Trakkr
The AI 500 How Trakkr structures rankings for AI discovery
The AI 500/Methodology

How Trakkr builds the AI 500

The AI 500 is designed to be crawled, cited, and interpreted as a public reference layer for AI-driven brand discovery.

Assistants tracked: ChatGPT, Claude, Gemini, Perplexity·Latest snapshot Apr 3·Public reference for brand, category, and comparison pages

Why this page exists

Trakkr separates AI visibility from traditional SEO

Search engines rank pages. AI assistants recommend brands inside the answer. Trakkr is built to measure that recommendation layer and make it legible through brand pages, industry rankings, and compare pages.

Industry-specific recommendation prompts

Trakkr queries leading AI assistants with recommendation-style prompts across hundreds of industries. The goal is to capture how assistants naturally surface brands when a user asks for the best option, not how a brand performs on a traditional search engine results page.

Position, frequency, and breadth all matter

The AI visibility score rewards appearing near the top of an answer, showing up consistently across relevant prompts, and appearing across multiple industries where that brand is genuinely recommended. Trakkr uses quadratic scaling with logarithmic compression so outliers do not flatten the rest of the ranking set.

Products roll up to parent brands

Product mentions are mapped back to parent brands using AI-assisted classification with manual overrides. That means brands can receive credit when assistants cite a flagship product name instead of the parent company directly.

Each rankings URL answers a different kind of question

Industry pages answer who AI recommends in a category. Brand pages answer when and why a brand shows up. Compare pages answer which brand currently leads in AI recommendations. Together, they form a crawlable knowledge graph around AI-driven product discovery.