https://lffafeqydexbjcnuctew.supabase.co/storage/v1/object/public/notes-audio/notes/35c8c368-5191-8075-a2c0-db9492576f54.mp3
🎧 Listen to this note (9 min): Open player
<aside> 🚨
Want more notes like this? Subscribe to weekly strategy emails: ainotebook.com
</aside>
First, I want to address the elephant in the room… Google’s recent stance on GEO/AEO.
Here are my thoughts.
With that out of the way…
Howdy Notebooker!
Last week I broke down Google's Q1 2026 earnings call and what it means for AEO. I focused on the AI Overviews and AI Mode quotes, plus the new ad format Google is testing inside AI Mode.
There's one quote from Sundar I didn't dig into that's been bugging me ever since. It's about Search latency.
"Even as we have brought new AI features into our results page, we have reduced Search latency by more than 35% over the past five years."
https://www.youtube.com/live/LPJoiDiVkTI?si=QtKbLIYAbU0jjiGr&t=459
Sundar said it like a brag, and by classical Search logic it is one. Faster Search has been a north star at Google for two decades.
But a recent study titled The Impact of Response Latency and Task Type on Human-LLM Interaction and Perception just dropped that suggests something more interesting than a Google product question. It suggests users have already used response speed to sort AI products into a trust hierarchy, and that hierarchy should change how you think about which LLMs your brand needs to be inside.
Felicia Tan and her co-authors at NYU ran a controlled experiment with 240 participants. Each person was assigned to one of three latency conditions (2, 9, or 20 seconds) and one of two task types (Creation or Advice). Same GPT-4o model. Same content. Same quality.
The only thing the researchers changed was how long the AI made you wait before it started typing.
Then they had participants rate the responses across seven dimensions of perceived quality.