Objective
Build a zero‑setup tool for creators and analysts to paste a JSON dump and instantly filter by outlier score, views, VPH, publish window, length, and more.
Why I built Outliers
Scrolling through YouTube Analytics or spreadsheets to find breakout ideas is painful.
I wanted a one‑screen tool where I could:
- Paste the raw JSON I already pull from the Outliers/Breakout endpoint.
- See every potential winner ranked by outlier score.
- Filter down by publish date, views per hour, length, and channel size.
- Click a card to watch the video in‑app and grab instant context.
No log‑ins, no spreadsheets, no waiting for a heavy BI dashboard to load.
How it works
- Drop your JSON – Hit Load Data and the app auto‑parses the array:
const vids: Video[] = obj.data.map(v => ({
id: v.video_id,
title: v.video_title,
// …other mapped fields
}));
- Client‑side filtering – Five Radix sliders + a publish‑date dropdown give you fine‑grained control with zero extra network calls.
- Modal preview – Click any card to open an embedded YouTube player and quick stats (views, VPH, duration, outlier score).
- Sample dataset – No JSON handy? The Load Sample button fetches
/example.jsonso visitors can play immediately.
What I learned
- Keep setup friction near zero – a pasted blob is faster than OAuth.
- Radix primitives save hours of accessibility edge‑cases.
- Small JSON > DB for side‑projects; everything lives in local memory and still feels instant.
Next steps
- Export to CSV / Google Sheets
- Save filter presets in
localStorage - Tag & group videos for multi‑channel research
- Serverless endpoint to pull fresh data on a schedule