Spacesoverview - Do things that don't scale.
Hi everybody 👋🏽
I'd like to share the story of how I started my second side project: https://spacesoverview.com
This post focuses on the beginning and the pure search. I create separate posts about upcoming features and my process of trying to acquire users.
Problem Statement
This time I had the idea of building a simple search interface for Twitter Spaces. I saw the release of the V2 API of Twitter and also saw the ongoing rise of Twitter Spaces but the only way for searching these spaces was going through the complicated advanced search of Twitter. For example with filter:spaces aws You could find spaces with the keyword aws. Somehow some other results are still occurring.
My idea was to create a simple interface for searching spaces by keywords. This is the final result:
Twitter API doesn't scale
I started out researching the Twitter API and found a cool endpoint for the exact use case. The endpoint /2/spaces/search gives you the exact capability or searching for spaces by a certain query.
Rate limiting on Twitter is really bad. You basically get 300 requests per 15 minutes for searches. I instantly thought about approaches on how to overcome this issue like caching popular searches, saving all spaces, getting rate increases, etc.
At some point, I realized that I want to validate this problem quickly instead of spending too much time on that project (again). So I just abstracted the API away, built it with the original rate limit in mind, and published it 🎉
This time I really followed the approach of Build things that don't scale
Goals
To go one step back. I had different main goals for this project. First of all, it was getting more proficient in a tech stack for my new job. I knew that I'll work a lot more with Vercel than I used to so I wanted to skill up. Building a project is the perfect opportunity for that.
The second goal was to build a go-to project that I can clone for future smaller projects. This project should include everything I need for a micro saas:
- Payment option with stripe
- AWS Infrastructure with CDK
- Frontend with Next, React, Tailwind
I accomplished both goals very fast and was super happy with the progress.
Analytics
How many people are visiting spacesoverview.com? I was amazed that the page was on number one search results for the search term twitter spaces search. Unfortunately, only on Bing & Microsoft. Google didn't even list my page. Maybe because I had Twitter in the name of my domain, so we'll never know.
I've connected plausible.io to my domain and had around 100 visitors each week. The main channel where people are coming from are:
- DuckDuckGo
- Bing
But see for yourself here:
I also created a Sign-Up and Sign-In flow with Auth0 within a couple of minutes that users are able to create an account and get future updates. I simply added a note with upcoming features.
50 people signed up without having any benefit purely for upcoming news. That is really great!
Tech Stack
I don't want to go too deep into the tech stack but what I currently use is:
- Vercel for Deployment and API
- AWS CDK for provisioning everything
- Next.js
- Tailwind
- SQS, Lambda (for upcoming features, see next post)
Domain Change
In the beginning, I had the domain twitterspaces.io. Many people told me that it is a pretty bad idea to have a brand name in your own domain that is why I just recently changed it to spacesoverview.com. I lost a few backlinks on search engines but that is fine.
Future Plans
I'll share more of my progress here on Indiehackers and on my Twitter. I already implemented another feature (reminder function) and will soon contact all my users to give them free beta access.
Do you have any questions or topics that interest you in particular? Let me know and I try to answer them the best I can! 🙂
