GEO Audit for consensus.app

{"domain": "consensus.app", "overall_score": 78, "business_model": "B2B2C SaaS — AI-powered academic search engine for peer-reviewed literature (freemium: Free/Pro/Team/Enterprise)", "audience_tags": ["To-C Researcher/Student", "To-B University Library/Enterprise", "To-C Clinician/Healthcare"], "tech_stack": ["Next.js", "Cloudflare", "WordPress (marketing/blog)", "MCP server", "REST API", "Semantic Scholar/OpenAlex/PubMed"], "competitor_domains": ["elicit.com", "scite.ai", "semanticscholar.org"], "scorecard": [{"metric": "Technical Accessibility", "score": 11, "max": 15, "status": "PASSED"}, {"metric": "Structured Schema Quality", "score": 8, "max": 10, "status": "PASSED"}, {"metric": "Content Extractability", "score": 10, "max": 15, "status": "WARNING"}, {"metric": "Answer Density (BLUF)", "score": 8, "max": 10, "status": "PASSED"}, {"metric": "Citations & Empirical Data", "score": 13, "max": 15, "status": "PASSED"}, {"metric": "Expertise & Attribution (EEAT)", "score": 7, "max": 10, "status": "PASSED"}, {"metric": "Off-Page Entity Footprint", "score": 12, "max": 15, "status": "PASSED"}, {"metric": "Machine-Readable Discovery", "score": 9, "max": 10, "status": "PASSED"}], "prompt_fan_out": ["AI-powered search engine for scientific research papers with citations and consensus", "Best tools to find the scientific consensus on medical and health questions", "Consensus.app vs Elicit vs Scite: AI academic search engine comparison", "How to search peer-reviewed literature with AI summaries instead of reading papers", "Consensus Meter: how to see if research supports or refutes a yes/no question"], "recommendations": ["Fix robots.txt: remove 'Disallow: /blog/' or add a separate AI-crawler rule allowing the blog — currently the entire marketing blog is blocked from OAI-SearchBot, ChatGPT-User, and PerplexityBot despite the explicit Allow: / for those agents", "Add WebSite schema with SearchAction and an FAQPage schema (what is Consensus, how does the Consensus Meter work) to the homepage @graph — LLMs heavily extract FAQPage for 'what is X' answers", "Fix canonical URL on homepage: <link rel=canonical> currently points to https://consensus.app/search/ instead of https://consensus.app/ — this dilutes the entity URL for the homepage", "Serve agents.md and .well-known/mcp/server-card.json as plain text/JSON without the Cloudflare managed challenge — anonymous AI agents that are not on the OAI/Perplexity allow-list receive the HTML challenge shell instead of the machine-readable files", "Publish author bylines and team credentials on blog/medical content to strengthen EEAT — the 50,000 clinical-guidelines and medical claims would benefit from named expert attribution", "Add a Wikipedia page and public GitHub org for consensus.app — currently no Wikipedia entry and no first-party open-source presence, which weakens off-page entity footprint vs competitors", "Standardize paper count across surfaces (llms.txt says 200M+, schema featureList 220M+, homepage copy 250M+) for consistency in machine-readable claims"], "agent_repair_prompt": "You are an AI coding agent tasked with improving consensus.app's Generative Engine Optimization (GEO) readiness. This is a Next.js application (App Router, data-next-head meta tags) fronted by Cloudflare, with a WordPress marketing/blog surface. Implement the following changes:\n\n1. **Fix robots.txt** (served at https://consensus.app/robots.txt): currently has `User-agent: *` with `Disallow: /blog/` followed by `User-agent: OAI-SearchBot`, `User-agent: ChatGPT-User`, `User-agent: PerplexityBot`, `User-agent: Perplexity-User` each with `Allow: /`. The blog disallow still applies to the AI bots via the wildcard group, blocking all blog content from AI crawlers. Add an explicit `Allow: /blog/` line to each AI bot user-agent group (order matters in robots.txt — later groups are more specific), and add a `User-agent: GPTBot` + `User-agent: ClaudeBot` group with `Allow: /` to cover other major AI crawlers.\n\n2. **Fix homepage canonical**: in the homepage layout metadata, change the canonical from `https://consensus.app/search/` to `https://consensus.app/` so the entity URL resolves to the root.\n\n3. **Add WebSite schema with SearchAction** to the existing JSON-LD @graph in the homepage (currently WebApplication + Organization): add `{\"@type\": \"WebSite\", \"@id\": \"https://consensus.app/#website\", \"url\": \"https://consensus.app/\", \"name\": \"Consensus\", \"publisher\": {\"@id\": \"https://consensus.app/#organization\"}, \"potentialAction\": {\"@type\": \"SearchAction\", \"target\": \"https://consensus.app/results/?q={search_term_string}\", \"query-input\": \"required name=search_term_string\"}}`.\n\n4. **Add FAQPage schema** to the homepage @graph with 3-4 questions LLMs answer from the site: 'What is Consensus?', 'How does the Consensus Meter work?', 'Is Consensus free to use?', 'How many papers does Consensus index?' — each with concise answers from the existing homepage/llms.txt copy.\n\n5. **Serve machine-readable files without bot challenge**: ensure https://consensus.app/agents.md and https://consensus.app/.well-known/mcp/server-card.json are served as static plain-text/JSON (bypass Cloudflare managed challenge for these paths via WAF rule / cache rule, or serve from the CDN edge with `cf-cache-status: HIT` for anonymous GET).\n\n6. **Add author bylines to blog/medical content**: add visible author name + credentials + date to all blog posts and clinical-content pages.\n\n7. **Standardize paper-count claims**: pick one number (e.g. '220M+ papers' from the schema featureList) and use it consistently across homepage copy, llms.txt, llms-full.txt, and JSON-LD.\n\nKeep changes within the Next.js App Router structure (app/layout.tsx, app/page.tsx, public/robots.txt) and WordPress theme for the blog. Validate JSON-LD with Google Rich Results Test after deployment."}