tech world works, but they're not building it with a depth of understanding of how hospitality works. And it leads to huge shortcomings. So, if you are a business owner that's looking for a true advantage, your wisdom in the space, your context in the space is going to be one of your biggest advantages.
And so, one of my big bits of advice for you here is, black-box everything. And this is a big statement for me. If you watch this YouTube channel, it's been the most forthcoming how-to Airbnb channel in the world. It's how actually I built my reputation and became a coach. I've got students in 43 countries, and they've made billions of dollars with my coaching, and we've just run organic, right? We've never had to run ads. And it's from the free content on this channel. And I've got to change my ways. I've actually got to be less forthcoming, and I've got to let less people in to know what I'm doing because as I build my internal systems, it's easier and easier now for you to make a bot that goes through all of my content and learns everything that I've ever made public and make a pretty good co-host. It doesn't even take watching my content. So, I have to start keeping everything in black-box, and I think that you should too.
And so, number three, the other thing that I'm using, I'm actually vibe-coding. This is a big part of my workflow lately, and I want to give you some warnings here, too. My business partner for one of my companies, the B&B Photo Factory, he worked for Dave Ramsey for a decade. Pretty much built a bank. And I learned PHP and JavaScript and all that stuff back in the newspaper industry. That's where the grays come from. But I knew enough about code, and with my business partner, that we've run into some really cool situations that we've handled really, really well with a cloudbot. If you use a website like Lovable, Lovable has built-in four dummy fail-safes that you could say, "Build me this thing," and it's going to know what you don't know. But a clawbot's not going to know the things that you don't know. So it's going to have deployment issues. And one of the things that you can do is you could say, "Hey, I'd like to build this feature for my website, and I wanted to do X, Y, and Z." Tell it to do that in a branch. Okay? Tell it to make a version of your site and debug and test to make sure it's ready for deployment, and then tell it to merge later.
Also, like my criticism of people graduating school computer science not knowing what problems to solve, I don't think that you should fall in love with Claude and like look for things to use Claude for. I think that you should get really deep in a space and then find where you have limitations to your own expertise or ability in this area that you're a really good human in the space. So, for example, another context thing for the vibe-coding. The B&B Photo Factory is an AI photo improvement software. Anybody can do it. The Photo Factory takes what I know about marketing and changes photos to my design. It takes everything that I know about marketing in Airbnb and includes that in our like prompt architecture. I've also designed a website that makes bulk upload really easy and different user journeys really easy for the software, too. There's tons of businesses that do photo transformation. You could make one, and you could probably get a couple customers if you make a compelling argument. Um, but then again, remember, if there's a thousand people offering the same product, there's going to be a lot of noise in the space. And why would somebody pick your product if you're just some nobody offering a widget that everybody else can offer? You need to offer something with a real order of magnitude improvement or some sort of built-in trust vehicle. So your position with people is going to trump the fact that cloud bots can do cool stuff because you can build a thousand things and no one's going to buy it.
So my advice for you is to stay analog as much as possible. Get good at a business, get good at something, and then apply your Claudebot to that. Don't try to apply Claude to everything. Don't try to apply crypto to everything. Don't try to apply Chat to everything. Don't try to apply programming to everything. If you are a slave to the widget, you'll never master anything to be able to bring that widget far enough in the space. But what we're going to see more and more of are good businesses innovating faster using code and cloud bots, and more noise in the middle. Just tons and tons of businesses that are copycats, but don't quite get exactly the point. But there are also will be good 80 to 90% strong businesses that will be right behind the heels of these innovators. And maybe yes, those companies, if they're better at positioning and marketing, they can constantly copy the upgrades of a tech where the business isn't as well positioned, and they could ultimately get ahead once the innovations start to slow down because in any industry, innovations will slow down. I think Airbnb is turning from a blue ocean to a red ocean, uh, for example. And so the pace of innovation won't keep you as far ahead as everyone else like it used to, like three or four years ago. And that's why I'm keeping everything in a black box because even though I have the position of this YouTube channel and I've got a ton of context, I'm not the best marketer in the world, and I'm going to need to guard every advantage that I have. And as soon as I deploy something that can be taken apart and reverse-engineered, then I lose my advantage to somebody who's a better marketer. So, I'm going to try, as a person who's built lots of businesses that have stayed the test of time, I'm going to try to only reveal as much as absolutely necessary as I build what I build. And most of my companies will rely upon some form of controllable effort that's unscalable, unduplicatable because again, one of the great things about the world right now is people don't have any fight. They don't have any discipline. They're going to try something. They're not in love with it. They're going to try to clawbot their way through. They're going to hit a couple obstacles. It wasn't quite as good as they wanted. They're going to go look for greener pastures. And in the Airbnb space, a host who's been in the game 10 years and has stayed open-minded will have all of the tools for success. And a first-year Airbnb host might actually lose.
In my next Claude video, I will teach you how hard it is to train a toddler, uh, which is the AI bots. The mistakes that it makes emulating human behavior. Uh, it's been very fun. My systems for managing memory and skills and context. Um, they're evolving really quickly. Um, it's been really fun. It's just like training an employee if the employee was born yesterday. It's been really cool. I will show you guys all that in the future. So, subscribe for more Claudebot related content and of course, other business stuff. Thank you so much for watching this all the way to the end. I'm about to travel in Australia for a speaking gig for a whole month teaching Airbnb. If you want to learn more about that, here's another speech that I did a couple years ago about Airbnb. Watch that. It's kind of fun. And as always, I will see you on the other side.