What I Found
The hidden fees became the product decision—and I still may not release ShowNext.
ShowNext began as a personal app built around my Trakt history, viewing progress, recommendations, and the question of what to watch next.
In the rough scenario I discussed, the AI concierge could cost around $2 per active user each month. That is already about $24 per year before App Store fees, taxes, support, metadata licensing, or any margin. In the video I questioned whether that pushes the real annual price toward $40 or $50 just to make the numbers work.
I also discussed an external metadata cost of roughly $150 per month. I referred to IMDb in the recording, while the current ShowNext build actually uses TMDB. TMDB does not publish a flat $150 commercial price; a monetized release requires a separate commercial agreement, so the final cost is still unknown.
Bring-your-own-key moves variable AI usage to the user, but it adds account and setup friction. A subscription is easier to understand, but it puts those unpredictable costs back on me. That is why ShowNext is not yet committed for public release—and it may remain a personal app.
What this video covers
- Why I built ShowNext as a personal TV and movie tracker
- Bring-your-own-key versus an in-app subscription
- How Trakt, OpenAI, and extra accounts affect onboarding
- The rough $2-per-user AI estimate and the $40-to-$50 annual pricing dilemma
- Why a roughly $150 monthly metadata example changes the economics of a small app
- How Season Encore limits token usage for large viewing histories
- The unfinished theater-data and AI-assistant pieces
The short version
A personal app can be inexpensive to build but costly to operate. Roughly $2 per user per month for AI, platform fees, and the roughly $150 monthly metadata example mentioned in the video can turn a small app into a recurring bill before it earns anything. ShowNext is not yet committed for release; the unresolved cost model is the point of the video.
