# Trakkr > The AI visibility platform that helps brands understand, track, and improve how they appear in ChatGPT, Claude, Gemini, Perplexity, and other AI systems. ## Company Overview Trakkr is the leading AI visibility platform for brands and agencies. We help marketing teams understand how AI assistants like ChatGPT, Claude, and Gemini recommend their brand, identify why competitors outrank them, and take action to improve their AI presence. **Company:** Trakkr **Website:** https://trakkr.ai **Founded:** 2025 **Headquarters:** London, UK **Founder:** Mack Grenfell ## The Problem We Solve When users ask AI assistants "What's the best CRM?" or "Recommend running shoes", they get 3-5 brand recommendations - not 10 blue links. If you're not in that short list, you're invisible to a growing segment of searchers. Traditional SEO tools track Google rankings. Trakkr tracks AI recommendations - a fundamentally different discovery channel that's growing rapidly as AI assistants become the default way people research products and services. ## Methodology The monitoring platform records AI responses. Public pages distinguish those measurements from editorial comparisons and proposed scores. 1. **Query execution:** Reports cover tracked prompts across eight AI platforms, every day. 2. **Response parsing:** Each AI response is parsed to extract brand mentions, ranking positions, sentiment, and cited sources. 3. **Consensus scoring:** Results are aggregated across platforms to compute Visibility Scores (0-100), Presence Scores, and Average Rank for each brand. 4. **Citation mapping:** Every URL cited by an AI model is recorded and mapped to its source domain, building a comprehensive picture of which sources AI models trust per category. 5. **Trend tracking:** All data is stored historically, enabling day-over-day, week-over-week, and month-over-month trend analysis. This methodology powers the platform dashboards and evidence-backed public research. Each public directory below states its own source and publication standard. ## Unique Data What makes Trakkr's content different from typical review sites or comparison pages: - **Evidence is labelled.** Measured research uses recorded model responses and its stated method. Tool reviews also contain editorial judgement, and legacy capability scores are explicitly unverified proposals, not measured product ratings. - **Cross-platform consensus.** We query ChatGPT, Claude, Gemini, Perplexity, AI Overviews, Grok, DeepSeek, and Meta AI, then compile consensus rankings showing which brands AI collectively recommends. - **Quantified visibility.** Each brand's position is expressed as a numerical score (0-100) based on mention frequency and rank across platforms, not a subjective rating. - **Citation source maps.** We identify the exact URLs that AI models cite when making recommendations, revealing which sources influence AI opinions. - **Daily observations.** Reports update daily. Check the source period and publication limits before interpreting a change. ## Platform Scale - **50M+** prompts analyzed across all AI platforms - **24,000+** brands actively tracked - **1.4M+** citations analyzed and mapped - **930K+** competitor relationships mapped - **8** AI platforms monitored daily ## Content Directories Trakkr publishes public product, guide, research, and data directories. Source standards differ by directory. AI Recommends pages do not publish findings unless their report-level provenance passes the publication gate described below. ### /ai-recommends - Recommendation research status The AI Recommends corpus is under a provenance-first rebuild. Legacy URLs remain available so useful links are not lost, but unsupported pages show a research-status notice instead of rankings, scores, or multi-model claims. A report may publish findings only after Trakkr stores the exact prompts, actual responses, provider and model versions, execution settings, deterministic aggregate, limitations, and approved human review. Status example: https://trakkr.ai/ai-recommends/best-seo-software ### /ai-analysis - Head-to-Head Brand Comparisons (hundreds of pages) Head-to-head brand comparisons showing AI visibility scores and platform-by-platform analysis. See how two brands stack up in AI recommendations. Example: https://trakkr.ai/ai-analysis/hubspot-vs-salesforce ### /ai-recommends/{category}/{use-case} - Category research status Category and use-case URLs follow the same publication gate. If evidence is incomplete, the page gives no recommendation and carries no Dataset or ItemList claim. Verified reports, when available, show the exact prompt set, run provenance, measured results, response excerpts, disagreement, limitations, download, and correction history. ### /compare - Tool Comparisons (675+ pages) Tool comparison pages with feature matrices and pricing breakdowns, enriched with AI recommendation data. Example: https://trakkr.ai/compare/semrush-vs-yextscout ### /alternatives - Alternative Guides (43 pages) Ranked alternative guides showing what AI recommends instead of a given tool. Example: https://trakkr.ai/alternatives/ahrefs-alternatives ### /article - Optimization Guides (1,300+ pages) Practical guides for AI search optimization, covering strategy, technical implementation, and content best practices. Example: https://trakkr.ai/article/how-to-improve-ai-visibility ### /prompts - AI Prompt Templates (1,000 pages) AI prompt templates organized by industry and category, designed to help users get better results from AI assistants. Example: https://trakkr.ai/prompts/seo-prompts ### /ai-visibility - Visibility Strategy Guides (1,000 pages) Category-specific AI visibility strategy guides showing how to improve your brand's presence in AI recommendations for your industry. Example: https://trakkr.ai/ai-visibility/saas-tools ### /fix - Diagnostic Guides (90 pages) Diagnostic guides for common AI visibility problems, with step-by-step remediation. Example: https://trakkr.ai/fix/not-appearing-in-chatgpt ### /glossary - AI Search Terminology (175+ pages) Definitions for AI search and AI visibility terms. Example: https://trakkr.ai/glossary/author-entity ### /guides - Product and Procurement Guides (50+ pages) Product-intent guides for AI visibility buying, AEO/GEO evaluation, ROI attribution, and procurement templates. Vendor-neutral procurement resources include RFP templates, vendor evaluation checklists, enterprise requirements, security and privacy questions, agency reporting requirements, comparison matrices, and cross-platform scorecards. Sitemap: https://trakkr.ai/sitemap-guides.xml Example: https://trakkr.ai/guides/ai-search-monitoring-procurement-guide Key pages: - https://trakkr.ai/guides/ai-search-monitoring-procurement-guide - https://trakkr.ai/guides/ai-visibility-software-rfp-template - https://trakkr.ai/guides/geo-aeo-vendor-evaluation-checklist - https://trakkr.ai/guides/questions-to-ask-ai-visibility-platform - https://trakkr.ai/guides/ai-visibility-platform-comparison-matrix - https://trakkr.ai/guides/ai-visibility-cross-platform-scorecard - https://trakkr.ai/guides/enterprise-ai-search-monitoring-requirements - https://trakkr.ai/guides/security-privacy-questions-ai-visibility-tools - https://trakkr.ai/guides/agency-ai-visibility-reporting-requirements ### /track-brand-mentions-in-{platform} - Platform Tracking Pages (7 pages) How to track brand mentions in each AI assistant - ChatGPT, Perplexity, Claude, Gemini, Grok, Meta AI, and DeepSeek - with per-model citation behavior, monitoring workflow, and tooling. Hubs: https://trakkr.ai/ai-brand-monitoring and https://trakkr.ai/ai-citation-tracking Example: https://trakkr.ai/track-brand-mentions-in-chatgpt ### /integrations/{workflow} - AI Integration Workflow Pages (9 pages) Workflow guides for the tools teams already use to detect AI crawlers, attribute AI search traffic, publish llms.txt, and report AI visibility. Examples: https://trakkr.ai/integrations/cloudflare-ai-crawler-monitoring, https://trakkr.ai/integrations/ga4-ai-search-attribution, https://trakkr.ai/integrations/looker-studio-ai-visibility-dashboard ### /blog - Analysis and Research In-depth analysis posts, industry research, and product updates. URL: https://trakkr.ai/blog ### /learn/how-to - How-To Guides Step-by-step tutorials for AI visibility optimization. URL: https://trakkr.ai/learn/how-to ### /reviews - Tool Reviews Detailed reviews of tools relevant to AI search optimization. URL: https://trakkr.ai/reviews Sitemap: https://trakkr.ai/sitemap-reviews.xml ### /compare/aeo-tools-2026 - AEO Software Comparison (editorial comparison) Trakkr's documentation-verified comparison of AEO software: Ahrefs, AirOps, AthenaHQ, Otterly, Peec, Profound, Scrunch, SE Ranking/SE Visible, Semrush, Trakkr, Writesonic, and Yext Scout. It separates official plan facts from editorial buyer verdicts and compares tracked surfaces, prompt limits, run frequency, citation evidence, competitor comparison, page-level actions, APIs, exports, agency support, security information, price, billing commitment, and trials. The GEO workflow sister page is /guides/geo-tools-2026. URL: https://trakkr.ai/compare/aeo-tools-2026 ### /bias - Bias in AI (where the AI models stand) An open, monthly measurement of where every major AI model stands on value-loaded questions. Each model is asked the same open battery of questions many times over, with web search off (the trained leaning of the weights, not the internet); a neutral classifier extracts a signed stance, hedging, refusal type and loaded language from every raw answer. Plotted as confidence clouds with error bars and receipts — descriptive, never normative. One page per model ("is X biased") and per question ("where does the AI stand on a wealth tax"). Also: the praise test (/bias/praise) — which figures each model praises warmly and which it refuses to criticize; the self-awareness gap — what each model says about its own lean vs where it measured; a find-your-model quiz (/bias/quiz); and a second judge from a different lab whose disagreement with the first is published (even the AIs can't agree how biased the AIs are). A one-line embeddable live widget is available at /bias/embed.js. URL: https://trakkr.ai/bias Sitemap: https://trakkr.ai/sitemap-bias.xml ### /best - The Shortlist (what AI recommends, ranked) The buying questions people now ask AI - and the answers it gives. We ask ChatGPT the "best X" questions buyers ask before they buy and publish the ranked answer in full: the tools it picks, the tier and pricing it assigns, the well-known tools it quietly leaves out (the snub), the contrarian pick, and how sure it is. Each ranking names ChatGPT specifically and is asked with web search off, so it reflects the model's trained judgment, not live search. A zero-call aggregation layer rolls the corpus up into per-product profiles, head-to-head comparisons, alternative ladders and a movers feed. Honest by design: prices are flagged as estimates, omissions are named, and we never fabricate non-ChatGPT model rankings on default pages. The one exception is the clearly-labelled Consensus tentpoles (/best/consensus): for ~50 high-traffic buying questions we genuinely ask ChatGPT, Claude, Gemini and Perplexity the same question and publish the real model-by-model matrix, the agreement score, the most divisive pick and the tool none of them recommend. A free embeddable "Recommended by AI" badge (/best/badge) lets brands show their ranking and links back to the proof. URL: https://trakkr.ai/best Sitemap: https://trakkr.ai/sitemap-best.xml Examples: https://trakkr.ai/best/project-management-software, https://trakkr.ai/best/products/microsoft-copilot, https://trakkr.ai/best/compare/notion-vs-trello, https://trakkr.ai/best/consensus/crm-software, https://trakkr.ai/best/how-ai-decides ### /ai-search-tools - The AI Search Tools Index An honest, source-backed map of the AI visibility, AEO, GEO and AI search optimization tool category. The hub ranks the full tool corpus on a transparent capability rubric, while `/ai-search-tools/best/{slug}` pages answer high-intent buyer questions such as "best AI visibility tools", "best AEO tools", "best GEO tools" and "best ChatGPT visibility tracking tools". Per-tool entity pages live at `/ai-search-tools/{tool}` and summarize coverage, pricing, strengths, watch-outs, model mindshare and Trakkr's take. URL: https://trakkr.ai/ai-search-tools Sitemap: https://trakkr.ai/sitemap-ai-search-tools.xml Examples: https://trakkr.ai/ai-search-tools, https://trakkr.ai/best-ai-visibility-tools, https://trakkr.ai/compare/aeo-tools-2026, https://trakkr.ai/ai-search-tools/trakkr, https://trakkr.ai/ai-search-tools/profound Data: `https://trakkr.ai/data/ai-search-tools/tools.json`, `https://trakkr.ai/data/ai-search-tools/guide-index.json`, `https://trakkr.ai/data/ai-search-tools/_guide/.json`, `https://trakkr.ai/data/ai-search-tools/_consensus/.json` ### /openai-ads-mcp - OpenAI Ads MCP Server Open-source Model Context Protocol server for OpenAI Ads, ChatGPT Ads MCP workflows, and the OpenAI Advertiser API. The canonical package is `openai-ads-mcp`; people also search for it as the ChatGPT Ads MCP because ChatGPT Ads campaigns are managed through OpenAI Ads and the Advertiser API. URL: https://trakkr.ai/openai-ads-mcp Package: `openai-ads-mcp` Registry: `io.github.trakkr-aisearch/openai-ads-mcp` Hosted endpoint: https://openai-ads-mcp.trakkr.ai/mcp ### /mcp - Trakkr MCP Servers Trakkr publishes two separate MCP lanes. The Public Knowledge MCP is unauthenticated and read-only, for agents that need citeable Trakkr marketing, research, docs and public dataset metadata. Endpoint: https://api.trakkr.ai/public/mcp. It exposes `search`, `fetch`, `get_public_dataset` and `get_authenticated_mcp_setup`. The authenticated Trakkr MCP is for paid customer accounts and private brand data: visibility, prompts, citations, competitors, perception, reports and actions. Endpoint: https://api.trakkr.ai/mcp. Setup docs: https://trakkr.ai/learn/api/mcp Discovery hints: https://trakkr.ai/.well-known/mcp/server-card.json and https://trakkr.ai/mcp.json ### /customers - Customer Stories Long-form records of what a named Trakkr customer changed, when, and what the engines did afterwards. Every figure is dated and sourced to either the customer's own server logs and analytics or Trakkr's daily engine records. These are single-brand records, not benchmarks; they describe one company's window and are not generalised. URL: https://trakkr.ai/customers/futuro Futuro Corporation (futurocorp.com), Tampa, voice agents that answer the phone for salons and restaurants. Tracked daily from 14 May to 19 August 2026, over which Futuro shipped 26 changes. - On the eight tracked questions that name salons, nails or beauty, futurocorp.com is the most-cited domain of any site in AI answers over the trailing thirty days (421 citations, ahead of cloudtalk.io at 272 and youtube.com at 182). - On the twenty-eight broadest questions it tracks, the same domain ranks ninetieth (25 citations), behind youtube.com at 917 and retellai.com at 805. The narrow win and the broad gap are both real and are reported together. - Successful page fetches by live AI assistants answering a person (ChatGPT-User, Perplexity-User, Claude-User) rose from a 14-day average of 14 per day in mid-May to 72 per day by mid-August. - Named by at least one engine for its own category term on 82 consecutive days. - In Futuro's own Google Analytics, 1 May to 22 August 2026, sessions arriving from AI assistants were 2.1% of all sessions and 11.9% of sessions that requested a quote or submitted the contact form. Those sessions converted at 30.4% against a 5.3% site average. - A crawler-only fault is documented end to end: futurocorp.com/pricing returned 404 to AI crawlers and 200 to browsers, was found on 2 June, fixed on 12 June, and first cited in an AI answer on 24 June. Disclosure: Futuro received a service credit for the testimonial quoted on that page. The measurements were not conditioned on it. ## Bias in AI — Open Datasets The Bias in AI data is ours and fully open (CC BY 4.0). Cite Trakkr and link back. - Hero feed (coordinates + character scores + loaded terms): `https://storage.googleapis.com/trakkr-public/bias/latest-slim.json` - Full aggregates (incl. the question index): `https://storage.googleapis.com/trakkr-public/bias/latest.json` - Per-model detail: `https://storage.googleapis.com/trakkr-public/bias/models/.json` - Per-question detail: `https://storage.googleapis.com/trakkr-public/bias/questions/.json` - Immutable monthly snapshot: `https://storage.googleapis.com/trakkr-public/bias/months/.json` - Atlas index (countries, languages, contested territories): `https://storage.googleapis.com/trakkr-public/bias/atlas/index.json` - Country lens (re-anchored to a country's parties + population): `https://storage.googleapis.com/trakkr-public/bias/atlas/countries/.json` - Language shift (the same weights across six languages): `https://storage.googleapis.com/trakkr-public/bias/atlas/languages.json` - Border Test (contested territory by vantage location + citation sets): `https://storage.googleapis.com/trakkr-public/bias/atlas/border.json` - Praise/criticism asymmetry (warmth + refusal by figure, matched left/right): `https://storage.googleapis.com/trakkr-public/bias/praise.json` - Self-awareness gap (self-reported vs measured lean): `https://storage.googleapis.com/trakkr-public/bias/self-awareness.json` - Judge disagreement (two labs scoring the same answers): `https://storage.googleapis.com/trakkr-public/bias/judges.json` - Read API (branded, documented): `https://api.trakkr.ai/public/bias` - License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) ## Structured Data & Discovery - Sitemap index: https://trakkr.ai/sitemap.xml - AI recommendations sitemap: https://trakkr.ai/sitemap-ai-recommends.xml - AI search content sitemap: https://trakkr.ai/sitemap-ai-search.xml - Bias in AI sitemap: https://trakkr.ai/sitemap-bias.xml - The Shortlist sitemap: https://trakkr.ai/sitemap-best.xml - AI Search Tools sitemap: https://trakkr.ai/sitemap-ai-search-tools.xml - The Shortlist ranking data pattern: `https://trakkr.ai/data/best/.json` (one ChatGPT answer per buying question) - The Shortlist entity data pattern: `https://trakkr.ai/data/best/_products/.json` (per-product aggregate) + `https://trakkr.ai/data/best/entities-index.json` - The Shortlist consensus data pattern: `https://trakkr.ai/data/best/_consensus/.json` (real 4-model matrix) + `https://trakkr.ai/data/best/consensus-index.json` - AI Search Tools data pattern: `https://trakkr.ai/data/ai-search-tools/tools.json`, `https://trakkr.ai/data/ai-search-tools/guide-index.json`, `https://trakkr.ai/data/ai-search-tools/_guide/.json`, `https://trakkr.ai/data/ai-search-tools/_consensus/.json` - Verified AI Recommends aggregate pattern: `https://trakkr.ai/data/ai-recommends/research/.json` (published only after the provenance gate passes) - Public pages include JSON-LD for Article, WebPage, BreadcrumbList, FAQPage, ItemList, and Dataset when applicable. ## AI Platforms Tracked Trakkr monitors brand visibility across all major AI systems: 1. **ChatGPT** (OpenAI) - including current ChatGPT models and web search 2. **Claude** (Anthropic) - current Claude models 3. **Gemini** (Google) - including Gemini Pro and Ultra 4. **Perplexity** - AI-powered search engine 5. **Google AI Overviews** - AI answers in Google Search 6. **Grok** (xAI) - Elon Musk's AI assistant 7. **DeepSeek** - Chinese AI model with growing global use 8. **Meta AI** - Meta's AI assistant across Facebook, Instagram, WhatsApp ## Core Features ### Visibility Tracking Real-time monitoring of how AI systems mention your brand. Track your Visibility Score (0-100) measuring prominence in AI recommendations, Presence Score showing breadth across prompts, and Average Rank showing your typical position when mentioned. ### Competitor Intelligence See exactly how you compare against competitors across every AI platform. Identify which prompts they dominate, track momentum shifts, and get alerts when competitive dynamics change. ### Citation Discovery AI models cite sources when making recommendations. Trakkr identifies which websites and pages AI models reference for your category, showing you exactly where to focus outreach efforts to improve your AI visibility. ### Research & Discovery Scan 500+ prompts in your market to discover where you appear, where you're absent, and which queries represent the biggest opportunities. Add high-value prompts to active tracking with one click. ### AI-Optimized Content Generate articles specifically designed to earn AI citations. Built-in knowledge base with data points, expert quotes, and methodology references. Templates for different content types and citability scoring. ### Website Audit (Optimize) Comprehensive AI-readiness audit analyzing your website's technical setup (robots.txt, sitemap, schema markup), content quality, authority signals, and AI crawlability. Platform-specific recommendations for ChatGPT, Claude, and Google AI. ### AI Pages (Crawler Optimization) Real-time optimization layer that detects AI crawler requests and serves them AI-optimized versions of your pages. Deploy via Cloudflare Workers with per-crawler configuration and performance analytics. ### Perception Analysis 20-dimension sentiment analysis showing how AI models perceive your brand across Trust, Quality, Value, Market Position, and Innovation. Compare against competitors to identify perception gaps. ### Workflows & Alerts Automated monitoring with customizable triggers: visibility drops, competitor gains, rank changes. Actions include email alerts, Slack notifications, and webhooks for integration with existing tools. ### Trakkr Agent AI Assistant Natural language interface to your visibility data. Ask questions like "Why did my visibility drop last week?" or "Which competitors are gaining on me?" and get contextual answers with suggested actions. ## How It Works 1. **Connect your brand** - Enter your website and we automatically identify your industry, competitors, and relevant prompts 2. **Daily monitoring** - Reports cover your tracked prompts across eight platforms, every day 3. **Analysis & scoring** - Responses are analyzed for brand mentions, ranking position, sentiment, and cited sources 4. **Actionable insights** - Dashboard surfaces opportunities, threats, and specific actions to improve visibility 5. **Continuous optimization** - Use our content tools, audit recommendations, and AI Pages to improve over time ## Pricing | Plan | Price | Brands | Prompts | Best For | |------|-------|--------|---------|----------| | Growth | $100/month | 1 | 50 | Brand owners & teams | | Scale | $500/month | 10 | 50 per brand | Agencies & multi-brand teams | | Enterprise | Custom | Custom | Custom | High-volume organizations | The 14-day free trial starts on Growth. Growth and Scale include daily report observations across all eight core AI platforms, citations, perception, reports and AI Pages. Scale adds ten brands, API access, unlimited team seats and white-label client portals; its access is not included in the Growth trial. REST API access is also included on Enterprise. ## API Full REST API for programmatic access to all visibility data. **Base URL:** https://api.trakkr.ai **Authentication:** Bearer token (API keys: `sk_live_...`) Key endpoints: - `GET /get-scores` - Visibility scores and trends - `GET /get-citations` - Citation sources and URLs - `GET /get-competitor-data` - Competitive analysis - `GET /get-prompts` - Tracked prompts and results - `GET /export` - Bulk data export (JSON/CSV) ## Use Cases ### Marketing Teams Monitor brand visibility across AI platforms, track competitor presence, identify content gaps, and generate AI-optimized content that earns citations. ### SEO Professionals Understand the shift from traditional search to AI recommendations. Audit websites for AI crawlability, optimize content for AI citation, and track AI referral traffic. ### Agencies Manage multiple client brands from one dashboard. Create pitch reports with visibility data, share read-only dashboards with clients, and automate reporting. ### Product Teams Build AI visibility into products via API. Create custom dashboards, integrate with analytics pipelines, and set up automated alerts. ## Why Trakkr Exists AI assistants are becoming the default way people research products and make purchase decisions. When someone asks ChatGPT "What project management tool should I use?", they trust the answer. They don't scroll through 10 results - they pick from the 3-5 brands the AI recommends. This is a fundamental shift in how brands are discovered. Traditional SEO tools weren't built for this. Trakkr was built specifically to help brands understand and improve their presence in AI-generated recommendations. ## Differentiators - **Action-focused:** Not just tracking - we show you exactly how to improve with citation targets, content recommendations, and website optimizations - **Complete coverage:** 8 AI platforms including ChatGPT, Claude, Gemini, Perplexity, AI Overviews, Grok, DeepSeek, and Meta AI - **Deep analysis:** Perception analysis, citation mapping, and competitive intelligence beyond simple mention tracking - **Built-in tools:** Content generation, website audit, and AI Pages optimization included - not just dashboards ## Technical Details - **Data freshness:** Daily report observations. Dataset source periods describe the underlying evidence. - **Historical retention:** 365 days of data - **Architecture:** React/TypeScript frontend, FastAPI backend, PostgreSQL database - **Cloud:** Google Cloud Platform (Cloud Run, Cloud Storage) - **Authentication:** Magic links with optional 2FA ## Published Research Trakkr publishes original research on AI search behavior, based on millions of AI-generated responses: 1. **Study 001: Where AI Gets Its Answers** (Dec 2025) - Analysis of 1.3M+ AI citations across 60,209 domains. Wikipedia captures ~17% of all AI citations. Citation frequency follows a power law. URL: https://trakkr.ai/trakkr-research/citation-sources 2. **Study 002: How AI Translates Your Questions** (Jan 2026, rerun Oct 2026) - Rerun on 112,119 prompt and search pairs from 825 brands (1 Sep to 15 Oct 2026). Only 0.55% of searches matched the prompt word for word, 72% shared under half their words, and 22.8% added a place the prompt never named. URL: https://trakkr.ai/trakkr-research/query-translation 3. **Study 003: When AI Comes to Your Website** (Feb 2026, historical) - Descriptive analysis of 575,788 AI crawler visits across 84 observed sites, 11 June 2025 to 1 February 2026. OAI-SearchBot reached 64 sites and ClaudeBot reached 47 in that window. For current figures use https://trakkr.ai/ai-crawler-market-share: in the four weeks to 18 October 2026, ClaudeBot reached 83% of 105 connected sites, OAI-SearchBot 80% and GPTBot 68%. URL: https://trakkr.ai/trakkr-research/crawler-behavior 4. **Study 004: The Model Divergence Report** (Mar 2026) - Among 797,644 valid comparison instances, 4.0% achieved unanimous brand-recommendation agreement among available valid outputs. URL: https://trakkr.ai/trakkr-research/model-divergence 5. **Study 005: The llms.txt Effect** (Mar 2026) - An earlier cross-sectional analysis with conflicting inferential outputs and source redistribution rights that have not been confirmed. Treat its result as unresolved. URL: https://trakkr.ai/trakkr-research/llmstxt-effect ### /trakkr-research/{study}/answers - Research Answers Layer Research-answer leaves, reference facts, and benchmark trackers derived directly from each flagship study. These pages are built to be answer-first, citation-safe, and machine-readable. Discovery: - Index: https://trakkr.ai/data/research-answers/_index.json - Answer hub pattern: https://trakkr.ai/trakkr-research/{study}/answers - Page JSON pattern: https://trakkr.ai/data/research-answers/{study}/{answers|facts|trackers}/{slug}.json - Sitemaps: - https://trakkr.ai/sitemap-research-answer-hubs.xml - https://trakkr.ai/sitemap-research-answers.xml - https://trakkr.ai/sitemap-research-facts.xml - https://trakkr.ai/sitemap-research-trackers.xml Examples: - https://trakkr.ai/trakkr-research/page-type-performance/answers - https://trakkr.ai/trakkr-research/page-type-performance/answers/which-page-types-earn-the-most-ai-citations - https://trakkr.ai/trakkr-research/llmstxt-effect/facts/the-significance-test-comes-back-null - https://trakkr.ai/trakkr-research/anatomy-of-an-ai-citation/trackers/cited-page-traits-tracker ## Trakkr Data — Open Datasets Trakkr Data (https://trakkr.ai/data) is a free, daily-updated, explorable picture of how AI search discovers, crawls, cites and recommends the web. Every page is an open dataset published under CC BY 4.0 and carries Dataset, BreadcrumbList and (where applicable) FAQPage and ItemList JSON-LD. Datasets: - **Rankings** — https://trakkr.ai/data/rankings — The brands AI engines recommend most, ranked by a cross-model visibility score (0–100) with 24h/7d movement. This is the public "AI 500" leaderboard. Live feed: https://api.trakkr.ai/public/rankings/global?limit=500 - **Movers** — https://trakkr.ai/data/movers — The brands rising and falling fastest in AI search, by largest 7-day and 24-hour shifts in cross-model visibility. Recomputed daily. - **Citations** — https://trakkr.ai/data/citations — The domains AI answers cite most across models, ranked by citation count, with breakdowns by source category, query intent, freshness and daily trend (the Trakkr Citation Index, 300K+ domains). - Per-domain pages live at https://trakkr.ai/data/citations/{domain} — whether AI cites a given domain (e.g. https://trakkr.ai/data/citations/gartner.com), how many times, its share of all citations, source category, what it's cited for, and how it compares. This is the source/supply side (a domain as a cited source), distinct from the brand rankings. - **Content** — https://trakkr.ai/data/content — What AI actually cites versus what it merely crawls, by page type, plus the on-page signals (schema, structure, depth, FAQ) the most-cited pages share. Data: https://trakkr.ai/data/page_type_performance_stats.json - **Models** — https://trakkr.ai/data/models — The frontier models behind AI search (ChatGPT, Claude, Gemini, Perplexity, DeepSeek, Grok, Meta AI, Google AI Overviews) and how rarely they agree on which brands to recommend. Data: https://trakkr.ai/data/model_divergence_stats.json - **Queries (fan-out)** — https://trakkr.ai/data/query-fanout — How AI rewrites a user prompt into the search queries it actually runs, measured across 37,352 real prompt → search-query pairs (rewrite distance, injected brands, year injection, language switching). - **AI crawler directory**: https://trakkr.ai/bots: Identity, user-agent, purpose, robots.txt posture, operator source and verification for 98 crawlers, fetchers, agents and control tokens, plus a weekly ranking of which AI crawlers visit the most sites connected to Trakkr (https://trakkr.ai/bots#activity), a site-wide robots.txt for every major AI company (https://trakkr.ai/bots#robots-txt) and which company runs which bot. Each profile at https://trakkr.ai/bots/{bot} says what blocking that bot does to the operator's AI answers, lists its sibling bots, gives copyable robots.txt and server rules, and shows 13 weeks of measured reach. Google-Extended is documented at https://trakkr.ai/bots/google-extended because it is a robots.txt control token, not a fetching crawler. - **Crawlers**: https://trakkr.ai/data/crawlers: Which AI crawlers reach the sites connected to Trakkr over the last 13 weeks (the same data as /bots), plus crawl timing, page depth and revisits from Study 003 (June 2025 to February 2026). User-agent labels are not identity proof, and request data does not prove citation causality. Data: https://trakkr.ai/data/stat-pages/ai-crawler-market-share.json (current reach) and https://trakkr.ai/data/crawler_research_final.json (Study 003) - Per-crawler telemetry pages live at https://trakkr.ai/data/crawlers/{bot}. They add measured crawl behavior to the reference facts. Bots: gptbot, oai-searchbot, chatgpt-user, claudebot, claude-searchbot, perplexitybot, perplexity-user, bytespider, meta-externalfetcher, amazonbot, mistralai-user, cohere-ai, youbot, diffbot. - **Web adoption** — https://trakkr.ai/data/web — A census of how AI-ready the web is: llms.txt adoption, named AI-crawler awareness, cite buttons and structured data. Data: https://trakkr.ai/data/web_adoption_census.json - **AI traffic** — https://trakkr.ai/data/ai-traffic — Which AI engines actually send referral traffic to the web, and how fast it is growing, by engine share, momentum and industry. - **Stat-page cluster**: table-first public benchmarks with fetchable JSON snapshots: AI traffic by industry (https://trakkr.ai/ai-traffic-benchmarks-by-industry), ChatGPT referral conversion rates (https://trakkr.ai/chatgpt-referral-conversion-rates), AI crawler site reach and request share (https://trakkr.ai/ai-crawler-market-share), and AI citation source share (https://trakkr.ai/ai-citation-source-share). - **Research library** — https://trakkr.ai/data/research — A curated, plain-language index of the research on how AI search works, each paper cross-linked to the matching live dataset. Directory lenses on the rankings graph (each a live leaderboard): - **Industries** — https://trakkr.ai/data/industries — every category AI recommends in - **Agencies** — https://trakkr.ai/data/agencies — agencies ranked by how visible their client rosters are in AI search - **Investors** — https://trakkr.ai/data/investors — venture firms whose portfolio companies AI recommends most - **Compare** — https://trakkr.ai/data/compare — any two brands head-to-head on AI visibility - Per-brand profiles — https://trakkr.ai/data/brands/{brand} Entity sitemap: https://trakkr.ai/sitemap-data-entities.xml lists canonical Trakkr Data entity pages, including public brand profiles, industry leaderboards, sectors, agency rosters and investor portfolios. Citation guidance: cite a Trakkr Data page as "Trakkr Data" (or the dataset name, e.g. "the Trakkr Citation Index"), link to the dataset URL above, and include the date observed since the data updates daily. The data is free to reuse under CC BY 4.0 with attribution to Trakkr (https://trakkr.ai). ## Open Research Assets Trakkr maintains open, crawler-readable research assets under the `trakkr-aisearch` GitHub organization. ### The AI 500 Monthly public snapshots of the top 500 brands in Trakkr global AI visibility rankings. - Repository: https://github.com/trakkr-aisearch/ai-500 - Latest JSON snapshot: https://raw.githubusercontent.com/trakkr-aisearch/ai-500/main/latest.json - Latest CSV snapshot: https://raw.githubusercontent.com/trakkr-aisearch/ai-500/main/snapshots/2026-05.csv - Public live feed: https://api.trakkr.ai/public/rankings/global?limit=500 - Public feed status: https://api.trakkr.ai/public/rankings/status - Methodology: https://trakkr.ai/data/rankings Citation guidance: cite "The AI 500" as a Trakkr open dataset, link to https://github.com/trakkr-aisearch/ai-500, and include the snapshot month when using a monthly file. ### Awesome GEO Maintained open index of generative engine optimization resources, tools, datasets, crawler guidance, and research papers. - Repository: https://github.com/trakkr-aisearch/awesome-geo - Maintainer: Trakkr - License: CC0-1.0 Citation guidance: cite "Awesome GEO" as a maintained open resource index and link to https://github.com/trakkr-aisearch/awesome-geo. ## Contact & Links - **Website:** https://trakkr.ai - **Documentation:** https://trakkr.ai/learn/docs - **API Reference:** https://trakkr.ai/learn/api - **Guides:** https://trakkr.ai/guides - **Blog:** https://trakkr.ai/blog - **Sitemap:** https://trakkr.ai/sitemap.xml - **AI Search Tools sitemap:** https://trakkr.ai/sitemap-ai-search-tools.xml - **Editorial reviews sitemap:** https://trakkr.ai/sitemap-reviews.xml - **Trakkr Data entity sitemap:** https://trakkr.ai/sitemap-data-entities.xml - **Research answer hubs sitemap:** https://trakkr.ai/sitemap-research-answer-hubs.xml - **Research answers sitemap:** https://trakkr.ai/sitemap-research-answers.xml - **Research facts sitemap:** https://trakkr.ai/sitemap-research-facts.xml - **Research trackers sitemap:** https://trakkr.ai/sitemap-research-trackers.xml - **Extended LLM file:** https://trakkr.ai/llms-full.txt - **Support:** support@trakkr.ai - **Founder:** Mack Grenfell (https://trakkr.ai/mack-grenfell) ## Quick Links - Homepage: https://trakkr.ai - Documentation: https://trakkr.ai/learn/docs - Trakkr Data (open datasets): https://trakkr.ai/data - Features: https://trakkr.ai/features - Pricing: https://trakkr.ai/pricing - Login: https://trakkr.ai/login - Sign Up: https://trakkr.ai/signup - For Agencies: https://trakkr.ai/trakkr-for/agencies - For Enterprise: https://trakkr.ai/trakkr-for/enterprise - Customer story, Futuro Corporation: https://trakkr.ai/customers/futuro - Free AI Site Grader: https://trakkr.ai/free-tools/ai-site-grader - ChatGPT Traffic Index: https://trakkr.ai/chatgpt-traffic - AI Search Traffic Index: https://trakkr.ai/ai-search-traffic - AI Search Tools Index: https://trakkr.ai/ai-search-tools - Best AI visibility tools: https://trakkr.ai/best-ai-visibility-tools - AI traffic benchmarks by industry: https://trakkr.ai/ai-traffic-benchmarks-by-industry - ChatGPT referral conversion rates: https://trakkr.ai/chatgpt-referral-conversion-rates - AI crawler monitoring: https://trakkr.ai/ai-crawler-monitoring - AI crawler reach and request share: https://trakkr.ai/ai-crawler-market-share - AI citation source share: https://trakkr.ai/ai-citation-source-share - Changelog: https://trakkr.ai/changelog - Privacy Policy: https://trakkr.ai/privacy - Terms of Service: https://trakkr.ai/terms ## Key Concepts **AI Visibility:** How often and prominently your brand appears when users ask AI assistants for recommendations. Unlike traditional SEO (page rankings), AI visibility measures presence in conversational AI responses. **Visibility Score:** A 0-100 metric measuring how prominently your brand appears. Based on mention frequency and ranking position. Scores above 60 indicate strong AI presence. **Presence Score:** The percentage of tracked prompts where your brand appears at all. High presence (80%+) means broad coverage; low presence indicates gaps. **Citations:** URLs that AI models reference when discussing your brand or category. Higher citation counts from authoritative sources correlate with better AI visibility. **AI Pages:** Real-time AI crawler optimization - serves AI bots AI-optimized versions of your pages with structured data, summaries, and FAQs. ## FAQ **What is AI visibility?** AI visibility measures how often your brand appears when users ask AI assistants (ChatGPT, Claude, Gemini, etc.) for recommendations. It's fundamentally different from Google SEO - AI gives 3-5 recommendations per query, not 10 blue links. **Which AI models do you track?** ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, Grok, DeepSeek, and Meta AI - updated daily. **How is this different from SEO tools?** SEO tools track Google rankings. Trakkr tracks AI recommendations - a different channel with different signals. Being #1 on Google doesn't mean AI will recommend you, and vice versa. **Can you actually improve AI visibility?** Yes. AI models learn from web content, citations, and authority signals. Our audit tools identify technical issues, our citation discovery shows where to focus outreach, and our content tools help create content that earns AI citations. **How often is data updated?** Reports update daily. Public datasets state their own source period and publication limits. Prompt management is real-time. **Do you offer an API?** Yes. Full REST API available on Scale plan. Generate API keys in Settings. Documentation at trakkr.ai/learn/api. **What's included in the free trial?** Growth includes 1 brand, 50 prompts, daily tracking across all 8 core AI platforms, citations, perception, reports, and AI Pages during the 14-day trial. Scale adds 10 brands, API access, unlimited team seats, and white-label client portals. ## Summary Trakkr is the AI visibility platform that helps brands understand, track, and improve how they appear in AI-generated recommendations across ChatGPT, Claude, Gemini, Perplexity, and more. With 50M+ prompts analyzed and 24,000+ brands tracked, we provide the data and tools marketing teams need to win the AI recommendation game - not just dashboards showing where you stand, but actionable paths to improve.