Google Says Don't Optimize for AI

Meanwhile, AI Search Results Are a Mess

Google's Search team recently weighed in on the "GEO" conversation—that's Generative Engine Optimization, for those keeping score on marketing acronyms. Their advice? Stop chunking your content into bite-sized pieces for LLMs.

Fair enough. But that guidance sidesteps a much bigger problem—and doesn't offer businesses much to work with.

Here's the good news

Unlike a year ago, you're not flying blind anymore. The tools to measure AI visibility have caught up. You can actually see where you stand, where competitors are beating you, and what's worth fixing. The situation is messy, but it's not hopeless.

Let me break down what Google said, what they're not talking about, and what you can actually do about it.

What Google Actually Said

Danny Sullivan addressed the chunking trend directly on Google's Search Off The Record podcast:

We don't want you to do that. I was talking to some engineers about that. We don't want you to do that. We really don't. We don't want people to have to be crafting anything for Search specifically.

His point is reasonable: a well-structured webpage with proper headings and logical organization already provides the structure that both humans and machines need. You don't need to artificially chop up your content.

Sullivan also warned against optimizing for today's AI systems:

And then the systems improve, probably the way the systems always try to improve, to reward content written for humans. All that stuff that you did to please this LLM system that may or may not have worked, may not carry through for the long term.

Anyone who's been in this industry for more than a few years has seen this movie before. Tactics that game current systems rarely survive algorithm updates.

The Real Problem Nobody at Google Wants to Discuss

Here's what's missing from this conversation: referrals. Traffic. The thing that actually matters to businesses investing in content.

When an AI Overview appears, click-through rates for the #1 organic result drop by 34.5%. And 69% of searches now end with zero clicks at all. That's not a minor adjustment. That's a fundamental shift in how search works.

But it gets worse. Google's AI systems use something called query fan-out. When someone searches for one thing, the AI decomposes that into dozens of related sub-queries behind the scenes, then synthesizes an answer. One user question, multiple queries, one consolidated response.

What does that mean for you? When one AI Overview addresses five questions at once, that's four fewer opportunities for your content to appear. Four fewer chances to earn a click. The math is brutal.

And here's the kicker nobody's talking about: the sources Google's AI is actually citing are often garbage.

Search Engine Journal's Roger Montti ran a simple search about styling a sweatshirt. The sources Google's AI Mode cited? An abandoned Medium blog from 2018 with broken images. A LinkedIn article. Content from a sneaker retailer's website.

Meanwhile, actual expert sources—publications like GQ that employ people who know fashion—were buried under the "More" tab.

This isn't an isolated example. Anyone using AI Overviews regularly has seen this pattern: authoritative sources get sidelined while random content gets promoted.

What This Actually Means

If you're running marketing for a mid-market business, this creates a real strategic problem.

• The traffic math changed

Query fan-out means fewer opportunities to rank, period. Your one-keyword-one-ranking optimization playbook is broken.

• Quality doesn't guarantee visibility

You can do everything right—create genuinely useful content, structure it properly, establish real expertise—and still watch AI search surface competitors with thinner credentials.

• You're flying blind

Most analytics tools weren't built to track AI visibility. You might be getting cited in AI Overviews without any way to know it. Or you might be completely invisible. Either way, your dashboard isn't telling you.

What To Do About It

Abandoning search isn't the answer. Neither is obsessing over every AI-specific tactic that trends on LinkedIn.

• Build for humans first

Google's advice here is actually sound. Content that genuinely serves your audience has always been the foundation of sustainable search performance. That hasn't changed.

• But don't ignore AI visibility

Understanding where and how AI search surfaces your content—or doesn't—gives you data to work with. You can't fix what you can't see.

• On measurement

The tools to track AI visibility actually exist now. Platforms like Semrush, SE Ranking, Otterly, and Profound can monitor whether your brand gets cited across ChatGPT, Perplexity, Google AI Overviews, Gemini, and Claude.

They show you where competitors appear and you don't. They track citation context—not just "mentioned" but "how you're positioned."

The category has matured fast; AI Overviews grew 115% since March 2025, and the measurement tools caught up. We run these audits for clients regularly. It's not guesswork anymore. If you want to know where you stand in AI search, you can knowand we can show you.

• Diversify your traffic sources

Any channel that controls 60%+ of your leads is a business risk. If Google's AI features cut your search traffic by 20%, what's your backup plan? If you don't have one, that's the conversation to have right now.

• Focus on what AI can't replicate

Original research. Proprietary data. Genuine expertise from practitioners. These assets stand out in a landscape increasingly cluttered with AI-generated content citing other AI-generated content.

Bottom Line

The GEO versus SEO versus AEO debate is noise. What matters is whether your target customers can find you when they need what you sell.

Google is still figuring out how to balance AI convenience with result quality. That's their problem. Your problem is maintaining visibility and lead flow while they sort it out.

The businesses that come through this in the strongest position are the ones treating AI search as a variable to monitor and adapt to—not a trend to chase or a threat to panic over.

No secret tactics. No magic frameworks. Just clear-eyed assessment and disciplined execution.

Have questions about how AI search is affecting your specific market?

Let's talk about what the data actually shows for your industry.

Sources:

• Google Search Off The Record Podcast

• Search Engine Journal

• OnSaaS

• Backlinko

• Otterly.AI

• SE Ranking AI Search Toolkit

• LLMrefs

• Profound AI

• Peec AI

• Semrush AI Toolkit