I've decided I'm going to embark on an ambitious project, and I'm going to do it in public. This is my first post outlining my plans.
I am an organic search specialist (SEO). I have spent much of my professional career learning how search engines work so I can optimize websites to rank better. This involves many different tools, most of them paid. One of my favorite paid tools is SEOTesting.
At my day job, I recently lost access to this tool. I was disappointed because there is nothing else readily available that complies with our strict corporate compliance requirements.
Since LLMs have flattened the learning curve for building complex projects, I figured I'd leverage Claude to help me architect an SEOTesting clone.
What is SEOTesting?
For those unfamiliar, SEOTesting is a web app which connects to Google Search Console. It allows you to create tests to track performance before and after implementing optimizations. You can also create content groups to permanently segment your traffic into logical site sections based on URL rules you set. It also houses your historical GSC data so you can see trends far beyond the 16-month window Google offers you.
Requirements
The MVP is going to be dead simple (a naรฏve statement, I'm sure). I just want to be able to connect to Google Search Console and display basic metrics--clicks, impressions, CTR, and position--in a dashboard view. Once that's working, I will get testing functionality and other quality-of-life improvements online.
Build plans
Here's the rough sketch of the architecture stack I'm planning to build out:
- Data source: Google Search Console
- Ingestion: Python script + cron job
- Database: SQLite
- API: FastAPI
- Frontend UI: Next.js + Tailwind + Tremor
- Hosting: Pi
- External access: Cloudflare Tunnel
As of now, that's all I have mapped out. Stay tuned for updates.
Top comments (1)
This looks like a solid MVP. Starting with Search Console data first is the right callโit's much easier to validate the workflow before adding testing features.