I Spent One Year Building A SaaS For Designers. Here Are 4 Tech Stack Mistakes I Made And What To Choose To Build Yours Faster.
Last year I focused on building Designvote — A platform for designers to get feedback faster. It got to have 500 users, but I made rushed tech decisions to build faster.
In the first month, everything went well, but my tech stack decisions started to weigh down on me after a while:
Mistake #1: I used Create React App
CRA is excellent, but there are many limitations with a SPA. Next.js is by far better. Here are just some benefits:
- bundle size
- page load speed
- SEO tooling
- image optimization
Mistake #2: I created my own backend REST API server
The problem with classic servers is that you also need to implement: authentication, bug tracking and reporting, admin dashboards for user and content management. This takes an incredible amount of time.
- secure by default
- authentication implemented
- admin dashboards provided
- cheaper than traditional hosting
And many MANY more.
Mistake #3: I didn’t use pre-made components
I wanted to style everything myself, which takes a lot of time. There are great solutions like Tailwind UI or Pro Chakra-UI, which offer pre-made, well-designed, responsive by default, and customizable components for you to use.
Once I bought Chakra Pro, my development time was cut in half because I wasn’t wasting time experimenting with layout and “how bold should this text be.”
They cost money (around 200$), but you pay once and have them forever, and I guarantee you get your money’s worth.
Mistake #4: I tried to learn a new technology
When I first started, I decided this SaaS would be an excellent opportunity to learn Clojure. I love Clojure, but creating a new business was not the right time to start because you should focus on problem validation and getting your app in front of clients, not figuring out CORS in the Clojure context.
When building a new business from which you want to make money, stick to technology you know.
This post was created with Typeshare