Clipmeapp/kick-clip-yield-2026

Kick Clip Yield Dataset (June–July 2026)

Attribution: ClipMe — https://clipme.com/research

Summary

Production-pipeline capture data from ClipMe, an AI clipping tool for live Kick streams. 51 capture sessions across 8 anonymized Kick channels, June 19 – July 14, 2026. Every number is measured directly from clip files on disk with ffprobe — no engagement, view, or engine-internal data included.

Dataset structure

Per-session table (CSV, kick-clip-yield-2026.csv):

• session_id, channel_id (anonymized), session_date, session_duration
• clips_delivered, live_clips (cut during broadcast vs. post-VOD)
• clip durations (median, mean, distribution buckets)

613 total finished clips, 173 (28%) cut live during the broadcast — the rest processed from the VOD after the session ended. Median clip length: 30.0s. 88% of clips land in the 30–60s band.

Collection methodology

Capture: ClipMe ingests the Kick live HLS feed and/or VOD directly. Every delivered clip file is measured with ffprobe for duration; session/channel metadata comes from the pipeline's own job records.

Selection: which moments become clips is decided by ClipMe's production ranking system — proprietary, not disclosed here. This dataset reports yield (what was delivered), not the ranking mechanism. No signal weights, thresholds, or taxonomy are included or implied by any column in this dataset.

Anonymization: channel_id values are opaque identifiers with no mapping to real Kick usernames published anywhere. No per-streamer identifiable data.

Intended use

Benchmarking clip-yield expectations for live-stream clipping tools, research on short-form content patterns from live streaming, journalism/analysis citing real production numbers instead of marketing claims.

Citation

ClipMe Research (2026). Kick Clip Yield Dataset: 613 clips from 51
live-stream capture sessions, June–July 2026. CC BY 4.0.
https://clipme.com/research