My business partner had been dropping increasingly less-subtle hints about how amateur our site looked and was threatening to fix it himself. With little to lose I signed up to Loveable and started prompting (or vibe coding as the influencers call it)… and it worked.
Although I consider myself a developer at heart, I hate building web sites. I could blame the complexity of developing for multiple screens, the finickiness of web building tools like Wix and WordPress, or the pain that comes with building and refining content. But the reality is, I'm just not that good at it.
I'm a data guy by training. SQL and the data bits of Python are second nature to me, but ask me to build anything with a front end that needs creativity visual design and I'll avoid it like the plague.
As I mentioned, my business partner has little love for our website. To be fair, I built it in Wix like so many others starting a business would - found a template I liked, scoured the net for some stock images, and pulled together a bunch of content. His suggestions for improvements were annoyingly fair but all required me to torture Wix (and myself) to implement them.
Enter lovable.dev… one of many new prompt-based coding platforms that promise to turn design-challenged developers like me into polished web developers. Lovable is also the platform that my similarly design-challenged partner had previously used to build and deploy a website in a pub… in the time it took to finish a beer [insert post link]. Naturally, I had to see if I could recreate this sorcery.
Spoiler alert - it worked (check it out here and see what all the fuss is about) and, I’d happily use Lovable again but… there’s some lessons I learnt that will make my next project much quicker.
You need a plan - You can't just throw prompts at these platforms and expect a good product. Think of AI coding platforms like an eager new hire who says "yes" to everything - desperately wants to please you and even more desperately wants to write code. You need to manage this enthusiasm by clearly prompting what you want AND how you want it to look.
It's not free, no matter what they say - Once you move beyond a coloured-in "Hello World," you'll pay money. The trick is thinking like a product manager and evolving in logical steps. Don't ask it to build Rome in a day - ask it to lay one brick at a time.
Take the refactoring offer - Lovable will nag you about refactoring code a lot. Do it. It shrinks the lines of code and reduces your token spend, which is basically the platform's way of charging you for being indecisive.
Test everything, all the time - These platforms will do exactly what you ask (except when they don't). You need to constantly check that your changes haven't broken something else. It's like managing a remote team, but with more debugging.
You still need to understand the fundamentals - Despite what the LinkedIn influencers say, you still need to understand databases, security, and how the web works. The people getting the most out of these platforms aren't complete beginners - they're developers who want to move faster. Take the time to understand what’s actually happening and the interdependencies between the components that make your app work.
Deployment is definitely easier - Getting a functioning site (security, data base, links that work etc) on the internet was painless. Even better, moving our domain from Wix to Lovable was a few clicks, not the three-hour customer service nightmare I'd braced for.
Final thoughts….
Was it quick? No. But was it quicker than playing with Wix? Yes, in about half the time we spent on our original Wix site we got a functioning website with a bit of design flair. Better still, we can tweak, change and improve the site much faster than playing with templates. Most importantly, my business partner has stopped making passive-aggressive comments about our site and I can focus on paid client work instead of web design.
If my experience is anything to go by, it’s an interesting point in time for developers. There will always be a need for technically capable people but I can’t see how we won’t be in a world where more people can get computers to do more things using human language rather than coding.
At JOURN3Y we’re working with businesses who want to understand the opportunity AI presents to change how they work and what they work on. If you’re keen to understand this as well, get in touch.