You’ve gone through phase one, right? You understand your workflows. You know maybe what you want, and you’re trying to get to the point now where how do I build this? How do I get this into my company and get it working?
And most importantly, how do I not waste a lot of money? How do I make sure it’s actually useful and it’s helping me and that it actually gets completed and maintained?
Do it yourself
Option number one is the easiest. I recommend this for anyone who has a little bit more time. This would be the best option because it requires time but it doesn’t require as much knowledge and you can start really at the beginning.
You have your idea of what your automation will be, but you need it coded. You need it converted to software. So you can actually use English, which is the most popular coding language now. You actually use English to just translate your idea into code.
What you get from building it yourself
There’s a lot of benefits of doing it this way. Number one is you start to actually understand what’s going on if you’re paying attention.
You can be involved in every part of the build process. And what’s great about that is that you’re not having to guess, you’re not being in the dark about where your project’s going.
It takes a lot of time
It’s still going to take you a couple weeks, maybe a month, maybe a month and a half just to crank out the initial version of the project if it’s a big one. If it’s a small little one, you probably get it done in a weekend, maybe a few days.
If we do it ourselves, we have to be prepared with the option of putting in 10 to 20 hours per week in maintaining it.
The part you should worry about most
When you’re doing it yourself and you don’t have a background in engineering and in computer development, you’ll tend to overlook things that you don’t want to overlook. And what these things end up being is helping you later on from not getting hacked, not losing all your work, and developing a project that can actually be maintained by someone else.
Because eventually, if you keep compiling more and more features without trying to compress and trying to build an architecture for your project, for your automations, what you’ll end up is with a very large project.
And it’ll just be way too big of a project and you’ll either pay for it in complexity, the AI won’t be able to figure out all the moving parts and it’ll start getting worse at its job. It’ll start adding features that are already there and so you’ll have duplicate of the same features.
Or on the flip side, it’ll cost you a lot of money because the AI will have to read through your thousand files that you’ve created and that’ll just turn into a token burning.
Where vibe coding projects break down
And that’s what we’re seeing a lot with vibe coding projects and do-it-yourself projects. They’re really great and they’re really cool and a lot of people are able to get a posted on LinkedIn screenshot version very quickly, within a few days.
But when they themselves try to take that project and make it a long term and make it a core part of their business, it starts to break down in these ways. And if it’s not breaking down, then they’re spending a lot of money with the tokens to try to fix it.
And while I always encourage people to get in there and fix it themselves, you want to expect the amount of time it’s going to take to really learn this system and figure it out and be able to maintain it.
Who should do it yourself
But I think it’s a very great start. And if you are a new business owner, I think that’s the best way to do it. Or if you have a little bit more time on your hands, this is the best option because you learn a lot about the process.
I think that if you’re starting out and you’re a smaller company, I think the do-it-yourself version is a really great angle. First of all, it’s free and you learn a lot. For the most part, it’s nothing that an experienced team can’t pick up after the fact. It’s a good start.