Independent AI market intelligence
See who AI recommends.
Benchmark Bureau measures how leading AI systems represent, rank, and recommend companies—then publishes the evidence.
Coverage
Cross-platform
Retrieval
Search-aware
Method
Repeat observations
Standard
Open to scrutiny
Why this exists
Recommendation is becoming a new front door.
AI does more than surface links. It narrows choices, frames reputations, and often makes a recommendation. Businesses can measure search rank and social attention. The recommendation layer needs an independent benchmark.
Measure the market
We test real buyer questions across multiple AI systems. The result is a market view, not a private visibility score.
Publish the method
Models, dates, prompts, repetitions, search settings, scoring rules, and limitations belong beside the result.
Preserve disagreement
ChatGPT, Claude, Gemini, and Perplexity do not always agree. That disagreement is intelligence, not noise to hide.
Keep rank independent
Commercial relationships never change a public ranking. Corrections change facts; payments do not.
First research program
Let the evidence choose the market.
Our initial study compares five commercially important markets. We will measure recommendation concentration, model disagreement, prompt sensitivity, citation quality, local specificity, and response stability before selecting the first sustained public benchmark.
Traditional AI-search tools help a company monitor itself. Benchmark Bureau measures the market.
The institutional distinction