[ Original Research ]

Research the AI-search industry can check.

We sell citation work — which is exactly why every study here discloses its methodology, names its data sources, and flags what we couldn't measure. Read our research the way we'd tell you to read any vendor's: check the method, not the headline.

Published studies

[ Research — Project 02 ]

The Vendor Bias Study: Who Funds GEO Research

30 GEO studies classified by funding source. Vendor-funded research reports effects 2.4x larger than independent work — plus a 5-question vetting checklist.

2.4x

vendor-funded studies report headline effects 2.4x larger than independent research

May 8, 202621 min readRead the study →

[ Research — Project 03 ]

The Surface Diet: What Each AI Engine Actually Cites

ChatGPT, Google AI Overviews, and Perplexity cite different sources for the same question. 25 live queries, the reconciliation, and the minimum viable Surface Diet.

46.7%

of Perplexity citations are Reddit threads — versus roughly 11% on ChatGPT

May 8, 202625 min readRead the study →

[ In the pipeline ]

  • The Citation Volatility Tracker

    a recurring monthly index of how much AI citations churn — 40–60% of cited domains change every month

  • The SSR Audit

    100 B2B SaaS sites tested for whether AI crawlers can read them at all

  • The Claude Benchmark

    the most selective AI engine, measured on B2B SaaS brands specifically

Want the next study when it publishes — or your brand included in a benchmark? Tell us here.