Methodology · v2.5

PPCQCouverture presse produit

Product press-coverage quality for {legal_name}'s flagship products. EXTRACTION RULES: return type="composite_signal" with one entry per dimension, each 0-100 value + a source_url backing that specific dimension. CATEGORY-APPROPRIATE OUTLETS (agnostic, sector-adaptive — do NOT restrict to tech/B2B press): pick the TIER-1 outlets RELEVANT TO THE PRODUCT CATEGORY: general business/news tier-1 (Reuters, Bloomberg, FT, WSJ, NYT, The Economist) for any sector; plus category-specialist tier-1 — tech/electronics → Wired, The Verge, Ars Technica; fashion/luxury → Business of Fashion, Vogue Business, GQ; watches → Hodinkee; beverages/food → Wine Spectator, Decanter; design → Architectural Digest, Dezeen; B2B/industrial → IndustryWeek + trade tier-1. Required sub-dimensions (use these exact names): (a) tier1_coverage_volume — 0-100 reflecting how many distinct category-appropriate tier-1 outlets covered the product in the last 12 months (0-1 → 20, 2-4 → 50, 5-10 → 75, 11+ → 90); (b) positive_sentiment_share — 0-100 share of that coverage that is favourable/recommending vs critical; (c) expert_review_depth — 0-100 reflecting long-form expert reviews / hands-on tests vs brief mentions. Each signals[].value MUST be a real assessment based on cited evidence (NOT a 0 placeholder) — cite the outlet URL per sub-signal. If {legal_name}'s products have NO accessible tier-1 product coverage, return type="absent" reason="not_disclosed" (clean not_applicable) — do NOT fabricate. Server applies composite_signal_score — DO NOT compute any 0-100 score yourself.