Oh, friends. We've had quite the kerfuffle lately, haven't we? We were all eagerly awaiting new search announcements from Google I/O, only to have AI Mode jump out from behind a bush and declare itself the new SERP. This was immediately followed with the May Core Update, which felt a bit rude. Then the company gave us Search profiles. By mid-month, we're feeling like a kid getting a toy from the dentist. My jaw hurts but this sure is a fun sticky hand. It'll definitely last a looooong time.

Not all users were chuffed to see AI saturate their search experience. 57% prefer traditional search for important topics. Over 80% haven't used genAI regularly. DuckDuckGo reported that traffic to their "No AI" search page tripled. SEO hot takes quadrupled. Neither mattered in the scale of Google. Math is funny that way.

It's still worth trying to understand how users interact with AI and what's standing in the way for those that don't want the devil's SERP. Dan Taylor puts it so well:

The tripling of traffic to DuckDuckGo’s “No AI” search page is a direct reaction to users not having a choice.

Did that kick you right in the autonomy? You'll love the rest of the article. In a breezy five minutes, you'll get useful stats and the five barriers to user trust. TL;DR– some of you don't have trust issues and it shows.

The punchline of all this is that the path to reach both users is the same: you need to be the one they trust. Google’s AI Mode, Gemini, ChatGPT, and all LLMs work with the same critical mechanic. They lean heavily on memory and personalization to curate what lands in front of your face. It turns out that showing up in these bespoke answers isn't just a domain authority flex. More often, it’s about the existing vibes: the newsletter you actually read, that YouTube creator you follow, or the loyalty program you visit before hitting "book."

Intrigued? 👑 Crystal Carter has 10 ways AI personalization is powered by engagement and how to make the most of it. This isn't a ridiculous "get users to shar your schema on Facebook" listicle. We're talking about how memory is inferred from habitats and patterns. There's things we know about ourself and things others know about us that we don't.

LLMs are essentially lurking in the shadows of your digital habitat, quietly noting your queries and the creators you actually bother to follow. They’re stitching together a version of you based on your recurring patterns and vibes, and here’s the kicker: you don’t get a vote in what they decide you care about.

The advice on using brand activations and visual search to build affinity at scale is particularly clever. Your Google Photos are now formally linked to Gemini’s Personalized Intelligence. This means the snapshots lurking in your account—from event captures and social media saves to those featuring specific brands or locales—are being stitched into the context the AI uses to map out your world.

This is meaty. It could be a whole article by itself. Oh look, it is. Myriam Jessier has previously given us magnificent guides to image and video optimization for AI. This time, they're back with an in-depth look at how human brains and LLMs default to the familiar.

For fun– open your favorite AI system and play with a couple of makeup prompts. Chances are, you'll get content about Clinque's Black Honey lipstick in less time than it takes to get a coffee. There's a reason for this. The way Google gauges your brand's importance is essentially via a digital map called RepositoryWebrefLatentEntities. It connects your brand to other hidden entities, alongside systems like RepositoryWebrefKGCollection. The more co-occurrences you have, the higher your mutual information score, and the more often you appear in answers.

Clinque's team is very, very good at leveraging this. Or they made a good product? Even if you believe they definitely retro-engineered the lipstick to match a banging GEO campaign, the end goal is still the same. We want to be the one they remember.