My AI development experiment: a gaming portal, SEO traffic and selling the project
Without a developer, I vibe-coded a game guides site, grew it to 1,000+ daily visits and 500+ sign-ups in a couple of months, then sold it.
Here's a project going into my portfolio: wtg.yason.studio, a site with video game guides and walkthroughs that I launched and have already sold.
It was an experiment: I built the site from scratch with AI, without a developer. In a couple of months SEO took it to 1,000+ unique visits a day and 500+ sign-ups. At its peak it even made up to $50 a month from ads.
Features in hours, not weeks
Building it was incredibly addictive: in a matter of hours I shipped features that used to take weeks with a developer.
- multiple languages (Russian and English);
- sign-in with Google and Yandex;
- a personal account for notes on guides, favourite games and walkthrough logs;
- a comment system;
- navigation that brings each game's guides, missions and characters together;
- a shared guest chat;
- automatic structured data, and much more.
The content machine
- Perplexity automatically found popular topics for guides.
- A content manager (humans are still needed) adapted and published the material, checked facts and moderated.
- The site filled in its own SEO data for promotion.
What's next
In the end I sold the gaming site and moved on to more serious projects: a comments widget for Tilda, a corporate website for an orthopaedic surgeon, and next, an e-learning platform. For now I have more ideas than time.
I'm gradually moving away from website builders and switching clients to custom-built engines: builders no longer speed up development, they hold it back. More on that in upcoming posts.
