Search Disrupted Newsletter (Issue 33)

Google's AI Mode report shows search becoming a 'do' surface. How marketers fail at Reddit. Google really wants to look like Claude. We can drop AEO and GEO terms as AI is all search now.

Michael Buckbee

Google’s AI Mode Usage Report Leaves a Lot Unsaid

AI Mode passed one billion monthly active users globally, with queries doubling every quarter. AI Mode and similar conversational search experiences are rapidly becoming the standard, mainstream way people interact with Google.

Three numbers from the report stand out:

  • AI Mode queries are 3x the length of a traditional Google search query. The top first words are “What,” “How,” “I,” “Is,” and “Can”: full questions, conversationally phrased. (Which is why within Knowatoa we track questions, not keywords.)
  • More than 1 in 6 AI Mode searches are non-text (voice, image, video), and image-input queries are growing more than 40% month-over-month.
  • Planning queries grew 80% faster than AI Mode overall, and “which” queries (comparison-shopping language) grew 40% faster. Search is increasingly a “do” surface, not just a “find” one.

As people shift to AI Mode for planning trips, comparing products, and building study guides, the result is search evaporation. It’s unsaid in the report, but each AI Mode query is replacing dozens of individual queries people used to stitch together manually.

Showing up in the bottom-of-funnel questions that directly ask for brand recommendations is what we focus on at Knowatoa.

Link: How AI Mode is changing the way people search in the U.S. (PDF)

AI Mode one-year retrospective
AI Mode one-year retrospective

Marketers keep misunderstanding Reddit

Reddit has an undeniably oversized influence on search results. Too many people treat it as just a tactical SEO lever, when it’s a community and ecosystem that shapes how your brand is perceived as a whole.

Ann Robison at Search Engine Land does a great job of untangling three things:

  • Lived experience. The product page says “battery lasts 30 hours.” The Reddit user says “mine drained in a day once I started tracking workouts.” Same information, plus the context people actually use to decide.
  • Disagreement. Polished brand content avoids nuance and caveats. Reddit threads contain conflict, humor, unexpected use cases, and people changing their minds mid-discussion. That makes it more useful for AI systems trying to answer subjective questions.
  • Perceived authenticity. Users aren’t paid to persuade you, and that perception matters more as influencer content gets more sponsored.

The takeaway is broader than “post on Reddit.” Content that wins citations brings real customer experience, tradeoffs, and acknowledged limitations. The same logic applies anywhere the lived experience for your industry actually lives: forums, Stack Exchange, Yelp, professional communities.

All of which you can track under the Citations tab in Knowatoa.

Link: Reddit’s influence on AI search extends beyond training data

Reddit and AI search
Reddit and AI search

Google’s idea of the future of search looks like Claude

Google released a sizzle reel that’s a flashy visualization of how they view the future of search. It emphasizes less of search as a feature and more that larger projects and tasks will be done via what used to be a search interface. In this way it looks very much like Claude or ChatGPT, where it’s a blend of search, generation, discernment, decision-making, research, and reporting all rolled into one.

Link: What’s New in Search (Google I/O 2026)

What's New in Search
What's New in Search

Google’s AI optimization guide is… fine

I’m not sure I’ve ever seen so many wrong hot takes as I did in the wake of Google’s official AI optimization guide. Undeniably we have a biased perspective here, but based solely on the words Google wrote down in this guide, there is a massive future for AI search.

A few points worth calling out:

  • It’s clear the article isn’t signaling the death of AI search so much as it is that there’s no longer any traditional “10 blue links” search. All search is AI search now (AI Overviews, AI Mode, etc.).
  • This is a guideline purely for Google’s search index. It’s not a global statement on how Claude or ChatGPT or any other service works.
  • Search is just one kind of AI activity happening now, and these recommendations only apply to that. Case in point: llms.txt won’t help your search visibility, but if you’re building a service that interacts with AI tools, it can be a useful public set of guidelines for how your site should be approached and what’s available.
  • “Just write good content” has been Google’s advice for years, and in today’s ultra-competitive internet that feels lacking and too broad. The article specifically calls out a shift we’ve been trying to get content strategists to make: broad informational top-of-funnel content (which used to be the largest driver of traffic) is going away, and your strategy should be much more focused on middle and bottom of the funnel content.

The old approaches to content and search ranking have dramatically shifted.

Link: Optimizing your website for generative AI features on Google Search

Google AI optimization guide
Google AI optimization guide

Thanks

Thanks for reading. These take real time to put together and the replies are what keep them going. Hit reply and tell me what you’re seeing on your dashboards.

Also wanted to give a quick thanks to everyone who hopped on a call this week and shared kind feedback on our BISCUIT framework. Really appreciate it.

p.s. It would really help me out if you could Follow me on LinkedIn

Get all the words above (and more) delivered to your inbox every week.