Raw data (Level 1)

The underlying dataset behind every claim in this study. PII-safe: author handles removed.

What's inside

corpus_piisafe.jsonl — all 6,024 records (Reddit, YouTube, TikTok, Instagram), one JSON per line, author handles stripped.
_profile.md · _persona-roster.md — the frozen denominators + personas.
_coverage.md · _corroboration.md — the coverage matrix + the two-modality corroboration check.
lens-*.md — the five analytical lenses (sentiment, competitive, personas, triggers, angles), each cited to record ids.
_scores.json — per-brand organic scorecard numbers.

Provenance. Captured 2026-07-17 via Digital Position's VOC methodology. Public content only. On-site reviews and Amazon are documented gaps (spirits aren't sold with public review widgets or via Amazon). Total scrape cost: $1.14.