The Startup Ideas Podcast - How to get 10x more out of Lovable, Bolt, Cursor

Episode Date: June 9, 2025

Join me as I chat with Meng To, Founder of Aura, to discuss how to elevate AI-generated designs beyond generic templates by using proper prompting techniques. He demonstrates how providing AI with specific context about layouts, typography, colors, and animations results in more professional and unique designs. The conversation highlights how thoughtful design creates user trust and can be a competitive advantage for startups. Timestamps: 00:00 - Intro 00:32 - Giveaway Winner 1 Announcement 02:37 - Problem with generic AI-generated designs 03:22 - Developer-focused vs designer-focused AI tools 05:03 - Explanation of design framing and styling elements 08:06 - Demonstration of prompt builder and results 19:17 - Advice for product builders 26:18 - Meng's journey building Aura Key Points: • Current AI tools like V0 and Lovable create generic designs that lack uniqueness and taste • Using prompt builders with specific design vocabulary helps create more professional, unique designs • Understanding design elements (typography, layout, colors, animations) is crucial for effective AI prompting • Better design increases user trust and allows creators to charge more for their products LCA helps Fortune 500s and fast-growing startups build their future - from Warner Music to Fortnite to Dropbox. We turn 'what if' into reality with AI, apps, and next-gen products https://latecheckout.agency/ BoringMarketing — Vibe Marketing for Sale: http://boringmarketing.com/ Startup Empire - a membership for builders who want to build cash-flowing businesses https://www.skool.com/startupempire/about FIND ME ON SOCIAL X/Twitter: https://twitter.com/gregisenberg Instagram: https://instagram.com/gregisenberg/ LinkedIn: https://www.linkedin.com/in/gisenberg/ FIND MENG ON SOCIAL Aura: https://aura.build X/Twitter: https://x.com/MengTo

I found a hack to prompt lovable bolt V0 10X better. And it's this tool that no one is talking about. I brought the founder of that tool to show you how to prompt for UI. And this might be an episode that for people who watch, they're going to be able to charge more. They're going to get more trust with the products. And their vibe coding products are just going to get absolutely stunning looking gorgeous designs. And this is going to be an unfair advantage for you.

All right, Colin. Colin David Kennedy, the guy who just won 5K. Dude. Hey, what's up? You won $5,000. That's sweet. That's awesome. What are you going to do with it?

I'm going to probably actually pay for Bolt.com new to get started. Yeah. All right. And what are you building? I'm staying on your goals throughout the week is my goal for it. Cool. And did you think you were going to win the 5,000? Like, how stoked are you?

All right, man. Well, appreciate you. Meng, by the end of this episode, what are people going to learn? Sure. So I think there's such a problem right now with AI creating generic content and creating generic designs. And I think what do you need to learn is how to do it with taste. Because ultimately, that's what. makes us a human and unique from AI. So you're going to be learning how to prompt the right way,

how to find the right resources to prompt, how to find the right context to give to AI, the right design resources, the right taste, such as typography, colors, visual hierarchy, all of these things that bring your design to the next level and not just something that is generic. All right. Let's get into it. The way that I see is that right, right now, everyone knows V0, everyone knows lovable. And while I think it's great, because we're all unifying around a platform where you can build any product that you want, at the same time, you know, it's very developer-centric. And most of the, like, what they don't show you here is 99% of the results are very, very

generic design, visual design, you know, I can pick anything random, but it's not the most design-oriented result that you're going to find. And, you know, I think what I'm trying to bring with my videos, but also my platform, which is called ORA, is to make something a little bit more design-oriented. So something that someone who's really good with Figma would be able to go and prompt something using a prompt builder. So you're going to be able to kind of adapt from existing templates. So we're like having thousands. Of course, yes, V0 and all those they have templates.

But as you can see, they have a very different flavor. And most of them are very basic and very generic. The same with V0, to me, V0 is kind of like more for developers and lovable is more for designers. As you can see, it's not the most attractive. You're not going to convince designers who's been, you know, like not 10, but 5 years into Figma and then move into this and try to convince stakeholders to have a seat at the table.

But it does give you the power to have a startup, to found your company and to make something in a weekend. So what I would like to bring is, you know, As someone who is VipeCoding a bunch of products, such as ORA, but also a dream cut, both of them have over 150,000 lines of code. So these are months and months of work, and I think I have a unique perspective. So, you know, you need to give a lot of context to AI. And you need to sort of train with taste. So what are the designs that you can be using?

So, you know, what are the kinds of layout that you want to be building? And how do you expand your vocabulary in terms of, you know, what layout do you want to bring? And like, is it a document site or is it an updates page? Is it a portfolio? You know, are you building a pricing table? And also within that, you have many flavors of layout, such as is it a list? Is it a grid? Is it a sidebar or is it a bento layout, which is quite new?

And Apple has been using that a lot and everyone's following suit. But you can see there's so many terms that you need to be getting familiar with and how you're setting the separation of the layout, such as one third, one fourth, and then you have to fill the grid layout. And then you have also the framing, which helps so much in term of marketing. because marketing, if you think about it, is all about design. It's all about user experience. It's all about what people feel when they see your design, right?

It shouldn't feel generic. It should feel unique even though you could be spending two hours on it. You could be spending weeks on it or months. But they don't need to know that. They just need to know that you're building something and that you have that special sauce that allows you to make it. feel unique. So that's why framing is also another important aspect. That's why a lot of people in Instagram and Twitter, they will show not just the design in the full screen mode, but also the

design in the frame. So they have a browser or they have a card and they have a lot of bleeding around the card. Sometimes they have like a clay. And so, yeah, get familiar with all of these. And this prom builder allows you to set those things visually. And then you have the styling, you know, is it a flat layout? Is it an outline layout? Is it iOS or material design? You know, is it light mode, dark mode? And then you get into the primary colors. And of course, I can go on forever forever. But, you know, you have typography, which is extremely important. 80% of all designs are text. So you want to know what are the most important fonts, at least, you know, to know enough to be dangerous. So what are people using the most right now? Well, Inter and then Geist and

then Manrobe. And then, you know, what is the serif font? And then we also finish with the animation. So you want to know what kind of animation that you can bring into your design. So, you know, also what is very common is sequence animation. So Apple does that a lot. So it doesn't just all fade in. just one by one element by element and sometimes word by word or letter by letter. So all of these things, you know, you bring it out together and then you add it to a prompt. So let me just select one of the designs here. And then you add it to a prompt and then yeah, you know, this is your entire prompt. It has a ton of context.

And if you do this properly, yes, you can write it as you get. more experience, you can write it. But if you do it properly, you should have some really amazing results compared to just doing from scratch. And there are so many little nuances, little references. You should definitely name the websites that you love the most. And, you know, and again, it's something that you build over time. It's something that you get better. and the more vocabulary that you have, the better you can direct the AI to do the right things for you. And so you can see,

instead of having something that is extremely generic as this, for example, you have something that is a lot more flavorful, but also unique because you brought a lot more nuances to the table. And this is definitely not something that you're going to see. And, you know, like it's a very specific font, It's a very specific layout for the pricing, for example. You know, and then, yeah, you get into the details. You get into the animation.

As you can see, it just happened. And I didn't have to put weeks and months of efforts into this. I just needed to set it up in a prompt builder. And there we go. We have the result right in front of us. Yeah. So this is why I wanted to have you on because what's happening is a lot of people are listening to this show, thousands of people are prompting the V-Zeroes of the world, and they're getting

the typical Shad-C-N templates, and they're posting it on Instagram and Twitter and stuff like that, and people, you know, their conversion rates are very low. And one of the reason I believe that the conversion rates are low is because, you know, the UI and the designs are not resonating with people. And what this is, what's so cool about this, is it's almost like training wheels. Like it gives you, you know, it's training wheels for design in a lot of ways. It's saying like, okay, I'm not ready enough to go into V0 as a product manager or someone who doesn't know the terminology. I need a little bit of help acknowledging that. I need a little bit of help and some guardrails around what, around basically what to prompt it so that.

I could prompt it in the best possible way. Exactly. And funny that you mentioned Shatzian because it's one of the most common templates that you're going to be looking at most of the time when you're going to prompt V-Zero. It's very much embedded into the platform. And in Lovable, you're going to see something

a little bit more like shadows, but it doesn't mean better. I think it's just one, everything looks the same. the other one is more like maybe there's too much shadows and maybe there's too many gradients, right? Like you look at something, you know, like this one. Someone who's not trained would look at this, it's not too bad.

But someone who's more trained looks at this. You know in the first second that it's not made by someone who knows you why and that's someone who cares. And that's the key word, right? You want someone who cares about the product. and how do you prompt in such a way that you can communicate the thoughtfulness of the stuff that you're trying to design? And it's as little amount of time possible.

So you have something that looks like this, right? And again, I'm not putting more prompt effort into something like this versus something like this. yet I can build something that feels much more, you know, like it elicits the emotions that I want people to feel and the carefulness that I want people to feel. And so, you know, like understanding all of these states, all of these nuances, the subtleties of the outlines and the drop shadows and getting the right drop shadows. versus the wrong drop shadows means so much. So, you know, if you look at all the things that, you know, I'm putting out there, we have thousand, like more than a thousand of them, like this one. You're not going to find something like this, you know, on V0 or not saying that it's impossible,

but definitely it's a lot less common. And that's what I try to bring to the table. And then you also want a post tool, like a post design tool, that allows you to make the right correction and the least amount of time possible. So right now we have Figma, of course. But with Figma, first of all, they're still super new into the AI world. So you're not going to see people, in fact, Figma make just launch, right? but it's an amazing design tool. So what I try to bring is, for example,

you can just quickly just change the text like this. And you can quickly change the font. And I just put like the best font that people can use like this. And then I also bring something like image. So if people want to use unsplash, which I highly recommend when you're starting out, you don't want to have to scour the internet for like the best images. So you want to have the best tags.

Again, I put the best tags here. You want to use like render. And then depending on your design, let's say you have a blueish design. You're going to go to the blue color. And then based on relevance, you're going to select something as beautiful as this. Again, adding the uniqueness to the design that you're creating. And again, we're not spending hours doing this. We're just spending seconds and minutes to edit these things.

and increasing the quality of the output and setting, making it, you know, adaptive. So we have also the code mode. So you have, you know, the desktop. And you can see it in iPhone, it works just amazingly well. And also a feature that there's no one else does it is something that I want to add. This is all vibe coded, by the way. So we are on the same page here. You know, if you're a vibe coder, you make websites.

We're doing the same thing right now. Yet I'm able to do these things that are amazing and no one else is doing it. So for example, I click on the text, boom, it gives me the text and now I can just edit from here. So yes, it helps to know code, but the fact that you can just have a tool that is so friendly that you can just edit directly the stuff. And within the context of code, because we're all using code, I think that's so incredible. be powerful. The last thing you want is to make things unforgiving, right? Code in the traditional sense is very unforgiving. You make one mistake. It's going to give you an error. In fact, you know, in the early stages of V0 and lovable, I don't know if you remember your first project,

half of the time you would spend like with an error, right? Like you have to, you know, put that and prompt that and then sending back and forth. Now it's getting a lot. a lot better, but you do still get these errors. And when you do get these errors, you know, a lot of people actually give up, which is really sad. And, you know, I think we've come such a long way and having a tool that makes it so easy to both design and code without necessarily the years of experience of both design and code, I think that can be really powerful for a lot of people. Yeah. And the other thing I want to just say about the powerfulness of it is it all. also allows you to charge more when you have better design.

Because when you see something like this, like what we have on screen, it just invokes more confidence. Like I look at it this and I'm like, you know what? You know, I trust Sarah Martinez, the product manager, you know? Like she, she seems like, you know, this feels like a, you know, an A16 Z backed company, you know? And that's amazing. Like sometimes you can charge two, three, four, five times more when you have.

have better design. Absolutely. And also without the technologies that, you know, all of these competitors have way better than you. So for example, if I'm competing with V0, well, obviously they have millions and they have, you know, it's a company that is worth, you know, hundreds of millions and also lovable and also Figma. But I don't have the infrastructure. It's just me. So how do I set myself apart. The last thing you want to do is to follow exactly what they're doing. That's the last thing you want to do because you're not going to set yourself apart. So the tools that you have in front of you, of course, there's marketing. If you have like a huge channel on social media, regardless of how bad your product is, as long as you have the functionality,

people will buy it. But the people who don't know you, those are the hardest customers. And those people, they look at the good design and they want to. want to trust it. That's why we go on a website like Apple and we're like, take my money, right? Like we're just opening the wallet right away. And that's because we have a design that is so trustworthy. It feels like the people who makes it, they care about pushing the envelope and they care about, you know, putting something that looks and feels great. So how do you set yourself apart in this world where you're competing with all of these companies that soon, if they don't make themselves better with AI, will become dinosaurs, which gives you an opportunity

to set yourself apart. So you have marketing, you have design, you have the functionalities, of course, and all of these things that you juggle with, you need to spend as a little time as possible without necessarily communicating that you don't care about it. I think that's the biggest thing. You can spend as little time as possible, but you can still care as much as you can. And I think that's really, really powerful.

Maybe we can leave it at this. What advice do you have for, you know, the people who listen to the show, it's called the Startup Ideas podcast. These are people who are building products, often multiple products. What advice do you have for them in order to increase their likelihood of success?

Right. So, you know, I think going back to design, when you see a good design, but not just a good design, because you can have a good design, but that is generic. So, right, and this is not saying that Shatsian is bad. Yeah. But anything that is a template is bad, right? And that's, and you can absolutely make Shat Sian something unique. But the problem is if you don't push the envelope, they're all going to look like this. And it is by design, by the way. This is not something that is not intentional.

This is all intentional. It's a template. It's a standard. But with any rule and with any template, you need to figure out a way to make it yours. Right. And that's why everyone copies Apple. That's why everyone buys Apple without a second thought. So my first thing is try to learn the vocabulary,

try to use as many apps as you can, try to follow as many people that cares about these things as much as you can, try to find the best designs out there, either from Instagram, from dribble, from Twitter, and follow these people, collect these resources, collect your favorite design. Let me try to refresh this. But you have so many platforms where you can find all of these resources that can help you.

And now in term of prompting, you should absolutely find resources such as 21st Dev where you can bring the code and bring that into your project. Because, you know, AI is very limited in the amount of context. It's very trained in a certain way by Claude, by Open AI, but it's not trained to be unique. The only way it can be unique is if you tell it and if you handhold it and if you give you the resource and then reference to your favorite website and the best resources to be unique. So it's not impossible.

It's just that you have to take that little extra effort. So, you know, a lot of people, you can see like, Something as, you know, I wouldn't say dumb, but something as small as just like, you know, a shining background with a little bit of a gradient moving can add that little umph to your product. Or, you know, very small details. When you use Open AI, they have this text gradient as it loads. Again, small details, but you care about it. And then you find a code for it. And then you copy the code. You copy all the code. you bring that, you paste that into your project, and then you prompt from there. And again, we have more resources such as CodePen. And these are not new resources, right? These are all the open source community that has been around for decades.

And so you're following someone as awesome as Jay, for example, who has a lot of amazing JavaScript experiment that he shares online. And then you can look at the CodePen, you know, understanding the different technologies that gets there, such as is it HTML CSS, is it React, is it, you know, is it iOS, is it Swift UI, and all that stuff? So you can absolutely copy the code. You can find libraries of components and buttons and loaders

and background animation. And then, you know, again, depending on your level of comforts, this is one of my favorite called React Bits. so you can have like particles you can have like fonts and text animations and blur effects all of these things that you can bring with just the power of a single prompt attaching the the library code or whatever component that you may have so absolutely you know it's it's not something that you're going to learn overnight it's something that also it's not even something that you absolutely need right now to know all of these resources.

In fact, if you just know one of them, that is enough to get you started on one project. And then the next project, you move on to the next component that you always wanted to build or next animation that you wanted to build, and then you bring that special sauce. And then the next one. So again, don't feel scared that there's so many things happening at once. Just feel like find one thing that really drives you and be. bring that and make it really good and really unique and, you know, take your design to the next level. And that's, I think if you do that, half of the battle will be done, right? Because marketing

is a lot about design. And if you do good design, the product is going to market by itself. Yeah, absolutely. So I have a little confession I need to make. So, Last night, I got a DM from a co-founder of a $15 or $20 billion company. And he was commenting on a new project I launched called ideabrowser.com. And if you go to it, it's ideabrowser.com. Yep. He was like, I really like the design and like the subtle details. You're going to probably hate the design because you're a designer.

But what my confession is, there was no designer on this project. I believe you. This was used with 21st Dev, which is a resource you talked about. Like, you know, the little details do matter. Like the little, like, you know, for example, the logo is kind of like moving a little bit. And, you know, if you hover over the number one software to spot trends, like right there, Yeah, there's like a little, you know, things move. This is the thing that that moves the needles so that you end up getting these DMs of people you look up to.

Yeah. And they're like, the design is sick. And like, I wanted to tell them like there was no designer on this. I absolutely agree. And, you know, believe it or not, aura took me probably like a week to build. And that's insane because we're competing against like V0 and Figma and all that stuff. So you see a lot of features there. But obviously I started with a Mac app and that didn't take off, right?

That was my first Mac app. I vibe coded this thing to death. And this took me three weeks, actually. But then I moved this to a web platform because nobody could install the app. and I had some trouble with Mac performance and stuff like that. So I experimented with it. I also build a really big, you know, intro to prompting. And this is actually where my product took off.

At the beginning, nobody cared about the Mac app that I did. Because, you know, I started this project as a voice to chat. Even you can even see it from the design. I remember it. Yes, voice to text and prompt. That was the first idea. And right away, I just pivoted, right? Like, you know, you have to anticipate this kind of stuff.

Like, you will start with one project. You will give it a try. And it might not work out. But it doesn't stop you. Just don't give up. Keep using it. And be the first customer, solve your own problems. eventually this became, you know, as I created this app, it became an HTML generator, right?

And I thought it was so good, so powerful that people had, you know, to use it. And I use it myself. In fact, I created like 400 designs from this. You can see at my profile. I have so many designs that I created. And, you know, it requires someone who has so much. much curiosity. So when I build this, it took me three weeks, the Mac app.

And then eventually I was like, okay, I still need a website. That's where people are going to buy this thing. I still need a pricing page and all that stuff, a download page. And then I need a learn page. And I also need to let people share their creations. So I built all of this stuff while I was building the Mac app. And eventually I took one week, one single week, and I brought all the features that I built from the Mac app into the web app. And here we are. This is what we have.

Sure, in the two weeks later, this is where I came up with the, the first week I had already the prom builder. I already had the generation stuff, you know, to be able to generate. And also I already had like the live the live preview as you generate. So nobody else has that, by the way. When you're generating, it's going to live preview everything you're doing. You don't have to wait like five minutes for something to show up. But I didn't have like design. I didn't have code yet.

So this was the second week. And this was like third-ish week. But I was building it and I was adding a lot of features to it. And now in a matter of. out of three weeks, right, we have something that is extremely functional. In fact, we're doing 500 users a day right now. 500 users are signing up to this platform every single day. And we're, you know, we just started, obviously,

but we're making something like 3,000 MR at the moment. And we're growing a lot, as you can see. Like, you know, just by the number of users. And I have done zero promotion. I have done like zero ads or ad words. Yes, I do have a social channel. Yes. You know, one of the big things, as I mentioned before, is that I'm someone who teaches for a living.

And I have design code, which is a platform that teaches design and code. but it wasn't design code that brought the users. It was this typo-paracry prompting or tips of prompting. And this video that I create on YouTube, which got like 9.2,000 likes on Twitter and already 50,000 views on YouTube. But most importantly, it's not just the views, is how many people respond to it, is how many people give you feedback. And that's how you build a product, right? you know, you will fail at first. You will pivot within a week.

You will add new things on top. You will be the first customer of it. You will solve your own problems. And eventually you have something that might seem something that is worth spending 40 hours a week on. 100%. This is a great place to end. I mean, I'm not affiliated with this at all. You know, the reason I wanted to bring you on was because I like to give unfavorable.

fair advantages to people who listen to the show. And to me, great design is an unfair advantage. And ORA is like a hack, right? It's basically a hack to get beautiful design for people who are not designers. So, you know, you're going to get, you're about to get way more than 500 users because, you know, a day. Because a lot of people are going to listen to this. I encourage people to sign up, check it out. I'll include the links in the link in the show notes for people to click.

I'm also going to include Meng's social on his different platform so you can follow him there. Meng, is there anything else before we head off? You want to say? Yeah. So again, thank you so much, Greg, for having me. It's great. For those who don't know, Gray and I goes back way back like 15 years ago. I was doing freelancing.

and I met Greg in Montreal. We both move out of Montreal. So now I'm in Singapore. Greg is in Miami, I believe. That's right. And, you know, next thing you know, we're both creating startup. So that's amazing. And it's quite a coincidence.

But back to, you know, design itself. Like I said, you know, please try to don't give up on the first prom. Try to use the prom builder. It doesn't matter if it's from V-Zero or not. You can even use the prom builder from my side and just learn about all of this and then go to your favorite platform, lovable. I don't really care if you're using Aurora or not, but I do want to give you the advice that you should not just depend on like creating a beautiful card. Of course, you know, it's all going to look the same or a beautiful site, right? What you want to do is, you know, use a prom builder, learn about the prompting tips, do them remix super, super important.

There's so many ways to do a remix. So, for example, you can click on a design that you like. You remix it. And then you adapt it. So you adapt to like a front page. You adapt it to a pricing page. And then you also want to, you know, to attach an image. So image to HTML is very, very powerful.

In fact, I do use 03, which is one of the best models for taking an image, a screenshot of your favorite website. or your own design in Figma and then turn that to HTML and then bring that to lovable, to V0, or to Aura, whichever you want. So there's so many ways to prompt, not just typing. You can import from Figma. You can attach an image. You can attach an HTML file. You can remix an existing project.

And all of these design that you see here, they're free. You can go here and you can download for free. So you can copy the code, save the HTML file, and bring that to your favorite platform. So that's just, you know, FI, I think is going to be really useful to a lot of people. And, you know, if you have a chance to try Aura, that would be my pleasure. And if you have any questions, please let me know. And again, thanks you so much to Greg for inviting me.

And I hope to see you soon. Thank you for the sauce. I appreciate it. Now everyone knows how to prompt for UI. So thank you, Meng, and I hope to see you around soon. All right. Thanks so much.