How the AI Brand Index Works

The AI Brand Index measures brand visibility in AI answers. Every week it asks ChatGPT, Gemini, Google AI Mode, and Perplexity the same category question, then combines their top-10 brand lists into one ranking.

The index exists to answer a simple question: when a buyer asks an AI system which brands to consider in a category, who shows up, and how consistently. It identifies the leaders in each category and tracks how that visibility shifts over time.

How the methodology works

1. 1

   Ask the same category question

   Every Sunday we query ChatGPT, Gemini, Google AI Mode, and Perplexity. For each category we ask the same question: which ten brands are the most visible, significant, and recommended in that niche. Each system returns its own top ten.

2. 2

   Collect rankings and cited sources

   We record each brand's inclusion, position, and the sources each system cites in its answer at collection time. Brand rank and cited sources are stored separately, because sources help explain visibility but do not by themselves set the ranking.

3. 3

   Normalize brand names

   AI answers are not always tidy. We normalize obvious name variants to a single brand, keep parent and sub-brands separate when they are marketed separately, and set aside entities that are not real products in the category. Ambiguous cases get a manual review.

4. 4

   Combine presence and position

   A brand's position reflects how often it appears across the systems and how high each system places it. Consistent, high placements rank above occasional, low ones. Non-appearance counts as no support from that system, and we do not publish a numeric per-brand score.

5. 5

   Publish the weekly snapshot

   A fresh snapshot is collected every Sunday, so the index updates weekly and you can compare snapshots over time.

How to read the ranking

We read a brand's inclusion and rank as AI recommendation visibility, not as proof of product quality. Week-to-week movement can be noisy, so sustained movement across several snapshots is more meaningful than a single jump. Because each system uses different retrieval and ranking logic, treat cross-system comparisons as observed outputs under one standardized prompt, not identical experiments.

What the index does not measure

The index measures AI recommendation visibility, not product quality. A brand can rank highly because it has strong third-party coverage or long category history, not because it is the best fit for a given buyer. Rankings are based on observed position, not a published score, so treat them as a visibility signal and run your own fit check before choosing a tool. See the full AI Brand Index to explore categories.