Transcript
A (0:00)
Today I'm going to show you something every marketer should be doing right now. You should be building free web apps to generate emails or phone numbers for your sales team using AI coding tools. I'm going to show you how to get the ideas. Take that idea and turn it into a full product spec and actually build out that product spec using Claude artifacts so you can get an MVP live that you can share on your website or on social media Today. Let's get into today's show. Before we get into today's show, here's a quick word from HubSpot. Cutting your sales cycle in half sounds pretty incredible, but that's exactly what Sandler training did with HubSpot. They used Breeze, HubSpot's AI tools to tailor every customer interaction without losing their personal touch. And the results were pretty incredible. Click through rates jumped 25%, qualified leads quadrupled and people spent three times longer on their landing pages. Go to HubSpot.com to see how Breez can help your business grow. All right, hey everyone. On today's show, I'm going to show you something that every business should be doing today with AI and that is using AI code generation tools to make free web applications that you can use as lead magnets for your business. One of the biggest opportunities you have is to spend a few hours, or if you want to make it really good, maybe a few days to build a lightweight web application that will make and create a ton of value for your customers that you can give away for free. Maybe in exchange for email addresses, phone numbers, some way to build that website visitor or social media follower into a relationship where they could become a customer for you. And so I'm going to walk you through all this today. And I thought the best way to do this was to pick the most kind of basic and universal of small businesses to be an example. But understand that you can follow this process for literally any business, any startup that you have. So the first thing I did is I went to chat GPT and I went to chatGPT because I wanted to get some ideas. Cause it's not just like, hey, go build this thing. I want to build something good. And so here's the initial prompt. I said I have a friend that has a landscape design business and is looking to improve his online marketing. He wants to offer a free web app to attract customers and collect email addresses for marketing purposes. He's based in Columbus, Ohio and is looking to serve customers and businesses within a 50 mile radius. Can you give me some ideas for the app to build. I did this with the O3 app within ChatGPT. And you know what? It came back and it gave me seven ideas, all of which that I thought were pretty good. This is my favorite one. An AI yard vision photo mockup where it takes a picture of your yard and uses AI to generate new renderings of it. That was going to involve me getting API keys for OpenAI and doing a bunch of stuff. So I didn't do it for the purpose of the show, but if I were actually probably doing this for real and I had a landscape business, this is probably the thing. It would probably take maybe a week or two of some vibe code to get there, but would be a really compelling way to do it. There's a lawn health diagnostic quiz project budget estimator, seasonal planting calendar. Who doesn't want to know when to plant stuff? Storm damage, risk map, water use ROI calculator. That I would have never, ever thought that. Honestly, I thought that was pretty good. And then curb appeal scorecard. So if you have like an HOA or small businesses or you're selling your house, like, do I have a good curb appeal? The drive by when you pull up, is it going to be looking good? Because I have, you know, some type of commercial incentive to care that it's looking good. So I actually decided I was going to try to do two of these on the show. I did one ahead of time, then we're going to try to do one live. The one I tried to do ahead of time was the project budget estimator. Because one of the things, if you, like, look at your backyard and you're like, oh, my backyard doesn't look that good. I'd like to do something about it, but, like, how much would it cost? And I'm really ready to do that. What's that look like? That's like a big hurdle of friction in the process. It even gives like a distribution playbook. It's very good, right? Like you would go back and forth and refine it and get it better. But for the sake of this, we want to take ideas and building an app. And so I basically asked it, hey, I love number three, which is that budget app. Could you build me a prompt and a spec for a basic version to build an AI coding platform so that I get an early version live to test. The business name is Leaves of Grass. I made it up. Shout out Walt Whitman. You know, I thought it was a good landscaping company name. So what I did here was do something very, I think, important in this Prompt which is like, I want a very basic spec because I want to test if this idea is actually good. One of the advantages you have with these AI coding tools today is that previously you would have been like, oh cool, I have this idea for this budgeting calculator. I'm going to hire some type of web development company to go and build this for me. And they're probably going to charge you 10, 20, 30, 50k to do something like this. It's going to take, you know, a couple months probably for them to do. And it may or may not be the right thing. It may or may not work. So it's a lot of time and upfront money to know if it is the right thing to do for your business. So one of the advantages of these AI coding platforms is that you can get something that you can put on your website or share with a small number of people to say, hey, do people actually use this? Does it work? Does it give people value? And if so, then I can either take time to build a deeper version or maybe hire somebody to build like a really thorough, deep, high scale version. Right? And so this is like, can I get to prospect, market fit? And does a prospective customer really like and want to interact with this? And so I asked it for that and it gave me a one shot build the app prompt. Okay. And then it also gave me a full technical spec for me and any people that were going to work with the prompt that it generated. And it says, hey, here's the recommended build sequence scaffold with the one shot prompt. Configure Supabase or sendgrid. This is like, hey, if you want to build the entire app, do all of the back end, have the email automation work, do everything it's saying. It's going to take you about four hours, which is not a short amount of time, but not a crazy amount of time for something that would have taken months before. And it tells you exactly what to do and you can then basically ask questions and how tos to each of these numbers and depending on your level of proficiency, couple hours to maybe a day or two and you would have this done. I want to show you how you would go and do it even faster. One of my favorite AI products out right now. And you're going to see lots of different shows from Kieran and I coming around. This is all on Claude's artifact feature. And artifacts are a way that you build and deploy web apps in Claude. What you want to do is go to Claude and you want to click on artifacts. So you have new chat, chats, projects, and now there's this new artifacts. And you can go look at artifacts people have built, you can look at the things you built or you can go to create a new artifact. And what's really cool is that you can build apps, a website, documents, games, productivity tools, quizzes. Remember that quiz idea that ChatGPT gave us? We could build it right there and you can publish a public version of your app. So without having to do a lot of the backend work, especially if it's just a prototype, if you have advanced features, you need an API and that kind of stuff, you're not gonna probably put it on an artifact. But if you're doing like basic web apps, artifacts are a really quick and fast way to do that. And so this is what I have built for our calculator. So I took the one shot prompt from ChatGPT, one shot prompt from ChatGPT and I just pasted into Claude and Claude built out all the components, wrote all the code, everything that you need here, and basically tells you, hey, here are the features that I implemented. Here's what I did. And then it sells me how to get started, like how to copy the files, set up an environment, run all of this. And I just did this in regular cloud. I didn't go to the artifacts feature like I showed you at first. It's like, hey, can you just create an artifact for this so that I can see how it works? And so it did that. Then this pops up over here, which is my real live web app. And it made a logo. If I had given it a logo, it would have used that. But this is my made up business. And it says, hey, instant pricing, professional PDF estimates, local Columbus expertise. Start your estimate. My square feet of my lot's actually pretty big. I'm going to say I got 10,000 square feet. You could upload an image of your property. That's optional. I won't do that. Now, a premium patio installation. I'd like a 500 square foot patio. I don't want a pergola. But hey, I love gardens. So I want a good amount of planting beds, lighting fixtures. Yeah, I don't want to walk around in the dark. I'm going to need eight fixtures in there. Sod. You should see my grass, it's horrible. So I'm going to need most of the yard sodded, you know, where I don't have the other stuff. Do I want a custom water feature? Sure. Who doesn't like water? And then I'm going to say get my estimate, put in a fake email, get my estimate and then it is supposed to send me the actual PDF estimate or I can say download a PDF, but here it is, here's the actual estimate it gave me. If I went in and gave it all these variables, it's saying, hey, it's going to cost $34,000 in two weeks. Obviously this is an estimate wouldn't be binding. You'd go in and you'd put all of your actual costs in and estimates in. But I did this in 20 minutes. You could see how a little bit more time and you could make something really good. And what's crazy is I can hit publish right here and I can publish and copy the link and now I can just copy this link and, and put it wherever we want. And like I can put it in the comments of this show and you can go and see that it's live and available. Yes, I can go and do all the hosting and build the web app and everything. But let's say I just have like 10, 20 customers or I have a page on my website that I want to put it on for a week to see if people actually use it and what works and what doesn't work with it. I can use this cloud project link before I go and invest in the next big heavy version of this, which is pretty wild and pretty awesome. So, like, you can go from an idea for your business for a lead magnet to having a live working app in 30 to 60 minutes. Now you got to get the email collection right, collect it in your CRM, use HubSpot, that whole thing. But that is what needs to happen. Hey everyone, we'll be right back to the show. But first let me tell you about a podcast that I love. It's a podcast called I Digress. It is hosted by my friend, Troy Sandage. It's brought to you by the amazing HubSpot Podcast Network, the audio destination for anyone in business. And I Digress is great. It's got 30 minute episodes and the the podcast is all about helping you eliminate complexity, complications and confusions in your business. It's really heavy in frameworks and strategies to help you scale and sustain your success. And you know, recently Troy had a really great episode. It was called Branding Secrets. Leveraging style to change the narrative, boost influence and profits. And that was with his guest toy, Sweeney. It's really good. You should check it out and you can listen to that or all the episodes of I Digress and anywhere you get your podcast. Okay, so Then let's do another one just for fun. So I go over to ChatGPT and I want to see if I could do this again potentially. I really like the idea of this curb appeal scorecard and so I had it generate the same one shot prompt. And if we were doing this for real and this was really our business, we would probably iterate on this prompt and the details around how you wanted the app to work and everything. You go back and forth on that More probably take 15 to 20 minutes to really get this prompt deep and right. But once you got the prompt right, you're going to hit copy and you're going to go over here to our friendly artifacts feature that I had showed you before. We'll click apps and websites and then I'm going to paste in our lovely one shot prompt here and I'm going to hit build it. Perfect. Let's see what you can do. Leaves of grass, Curb appeal scorecard, nvp. It's telling me what it's going to build. The user experience flow, the strategy, everything there. And I'm going to say build this in an art effect for me. And now it's actually writing all the code. You can see it here, generating the code for our app. And so an artifact really is just a web app where the code is written, contained and hosted within Claude. And I suspect Claude is going to keep making artifacts more and more powerful. And for a lot of one off use cases, small business use cases, individual use cases, you're just going to end up using artifacts versus like going and hosting a full dedicated web app. Because I think Anthropic is going to make that product better and better. They'll probably charge you for the hosting and some of the costs of goods sold associated with that, but we'll cross that bridge when we come to it. So it is generating, it's writing, like I said, even though you're writing, quote unquote, an MVP app, it still has to write a good amount of code and is working through here. I feel like I'm kind of on like an old school cooking show, you know, where they put the thing in the oven and then they turn around and they take the done thing from the other oven. It's like whoosh. Except this time it really did because here is our scorecard. What's interesting is it used the same logo and branding as the other app and I didn't say to do any of that, which is pretty cool. And here it is. So you upload pictures of your property or choose sample Images and then you continue to your assessment. And we have a working app and we would upload photos, we would put photos in and let's see how it actually works here. All right, so I found a photo from Google of a house that had good curb appeal. I'm just going to see if it'll let me upload one and continue the assessment. All right, the uploaded picture. Now, how well maintained is this landscaping? Oh, it's excellent. Conditions of walkways are four. Maybe there's a couple little cracks. How much seasonal color is present? Oh, there's some flowers there. It's looking good. Outdoor lighting, it's five. Attractive is the signage and wayfinding. It's four. It's looking good. What's interesting is it created all these categories and did all of this for us. These are good things. Obviously, depending on your business, you would go and customize those and you could decide if you wanted this type of conversion mechanic or if you wanted to do something else. But this is a good example of that. And then it says, hey, you have a grade A photo analysis. Green coverage, 47%. Visual symmetry, 78%. Areas for attention. Debris visible, uneven edges. There's no debris visible in this photo. It's an early prototype of what we're doing here. But again, you can see that you would potentially hook up the right API to analyze the photo and do all of the work that you'd want to do here. So these are two apps that if you were a landscaper, you could go and build and test and you know what, likely increase the amount of people who are coming to your website that you're actually getting email addresses, reaching out to and trying to work with for landscaping services. And you would do this for literally any small business or any individual project that you are working on. I think cloud artifacts are critically important part of the up and coming AI coding stack. And as a marketer or as a founder, one of the things you're looking for is to validate an idea really quickly with your audience. And that's what today's show was all about. Can I, with the help of AI, get some ideas, bring those ideas to life and then either test them with an audience or test them with myself to know, hey, this is what I need to iterate on. This is the actual, like, technology that needs to exist for this to actually be a good experience? Experience. And we go and build it. We take it live to everyone. I would love to hear what you've been building with AI coding tools. If you're using Claude artifacts, publish links to stuff you're building below. I'd love to go and check that out. Thank you so much for checking out today's show. I'd love to hear how you're inspired to go and build some stuff for your business. Hit that. Like, hit that. Subscribe and I'll see you real soon. On Marketing Instagram SA.
