The CMO's Guide to the New Search Era: From Personalization to Customization

By the Numbers

• AI Overviews now trigger on 59% of informational queries, but only 19% of commercial ones.

• A "Preferred Source" label makes your brand 2x more likely to get the click.

• 73% of top-ranking pages now show author credentials, up from 58% before the 2026 updates.

Search has felt like chaos for about a year and a half. That's ending. This article explains the shift in plain terms, shows what the data says, and lays out what it means for how your brand earns visibility now. You won't need to blow up your strategy. You will need to update one belief: ranking is no longer the goal.

From Personalization to Customization

We're moving from personalization to customization. Google is rolling out a customization model, and it matters more than personalization ever did.

For years, Google used personalization to guess what you wanted. It read hidden signals about you and quietly shaped your results. So when a million people ran the same search, they got a million different windows, viewpoints, and contexts. Nobody chose any of it. The guessing happened behind the curtain.

Customization works differently. There's still one shared starting point for a search. The base is the same for everyone. But from there, each person decides what they see. They pick:

• The channels they want

• The sources they want

• Whether their results get regionalized or not

Those are choices, not guesses. That's the whole difference. Personalization means no choice. Customization means you get to pick.

That's what Google seems to be building. Choices.

For a modern CMO, that's a relief. After eighteen months of noise, it adds something we haven't had in a while: a little clarity, and a little predictability.

The Scale of the Shift, by the Numbers

AI in search is no longer a test. It's the infrastructure now. The pilot phase is over, and the change is happening at full scale. Current SEO playbooks are running straight into a "zero-click" problem, where people get their answer without ever clicking through.

Here's what the data shows:

• The click gap. In 2026, informational queries saw a 61% drop in organic click-through rate. People are asking questions and getting answers right on the page.

• The volatility. In early 2026, Google pushed three major updates close together: the February Discover Update, the March Spam Update, and the April Core Update. Rankings swung in ways we hadn't seen before.

• The intent divide. AI Overviews now show up on 59% of informational queries but only 19% of commercial ones. Translation: the top of the funnel, where people are just learning, is now almost entirely zero-click.

• The trust signal. 73% of top-ranking pages now show detailed author credentials, up from 58% before the 2026 updates.

Put those together and you get the same conclusion. The old backbone for visibility is cracking. The new one runs on a high-efficiency engine and a different kind of reader.

Gemini Flash: The Engine Behind Agentic Search

Modern search runs on Gemini Flash. It's a fast, low-cost model built for "agentic" work, meaning AI that goes off and completes tasks on its own.

This changes who you're writing for. Your reader is often not a person anymore. It's an AI agent, like Claude Code or Gemini, grabbing your content and turning it into an answer for the executive who actually asked the question.

So the audience you've been optimizing for may not be the one reading your page.

From SEO to Brand Governance

Here's the mindset shift. Agentic Engine Optimization, or AEO, isn't a technical chore. It's brand governance.

Think about what happens when an AI agent visits your site. If your documentation is bloated or badly structured, the agent does one of two things. It cuts your value proposition short, or it makes something up, sometimes pasting a competitor's features into your brand's profile. Either way, your brand gets misrepresented and you never see it happen.

That's why this is a governance question, not an SEO one. You're deciding how machines describe you when no human is watching.

The invisible audience. A human spends four to eight minutes skimming your site. An AI agent sends one request and "reads" the whole page in about 400 milliseconds. Your analytics will log that as a bounce with zero time on site. So your most valuable visitor, the one shaping the answer an executive sees, shows up in your reports as nothing at all.

The User-Controlled Results Page

The clearest sign of the move toward customization is a feature called "Preferred Sources." It's not just a ranking factor. It works like a loyalty program for search.

When a signed-in user picks your brand as a preferred source, your results get a "Preferred" label in places like Top Stories and AI Overviews. And when that label is showing, your brand is twice as likely to get the click.

So the goal has changed. Ranking number one in the general results isn't the prize anymore. Being a chosen source inside someone's dashboard is.

That sets off what I think of as a trust loop:

1. Selection. The user tells Google they trust your brand.

2. Prioritization. Google starts giving your content a "Preferred" label.

3. Authority. Future AI summaries get built using your brand as the main source of truth.

Each step feeds the next. Once you're in the loop, you compound.

The New Trust Signals

To fight off AI-generated filler, Google has leaned hard toward human-centric content and original reporting. If you watched LinkedIn's 2025 algorithm change, this will feel familiar. There, personal authority and real lived experience became the things that drove reach. Search is now doing the same.

Here's the trust hierarchy I'd watch in 2026:

• Information gain, meaning original reporting. Content with data or a point of view that doesn't already exist in the top ten results. New information, not a rewrite of everyone else's.

• Highly cited labels. A system Google expanded in 2026 that rewards whoever broke the story. If your brand is the origin of a fact, you get the citation, even if a bigger site has more domain authority.

• The perspectives carousel. A part of the page built to surface real human conversation from places like Reddit and TikTok. Authentic talk beats polished corporate prose here.

The pattern is simple. Be the source of something, not the echo of it.

GEO and AEO: A Practical Roadmap

Technical SEO has split into two jobs. GEO, or Generative Engine Optimization, is about how you show up inside AI summaries. AEO, or Agentic Engine Optimization, is about how AI agents read and use your site directly.

Here's the readiness checklist I'd run:

• Mind your token economics. A token is roughly a chunk of text an AI reads at once. Keep your core "Quick Start" pages under 15,000 tokens, and individual pages under 30,000. Go past that and an agent is likely to truncate your page or skip it.

• Choose structure over style. Serve content in Markdown, not heavy HTML or JavaScript. Agents read Markdown with far less effort, which means your message gets parsed accurately instead of mangled.

• Set the gatekeepers correctly. Add an llms.txt file at the root of your site. Think of it as a sitemap built for AI agents. And check your robots.txt. If it's set to block AI crawlers, you've quietly locked your brand out of AI search without realizing it.

None of this is glamorous. All of it decides whether you exist to a machine.

Measuring Value When Clicks Disappear

With clicks down 61%, "total traffic" is a vanity metric. It tells you almost nothing now. CMOs need to measure assisted conversions and cost avoidance instead.

Here's the framework I'd use.

1. The ROI formula. ROI = (Revenue − Cost) / Cost. Count revenue as the customer lifetime value of leads that organic search influenced, even when the final click came through another channel. Organic search often starts the journey it never gets credit for finishing.

2. The cost avoidance math. Compare your AI citations to what that visibility would have cost in paid search. Say you show up as a cited source in an AI Overview for a keyword with a $15.00 cost per click, and you earn 1,000 impressions in that overview. That's $15,000 in cost avoidance. You got the exposure without paying for the clicks.

3. Assisted brand lift. Watch your brand search growth. When AI mentions your product, it often pushes someone to go search your company name directly right after. A rise in branded search is one of the clearest signs your AEO is working.

The Approach: Three Pillars

The search channel isn't dying. It's becoming a curated, high-trust dashboard. The blue link is being replaced by the cited source. To lead in 2026, your marketing organization has to stop trying to "rank" and start becoming "source-worthy."

That comes down to three pillars:

1. Be chosen. Move from broad keyword targeting to the human authority it takes to become a Preferred Source. People have to want you in their dashboard.

2. Be cited. Invest in primary research and information gain. Originality is the only real defense against the AI-generated middle.

3. Be parsable. Treat your site structure as brand governance. If your site isn't token-efficient for AI agents, your brand effectively doesn't exist to them.

In this era, source worthiness is the new PageRank. Focus on the depth of your analysis and the clarity of your data, and the engines of the new search era will follow.