{"generated_at":"auto","methodology":{"title":"Structural Entropy of the Web and AI Visibility","tokenizer_method":"Word-count × 1.3 (approximates tiktoken cl100k_base)","semantic_regions":"TITLE, META, H1, CONTENT (main/article/body), JSONLD","structural_regions":"NAV, FOOTER, SCRIPTS, STYLES, ATTRS, HEADER, ASIDE","noise_formula":"NoiseRatio = StructuralTokens / TotalTokens","entropy_formula":"H = -Σ pᵢ log₂(pᵢ); NormalizedEntropy = H / log₂(k)","score_formula":"EntropyScore = 100 × (1 - α×(1-S/T) - β×(1-NormalizedEntropy))","alpha_weight":0.6,"beta_weight":0.4,"thresholds":{"Fair (50-79)":"Moderate bloat; some improvement needed for optimal AI visibility","Good (80-100)":"Low token bloat; AI agents can extract content efficiently","NoiseRatio_High":"\u003e= 0.7 (High bloat)","NoiseRatio_Low":"\u003c 0.4 (Low bloat)","NoiseRatio_Med":"0.4 - 0.69 (Medium bloat)","Poor (0-49)":"High token bloat; AI agents likely miss key content"}},"actionable_checklist":["1. Move \u003cmain\u003e content to the top of the DOM, before navigation and sidebars.","2. Add comprehensive JSON-LD (Organization, Product/Article, BreadcrumbList, FAQ).","3. Defer or externalize non-critical JavaScript — inline scripts inflate tokens by 20-40%.","4. Simplify navigation: mega-menus can consume 4,000+ tokens of redundant links.","5. Extract large CSS into external stylesheets; only inline critical above-fold CSS.","6. Remove duplicate \u003cnav\u003e blocks (mobile + desktop duplication is common).","7. Minimize footer link farms — link to a sitemap page instead.","8. Use semantic HTML (\u003carticle\u003e, \u003csection\u003e, \u003caside\u003e) instead of \u003cdiv\u003e soup.","9. Server-render critical content for AI bots (SSR or pre-rendered snapshots).","10. Monitor Entropy Score in CI/CD to prevent regressions on every deploy."]}