SEO Experts Develop Unofficial Methods to Track Google AI Queries

Image credit: Search Engine Journal

SEO professionals globally are developing unofficial methods to monitor Google AI Mode conversational queries within Search Console because Google’s official reporting offers limited insights.

Google introduced Search Generative AI performance reports in Search Console on June 3, 2026, providing data for AI Overviews and AI Mode impressions; however, the Search Analytics API and BigQuery bulk export do not officially expose query-level generative AI data, according to multiple industry observers.

Fragments of AI conversations sometimes appear as ordinary queries within Search Console, prompting experts to devise workarounds. Glenn Gabe, an SEO consultant, developed an API-to-Excel approach utilizing Claude to extract these queries.

Jean-Christophe Chouinard, another expert, created custom regex patterns for in-report filtering to identify AI Mode queries. Amin Foroutan contributed to these efforts with his Advanced GSC Visualizer Chrome extension, which aids in data extraction.

A proposed custom machine learning model, based on FacebookAI’s xlm-roberta-base and fine-tuned on multi-language Search Console data, aims to classify AI queries more accurately. This model is designed to handle edge cases and code-mixed languages, according to its developers.

While AI Mode clicks are tracked in Search Console, the specific queries leading to those clicks are almost exclusively anonymized. This anonymization presents a significant limitation for all current tracking methods, Barry Schwartz, a search industry reporter, reported.

Google officials, including John Mueller, have acknowledged the rollout of AI features in Search, but have not detailed plans for more granular AI query reporting. Anastasia Kourou and Metehan Yeşilyurt are among other experts contributing to the discussion around these tracking challenges.

The unofficial methods highlight the SEO community’s efforts to understand user behavior in search driven by generative AI. Vijay Chauhan and David Konitzny have also shared insights on the complexities of analyzing AI-driven search traffic.

Source: Search Engine Journal