Masters of Search Conference 2027
← The Brief

Landwehr's 190-day log finds only two LLMs leaving traces in GSC and Bing Webmaster Tools

A long-running check of both consoles shows Google's AI Overviews and AI Mode in GSC, and Bing Copilot in BWT, with everything else invisible.

Niklas Buschner
Screenshot from Malte Landwehr's original page
Screenshot: Malte Landwehr

Malte Landwehr spent 190 days checking Google Search Console and Bing Webmaster Tools to see which LLMs register inside either console. He published the results on LinkedIn on 9 November 2026.

Google's AI Overviews and AI Mode produced impressions in both the Search and Generative AI reports inside GSC. Bing Copilot produced citations in the AI Performance report inside BWT. On his data, those were the only LLMs that left a clear, repeatable trace in either console.

Gemini did not register impressions in GSC at all. Perplexity and Grok did not show the scraping patterns he expected across either tool. Yahoo Scout was also absent, which Landwehr flags as surprising given it is openly based on Bing WebIQ.

ChatGPT was the moving part. For logged-out users it ran on Google through July and August, then switched to Bing, with each phase visible as spikes in the matching console. Logged-in traffic stayed inconsistent across the whole period, which Landwehr reads as a mix of Google, Bing and ChatGPT's own index, Labrador. He notes the BWT spikes are now less reliable than the earlier GSC ones, which he takes as a sign more grounding is going through Labrador.

He ran his prompts from US and German IPs. GSC traffic on ChatGPT days also arrived from the UK and the Netherlands.

Reading the consoles against the wider set of assistants

We think the useful read for practitioners is narrow. GSC and BWT will tell you about four surfaces: AI Overviews, AI Mode, Bing Copilot, and whichever search engine ChatGPT is routing through this week. Everything else, including Gemini, Perplexity, Grok and Yahoo Scout, is invisible to both consoles on this sample.

That means a console-only view of AI Search understates your exposure. Referrer logs, server logs and direct testing in each assistant are the places where the other surfaces actually show up. Before changing anything based on a console trend, check whether the assistant you care about was ever counted in it.

The ChatGPT routing shift is worth watching rather than acting on. One practitioner's log over 190 days is a strong signal that the plumbing moved, not a guarantee it will stay where it is next quarter.