Methodology · v2.9

TLSThought Leadership

Thought-leadership output count for {legal_name} in the last 12 months. Count outputs across FOUR categories: (A) peer-reviewed papers on major scholarly indexes / arXiv / IEEE Xplore / ACM Digital Library where the company or a named executive is an author; (B) bylined op-eds / long-form articles in HBR, MIT Sloan Management Review, NYT Opinion, Time, FT, The Economist, WIRED, or equivalent tier-1 outlets — must be authored by a named executive (not a corporate press release); (C) the company's OWN research/engineering publishing track when it carries authorial bylines (typical patterns: first-party AI/engineering research blogs, in-house research-lab publications, vendor research blogs on the company's own domain) — count one ENTRY per published research artifact, NOT each blog page; (D) named keynote slots / annual developer-conference sessions at industry tier-1 events where the company or a named executive is the speaker — count one per SESSION, drop panel appearances without a verifiable speaker URL. EXTRACTION RULES: SEARCH each category before returning 0; emit a distinct sub-count per category in the reasoning. Drop ghostwritten content marketing, aggregator press releases, recycled marketing copy. Cite each output URL (peer-reviewed paper landing page, byline URL, research-blog post, conference session URL). Return exact_numerical{value: <thought_leadership_count_12mo>, unit: "outputs", source_urls: [≥1 URL per category counted]}. Server applies ratio_band (0→20, 1-5→50, 6-15→75, 16+→90) — DO NOT compute any 0-100 score.