What Is Claude Code? And How To Use It

1.5 million people left Chat GPT and went running to Claude for all the wrong reasons. Claude isn't just the more ethical AI tool. It's THE tool built for serious businesses.

One of the most popular topics of conversation right now is Claude Code. I myself have been obsessing over it lately, finding myself dreaming of code to the point where I wake up thinking I've finished a debug.

The bigger question isn't just HOW to use Claude Code but it's really, SHOULD you be using it? I'm estimating it'll take a good 200 hours of use before you start to "get it", but that doesn't mean you can't make some pretty cool things along the way.

This week's episode dives into Claude Code, how I used it, what you can do with it, and how to get started.

For the links mentioned in this episode visit: thetoycoach.com/606

Mentioned In This Episode

Claude & Claude Code:
https://claude.ai

Claude Max plan (used while building):
https://claude.ai

The Toy Matchmaker (beta):
https://www.toymatchmaker.com

Toy Creators Academy:
https://www.toycreatorsacademy.com

The Toy Coach:
https://www.thetoycoach.com

Have an idea for a toy or game? 🧸 Click here.

Do you have an idea for a toy or game that you think could be the next big thing? Join the Toy Creators Academy to learn how to bring your toy idea to life! Or book a 1:1 call with Azhelle to get personalized guidance on your toy journey.

Episode Transcript

Azhelle Wade: [00:00:00] I built a pretty fricking awesome app. The hardest thing is to find the right stores to pitch to you've gotta do so much research about the store. This idea, I floated by Chris at New York Toy Fair, but it actually started years ago. You are listening to Making It in The Toy Industry. Season six, episode six. Hey there, toy people, AzhAzhelle Wade here and welcome back to another episode of Making It In The Toy Industry. This is a podcast brought to you by thetoycoach.com. today we're talking Claude Code. What is it? What I built with it and whether or not you should use it. I wanna pull the curtain back on Claude Code just a little bit so that you can decide whether or not this is worth your time to learn, [00:01:00] or not. Okay, so first to kick this off, I gotta say I built a pretty fricking awesome app. This app has a database, it has user authentication. We got payment processing. There is an AI powered backend. There is an API powered backend. If you don't know what that is, don't worry about it. I built this app with the help of Claude Code. what started as an experiment to see what Claude Code could do Has evolved into the most useful toy specific sales tool I have ever seen. so today I just feel like we've gotta talk about Claude Code. I am obsessed. I have been obsessed, and I think you might be obsessed too if you learn a little bit about it. Now, the big hype in AI right now is about how about 1.5 to 2 million users left chat GPT and a big chunk of those [00:02:00] users shifted over to Claude. I'll just tell you quickly and objectively, Claude is a slower response time Claude is a more expensive tool meaning it burns a lot more tokens to answer questions. And Claude has a tendency to check in with you rather than give you more information. As opposed to chat chat GPT that will say, would you like me to now create a presentation or a deck or a script or making you feel like I guess if you're gonna just lay it out there and offer it, why not? So those are the front end differences that you'll see in those two apps. So Claude is known for many things. they're known for their sassy commercials against chat GPT. really giving us that Mac versus PC energy. and most famously, right now, Claude is known for Claude Code. Claude Code is. where Claude can act as a program or software developer for you instead of just having a conversation with you. It can read files, it can write files, [00:03:00] it will edit code, it will run commands it can create accounts for you on different websites and run approvals through those accounts if you allow it, let's just do a quick demo of Claude Code. If you're watching the YouTube or Spotify, video you'll be able to see this. when you first boot up ClaudeThere are three different ways to use Claude. First is the chat, which we're all familiar with. You type in a chat, you ask questions,There's cowork Then there's code, when we click into it. you'll see a little digital data monster at the bottom where your prompt should go. And then you type in a prompt here. And you will need to choose a folder to store the data that the app is going to write for you. You always want it to ask permissions. at the bottom right underneath where you type in your prompt, there's a thing that says Ask permissions. You always want it to ask permissions because it is literally writing code onto your computer. So you want to be able to read what it's about to write and say yes or no. And then you're gonna have to choose a folder for it to write whatever you're telling it to write in. [00:04:00] And that's the folder that you'll see here in my screen share. It says Test dash app. That's just a local folder on my drive. If it's your first time using Claude code, a popup, a dialogue box is gonna pop up and say, where do you wanna save this information? So then you type in something like, help me create an app to scan my toy and turn it into a digital character. I, just making this up and then it's gonna ask you, can we trust this workspace where we are gonna read, write, and execute files? I'm gonna say yes, I trust it. So I'm not gonna go too deep into this 'cause I don't wanna waste my Claude tokens on a test concept. But it will start generating your app. right now we're actually seeing a great example of the benefits of Claude over. For example, chat, GPT. So Claude is already asking, what platform should this app be built for? Is it a web app? Is it a React native mobile for iOS or Android? Is it a desktop standalone application? It's recommending we choose web app. I'm gonna say yes. then it's asking, what should scan mean in this concept, in this context? [00:05:00] Is it take a photo, upload an image, or both? So I'm gonna say, take a photo. Now it's asking what kind of digital character should be generated. I'm gonna say a 2D character, Then it's gonna say, how should character details, name stats, backstory be generated? I'm gonna say randomized. Now it's going to use all of that data to start designing the app that is essentially Claude code. So where you may have needed a developer before to do something like that, you will get all of the code that you need to do it right here. Now what you do with that code is. Part of why I didn't start Claude Code six months ago, I believe it was out six months ago, Because of the wonderment of what do I do with this code? This feels very overwhelming. But we are in the age of ai. So if you are very interested in coding and this is something that you would like to learn and you would like to do. That's when you pop back over to [00:06:00] the chat side. You could ask it, okay, what do I do with this thing that you've generated? What do I do now? until you get this thing online. So other examples of. Things you could build with Claude Code is you could just build a tool for your business. If there's something that you find yourself repeatedly doing over and over again, that's like very monotonous. For me it's like uploading podcast episodes. I could build an app to help me do that. I know in previous episodes of this season. I built a make automation to help me with podcast episodes, and since doing that, I've come to realize I now have the infrastructure built that I could build that same thing completely independent of make, which Limits how much data I can transfer back and forth. I could build that and have it all live on my own database, my own server, using my own API credits, my own upload credits and just cutting the cost of a tool in doing so. Okay. I do wanna flag [00:07:00] a few things here. Claude Code sounds like magic and it is in many ways, but it also isn't, you do have to learn how to think like a developer. And in that, what are the pitfalls of an app that you might create? What are the security pitfalls? Like? What are the user bugs or holes people might fall into in your program or loops a user might find themselves in? You've got to predict and then problem solve for all of these things. And you might be thinking, oh AzhAzhelle, I just tried it and Claude created the app right away. What, what people have to be very careful with AI is that it wants to give you an answer. It wants to give you what you're asking for. So if you ask for this app that takes a photo of your toy and then turns that into a character, it's gonna give you that. However, what it may not integrate are parameters to make sure that when a photo is taken of that character, that [00:08:00] photo is not sent anywhere else. It won't add in things like making sure that. outputs are very age appropriate. These are all context pieces that you have to add into your initial prompts and your follow up prompts to make sure it's building exactly what you want and nothing less. Again, like AI is going to take the path of least resistance. Sometimes it actually thinks itself into a circle,and says it can't do things that it can do. It is so easy to do things a very expensive way. A very risky way in app development when you don't know how to analyze what the AI is building and why. So I just would really, urge you to only go down this rabbit hole if you are seriously interested in learning like a developer so that You can develop good quality or production grade software. Additionally I wanna flag use I'm on the a hundred dollars a month Claude plan. There are lower plans, but from [00:09:00] my experience, you are not gonna get very far your prompting. If you choose a lower plan, I myself run out of tokens and run out of usage and have to wait a few hours sometimes. Granted, I'm sending like a ton of prompts a day. But if you're gonna be doing some serious work in this app then you know you're gonna have to invest. This is not just a hobby type thing. All right, so now that now we're on the same page, I would love to tell you a little bit about the app that I created and how it went from an idea Toy Fair to a fully functioning platform. So I was at New York Toy Fair when I met Chris Coat , the founder of Cent Co. And that toy company knows what they're doing right now. They're fart monster. It's such a clever little design. I'm sure retailers are loving it. And I just I wasn't intending to pitch him a concept, but I just got excited in a conversation about AI and told him like I'm building this thing and it's, a site that will allow you to search by location and [00:10:00] identify stores that are a good fit for your product and to be able to pitch them through this app. And he was like, oh, that sounds like something we could use. And so with that. I talked to a few of my clients and some of my students. They also were very interested. And then I finally started building it. really it just started as like a proof of concept. I wanted to know if this idea could work, if it was valuable. So what started as a simple, literal zip code search for toy stores evolved into something much bigger. The Toy Matchmaker is now a smart searching and smart pitching platform. It's a full web app and it does some pretty remarkable stuff. So you put in information about your product pricing description, name, brand, of course and you choose a location, you choose a search radius, and then it will find and match every toy store that is actually a good fit for your product [00:11:00] nearby. And it will rate your stores by a percentage so you know which stores you're a best fit for, and which. Or not. And it also tells you why, everything from like demographic data, like this has a high grandparent population, great for giftable to So something as specific as consumers mention Barbie a lot here. So they're very into Barbie or consumers mention Montessori here. the hardest thing is to find the right initial stores to pitch to and to have the right angle. You've gotta do so much research about the store. This idea, I floated by Chris at New York Toy Fair, but it actually started years ago. So years ago I was at Astra Marketplace and Academy talking to Miriam, one of the board members at the time and friend of mine, She owns a toy store called the Toy Chest with several chains, and I was asking her about getting my students' products placed [00:12:00] in her store. It was actually Joey Dolls Samantha's Doll Line. And at the time she said, we do have a very high Asian population in this one store that I'm going to be building. And at that time I remember thinking, huh, I hadn't thought about looking at demographics to choose where my students had pitched. So then I said to her, huh, do you think that demographic data is something that I should use and look at when having my students pitch their products? And she looked at me and she said, you know what? Yeah, you should. So like fast forward years later that conversation was like pre ai, pre-everything. I think I was a brand new board member at the time. Fast forward years later, this idea is back and I'm empowered with AI to do something about it. So the Toy Matchmaker combines that original idea. Combines the zip code search scores, product based on their product fit,based on demographic data, based on industry insights that I have myself. And then scores [00:13:00] products one to 100 on how good of a match they are for a store. But it doesn't stop there because once you find the stores that you're a great match for, you can generate an email pitch, an in-person pitch a phone call pitch, and even an Instagram DM pitch. And all of those pitches are intentionally designed to pull information from that demographic data, from your product data, your sell through. Your product, you unique, your unique positioning elements of your product and pull that in through the pitch. There are pitches where it's framed from the perspective of the store, like your store and your customers say this and this is why you need this product. And then there's other pitches that talk about the product. When your product is the hottest thing on the market, how do you pitch it? So you say, this product's selling. 50 a week and your competitor stores are carrying it, why aren't you? The pitches that this platform generates are just, they're just so well done and I took so [00:14:00] long perfecting the creation of all of that. And then you might be wondering like, okay, Azhelle, what happens if my product gets a low score? If your product gets such a low score for a store, you actually can't pitch them. You have to reach out and ask Hey, I really think that this product is a fit. Can you re-look at it for us and give us an override? But you can't pitch them. We will not give you pitch materials or contact info to pitch them. Oh, and did I leave that out? Yes. Not only for each of these stores are you generating the pitches to pitch them, but the system is also finding the contact information you need, whether it's an email, a phone number, a store address, or an. Instagram account. So not only do you get the pitches that are taking into consideration demographic data and sell through and product unique positioning, but also the direct button to hit, send this email to this store owner or send this Instagram pitch to this store owner. And if you're wondering, Azhelle, does it work for flagship stores [00:15:00] too? Yes, of course it does. So if you have a pro account, you'll unlock flagship stores. So you can do a search and find buyers for let's say Barnes and Noble, instead of pitching the owner directly. So if you're wondering like, how the heck did a show do this? First I'd have to say I am a child of the nineties and in the nineties we did quite a bit of coding on our own. I made many a website in my day. HTML coding was like. My whole life. not to say it's the same 'cause it is not, but it definitely makes you more comfortable seeing code being run up a screen and trying to read and decipher and understand what everything says. So in order to do this, I did have to think like a developer. I had to ask questions like a developer. I had to research security risks. I had to understand how to protect my user's data. I had to understand what a database was, what a server was, what API calls were. Not just so that I could write it myself, but so I could. Direct and orchestrate Claude code into writing [00:16:00] it for me and to catch things when things went wrong. But I also had to think like a toy entrepreneur, like the person who is trying to expand their reach and get their product into new stores who's trying to impress retailers and not scare them off. And I also had to think about retailers. What kind of messages do retailers want to get and what do they not wanna get? What kind of products do they think is a fit for their store? And what do they not think is a fit for their store? Because what I haven't told you is the Toy Matchmaker app is not just for people trying to sell their products, it's also for retailers looking for products. So if you have a product that you post on that app, it immediately becomes available to the retailers on that app who can then search for things they're looking for. And if your product is a match, it will show up. There are multiple searches on the retailer side. There's one that's just essentially a string search where they could type in something they maybe saw on TikTok or Instagram and see if they can find it on the app. But there's also an option where they can just say, here's my store and now tell [00:17:00] me what's a good fit for my store. And it'll search through our database and find the best fits for their store. And if you're thinking like Azhelle, there's no way you're gonna be able to know enough about my store. We ask questions, we get a Store a snapshot from you as a retailer. So you tell us like what sells, what doesn't sell, what price point works, what doesn't work, and we use all of that to determine which items on our database we should show you and which we shouldn't. Not only that both sides, the retailer and the vendor side have. Built-in trackers. So for the vendors, you have a lead tracker to track all of your pitches. Have you won? Have you lost? Have you followed up, have you not? For the retailers, you have a lead tracker to track all of the products you're interested in, if you've requested a sample, if you've placed a purchase order. I've added in little details, like after 90 days you get a notification to say, Hey, are you interested in reordering this item? Because that is a huge pain point for specialty retailers. Like when something sells, [00:18:00] nobody contacts them to say are you gonna reorder? This app will do that for you, which helps the retailers and helps the vendors who also might not have time to do that 90 day follow up with all of their retail clients. So yeah, this is just, it's just so exciting. If you're. Interested in checking it out go to thetoymatchmaker.com or toy matchmaker.com. We are in beta right now, so we're doing a few more weeks of beta testing. You can apply for access to beta testing. Just fill out the form or you'll just get on the wait list. If you don't make it into beta, I would love to have you try it if you're seriously trying to grow your toy business or if you're a retailer, seriously looking for new product. we are planning to officially open this app to everyone April, depending on how our beta testing goes. Alright, the burning question in your mind right now, should you learn Claude code? All right. So I know some of you're thinking like, oh my gosh, I wanna build something as cool as Azhelle. Just did. That's amazing and some of you're thinking that's intense. Take my money, don't wanna do it [00:19:00] myself. Both responses are totally valid should you get into cloud code? It really depends on what your goal is. If you have a goal to just run the business, you've already built the business that you love, that you are having success in, Then your time is almost certainly better spent running that business and not learning and managing servers and databases, but. If your goal is to build a tech product as part of your toy business something that maybe becomes a second revenue stream, a passive revenue stream, or a tool that you can offer to your audience to help them use your product better, then yeah, it might be worth the learning curve. because it's so new that not many people are going to be able to tell you how it can impact your business. For somebody to know what creative, amazing thing you can make for your business, they would have to understand your business inside and out. And they would have to understand Claude code inside and out so that some people may [00:20:00] be in the toy industry like myself, but not a lot of people are well versed in this enough to help you do that. So if you're like, I need to do this immediately, then it might be something you have to learn. But if you have more time and if you're like, I am drowning and work Azhelle, I cannot take another thing, then I would just keep your eyes and ears open when you hear someone has like an AI powered platform. Perk those ears up and try to learn more because they may have something like this that they have built to answer and solve problems for business owners. They have done the hard part and it can just be another tool that you plug into and how to know if this is worth it. If you see another AI app coming out and you're like, oh my gosh, I have so many subscriptions and so many services I pay for, I'm not gonna put in another thing. Fair. So you have to look at what is this replacing. so if you are looking to bring in something like, let's say the Toy Matchmaker, if you are thinking of I was gonna hire somebody to help me with outreach, to go through [00:21:00] emails and things that I've pulled together and send out email pitches, then the Toy Matchmaker becomes a no brainer because instead of hiring someone. Two to $3,000 a month, you could get this app and pay $50 a month and do the same thing at an even higher level and faster timeframe. it's gonna be about what is gonna save you the most time. unless you have carved out. 200 hours in your schedule to learn Claude code and you really want to be that techie person. I would recommend working with someone who has done it or investing in a tool that already exists that can help give you the results you want. Just keep your ears and eyes open. Those things are coming. Okay. Let's talk about a use case for you of an app like this. you could have an app that allows a user to take your character or take your game and kind of insert them into new AI generated [00:22:00] stories or scenes. Of course, there you wanna have like limits and protections on what kind of things can be generated, but those are the kind of things you can do. Is this possible? Yes. Is it gonna be hard? Oh yeah. Is it gonna make you question your life choices? Yes. Will it make you dream of quad code? Yes. Look, you cannot be afraid of code. If you wanna do this. You don't have to write it, but you do have to look at it, read it, and understand it. there is a real trap with building with AI that so few people talk about, and I wonder sometimes if people even realize it. And the trap is this, AI will always give you an answer, right? And it will always tell you that it's doing what you asked, but that doesn't mean that it actually is. Every time Claude Code creates something for you. If you are not being extremely specific and thinking about. What that thing should be or look like or do at every single step, and you're leaving gaps and expecting the AI to fill them in [00:23:00] intelligently. It's gonna fill them in, but with the easiest, simplest, and fastest possible solution, which is almost never the detailed, accurate thing that you probably wanted. if you don't give like an. excellent, prompt, upfront, you are gonna have to constantly be asking, how does this work in my app? Exactly. What are the steps happening behind the scenes? walk me through every call or API call or AI call being made and then rewrite this in this specific way and do not touch x, Y, Z. That detail is what keeps you from building something that just will fall apart. again, the real question to ask yourself before you jump head first into Claude code is, are you building an infrastructure that is going to help grow your business? That your business is gonna just continue to grow around? Grow your business's revenue? Or are you just trying to avoid doing real work? Are you just I heard that this would save me time, so I'm gonna do it? 'cause no, it will cost you more time upfront than it will save you, [00:24:00] but eventually it will save you an immense amount of time. last thing I wanna talk about is the death of prompt engineering and the real magic of ai. Multisystems like the AI multiverse, but AI multisystems. All right, this is something I'm hearing a lot in conversations right now. People are thinking that if you want to win with ai, you have got to know how to prompt. You've gotta be a prompt master. And they're not wrong. Prompts are important. People sell prompts, they're so important. the right prompt can genuinely unlock better results, but to really differentiate yourself and your business from other businesses that are also using AI, that are also prompt engineering. The real competitive advantage is from layering your AI usage. The first time I noticed this is from a good friend of mine, Nicole. She makes the most beautiful AI imagery I have. Ever seen. And whenever you ask her [00:25:00] like, girl, how did you make that? She's I don't know. I use so many programs. And she like layers them. She's almost like a painter. Like she'll start in mid journey, then she'll go to Higgs Field, then she'll go to like nano banana, then she'll go, she just bounces around. To different generators, and comes out with some stunning,really cool AI art pieces. And that is really the heart of creating an app and integrating AI into that app. It is the layering. It's not just, I made an app where you type in your business goals and it gives you like a calendar output of what you should do for the year to reach those goals. That's lame. No, the power of integrating AI is that you could have you could input your goals for the year. Then have AI analyze the recent six month trends of Google Trends and analyze the recent reports of you [00:26:00] in the news in the last six months. And then use AI to combine all that data with your one year plan, and then put out something that is hyper specific to you, and then not only that, you can make that app repeat that monthly so that information is up to date. and that is only two layers. That is so like what I just described is very cool, but also just that's so basic. That kind of multi layering though is where we're going next and what you can do with Claude Code and what companies who are bringing you apps are going to be able to give you. It will be where you put your information into an app, one like the Toy Matchmaker, the output you get is so astounding And when you go to try to replicate it in Gemini, you can't get the same data and you don't understand why. It's because of the layering. I truly believe that is actually going to be the IP of the [00:27:00] future. the prompts that you use in the way that you layer them to output information is going to be the new ip. That is going to be what needs to be protected. Okay. I wanna give you one more example in case that one didn't land. So Imagine you put in your product data and one AI tool is analyzing that data. Another is analyzing nearby demographic data. Together. They're figuring out how that product fits in with that nearby demographic data. Then they're figuring out the best way to get your product in front of the store owner, and they're figuring out who that store owner is. That is AI and data APIs all working together, that's like what we do with the Toy Matchmaker. to turn what is normally like a week's worth of work into five minutes, so you can focus your energy on making the best product, making the best content to sell that product, and then making the most personalized pitch to get a conversation with the owner of the store that you wanna sell to. [00:28:00] So essentially your business is gonna win when AI is working in the background of a tool that you're using. Whether you built that tool yourself or someone built it for you, that is the shift.A while back I did a mind map of where everything is going with ai and we talked a lot about existing companies that are gonna be integrating AI into their systems. One company I will bring up isClickup. They are trying to like force speed AI to all of their users, to the point where it's getting really frustrating to use the app because you're almost accidentally. Using AI and getting popups that say, upgrade to use this feature, and you're like, I didn't even, I don't even wanna use that. I don't, I'm not looking for that. And I think. There is a detriment happening where AI is getting integrated into existing tools. these tools where AI is being integrated after the fact are not gonna work as well as the tools that will come out with AI integration. From the forefront because they will have been thinking about AI the entire time and integrated it into [00:29:00] all of the aspects of the product instead of it being plugged in and like having bugs within your regular data. All right, let's bring this home. So Claude Code is genuinely one of my new favorite things. I'm obsessed. And the fact that look, I'm a toy person. I've built websites. Since I was 10, and I do still build my own sites. I still do some coding, but it's so much easier these days with ai. I used to have go to coding websites and look up like what is the code to do a link on text HREF. Like I, it was so annoying. It's so much easier today. We are going to change the way that people sell their toys and games to retailers, and we are gonna change the way that retailers find toys and games that fit with their products and their customers. I am so excited. Again, go to the toy matchmaker.com. If you're interested. Get on the wait list, learn more. Also, if you wanna try Claude Code, go to claude.ai. Open a project, start playing around. You don't need an app on day one. You just start small. See [00:30:00] how it thinks, see how it builds. Get curious with it. If you're a toy entrepreneur who wants to skip straight to just the system that is built for you built to help you grow. If you wanna find the right stores to get your product in, pitch them with confidence and know exactly the right approach to use while doing it. Again, go to thetoymatchmaker.com. We are in beta right now. If you want in early submit your information. I'll take a look at your, I'll take a look. Submit your information, All right. And I would like to leave you with a question to answer in the comments wherever you're listening to this podcast. What would you automate if you could, in your business today? Let me tell you something, there's a pretty good chance that the tools you need to do that already exist. Okay. Wherever you're listening, leave us a rating and review, leave us a comment and some love. As always, thank you so much for spending this time with me today. I know there are a ton of podcasts out there, so it truly means the world to me that you tune into this one. Until next [00:31:00] week, I'll see you later. Toy people.