Category
- 🚗 The Invention Lab Story
Date of issue
Sep 11, 2025 2:00 PM
🧑🏻💻 Of course, at first, outsourced development!
To be honest, we never thought we'd develop it ourselves from the beginning. As always, our initial thought was, "Let's do the planning ourselves and outsource the development!" We even met with the outsourced developer. We drafted a detailed plan and discussed the features we wanted and the flow of development. As a result, the project came together much more quickly than we'd anticipated. However, problems arose when we couldn't proceed with the outsourced developer, and the project was temporarily put on hold.
🐥 So I set out to find it myself
Whenever we worked on an in-house venture or new business planning project, something I always told my team suddenly came to mind.
"Show your customers how your product or service works."
"At least make it into an MVP, not a PPT."
I felt it was time for us to personally uphold this principle we'd been emphasizing to our teams. "Okay, let's give it a try!" naturally occurred to me, and I immediately began looking for tools.
That's how I found lovable (lovable.dev).
The world was flooded with tools claiming to be usable without any coding knowledge. One of them was Loveable. Lured by the promise of "conversing with AI to create web pages!", I gave it a try, and it was truly amazing. Just typing a few lines of text produced a screen in no time.
첫 번째 인벤션덱AI
I thought, "Wow! We can do this too," but I quickly hit a wall. Implementing core features like AI integration, membership registration and login, storing user-generated idea data, and exporting results proved challenging. Furthermore, the service only allowed five requests per day, after which a paid subscription was required. Before committing to a paid subscription, I decided to try other services first.
A message that can be seen when you give the command 5 times…
🏃🏻➡️ From Loveable to Google App Script, Replit, and Firebase Studio
Aside from Loveable, there were so many tools that marketed themselves as “useable even for non-developers!” that I started trying them out with high expectations.
The first thing that caught my eye was Google App Script . I could generate code with Claude, copy and paste it directly into App Script, and deploy it right away. It was easy to integrate with spreadsheets, and visual results were easily created. However, when I tried to expand its functionality, its limitations quickly became apparent. While it was suitable for initial experiments and simple automation, I felt it wasn't enough to scale the service or ensure stable operation as users increased.
Next came Firebase Studio . As expected from a Google service, it was reliable in terms of stability and additional features. It offered authentication features, thoroughly managed security, and even handled hosting. However, when it came to building the app, no matter how specific our instructions were, we couldn't achieve the desired results. In particular, errors kept occurring when integrating the card data database with Gemini, leading us to conclude that Firebase Studio wouldn't be able to create the service we wanted.
The last one I looked at was Replit . As a non-developer, I was able to manage version control and hosting like a developer, and the implementation was more satisfactory than Firebase. However, as I used it, I encountered another challenge. Digging deeper into the platform meant I had to understand and modify the code myself. While it offered version control, I had to study how to utilize it, which created a significant barrier to entry.
# Advantages/Disadvantages of Each Tool
Tool name | Merit | Limitations/Disadvantages |
App Script | • Easy integration with spreadsheets • Results are visible quickly | • Reaching limits quickly when expanding functionality • Lack of stable operation when user growth increases |
Firebase Studio | • Stability & Reliability Provided by Google • Provides basic functions such as authentication • Thorough security management • One-stop service including hosting | • Difficulty in achieving desired results • Multiple errors when connecting to API |
Replit | • Easy version control/hosting • Higher satisfaction with implementation results than Firebase | • Deep functional expansion requires understanding/modification of code. • Existence of entry barriers |
Lovable | • Optimized for rapid MVP creation, front-end-centric | • Basic knowledge of Supabase is required (or payment required) when connecting to the backend/DB. • Stability deteriorates beyond certain scenarios. |
In the end, the results were the same after using App script, Loveable, Firebase Studio, and Replit.
"Each has its own distinct advantages and disadvantages. But we haven't found the 'right answer' we're looking for yet."
We continued to cross-use each service and conducted various tests to implement the service we wanted.
And our journey didn't stop with choosing tools.
Beyond choosing the right tools, we faced a bigger challenge: how to sustainably update our app's most important asset, its content, and provide it in a way that real users could actually use .
👉 In EP.2, “The world is turning and keywords are pouring in, the challenge of content creation automation” continues.
Subscribe to 'The Invention Lab Blog'
If you subscribe to the site, you can be the first to receive the latest updates, including new posts, via notifications and email.
Join Slashpage and subscribe to 'The Invention Lab Blog'!
Join Slashpage and subscribe to 'The Invention Lab Blog'!
