Staying on top of everything happening in the world of web and search sometimes feels like drinking from a fire hose (and I actually *like* this stuff!) If you’re a small business owner who isn’t living in code all day, I imagine it feels even more overwhelming. So when Google announced some fairly sizable changes to how their search algorithm works, I wanted to make sure I understood it fully before sharing the info. This frequently means using my own site as a testing ground. Here’s what I learned, what I changed and why it might be worth doing the same on yours.
First, a Quick Vocabulary Lesson
Before I get into what I actually did, let’s get two terms out of the way because you’re going to keep seeing them (more than likely you’ve already seen them but aren’t sure what they mean, so let’s break it down): AEO (Answer Engine Optimization) and GEO (Generative Engine Optimization). Both refer to optimizing your site so that AI-powered search tools (Google’s AI Overviews, AI Mode, ChatGPT, Perplexity, etc.) find your content trustworthy enough to cite in their answers.
Here’s the thing Google actually said that I found reassuring: optimizing for AI search is still just SEO. (That was kinda how I have been approaching it, so it’s good to feel validation, ha!) Basically, the same principles apply: good content, clear structure, genuine authority. But how Google actually surfaces that content has shifted and this is the important part, so read on.
What’s Actually Different About Search Now
There Are More Places to Lose Someone Before They Reach Your Site
The old model was clean: type a search query → get a list of links → click one → land on a site. That path still exists, but there are now multiple places where a user can get enough of an answer to stop looking entirely and drop off.
It looks more like this now:
Query → AI Summary → cited sources → additional links → your site
I hadn’t noticed the pattern until I read this update from Google, then the dots started connecting for me. I mean, I was searching up viral cookie fries recipes and never once clicked through to a website. The answer was just… there.
Data from SparkToro showed that, in 2024, 58.5% of U.S. Google searches already ended without a single click. Updated numbers saw that number climb to 68% in 2026. When an AI summary is present, users are roughly half as likely to click through to a website compared to searches without one. That’s not the end of the world, but it does mean your site needs to either be the cited source inside the summary, or give people a compelling reason to click through anyway. Generic content isn’t going to cut it.
The First “Visitor” to Your Site Might Not Be Human
Google introduced what they’re calling “Information Agents” which is (are?) AI that runs searches in the background without a human actively present. You describe what you’re looking for, the agent crawls blogs, news sites, social posts, and service pages, then reports back with what it thinks you needed.
Thinking about this when I was looking at my own site meant that I stopped asking “does this read well?” and started asking “could an AI reading this figure out exactly what I do, who I help, where I’m located, and what it costs to work with me?” Those are different questions, and the answers led me to make some specific changes… more on that in a minute.
Google Is Now Searching Across Questions You Didn’t Ask
Query fan-out is the term for something Google expanded with this update. When someone searches, Google can now break that query into multiple related sub-questions and pull answers from across all of them simultaneously. A search for “WordPress designer in Napa” might lead to sub-searches for “how to hire a WordPress designer,” “what does a website redesign cost,” and “what to look for in a web developer” and then weave the results together into one response.
The more clearly your site answers the real questions people have about what you do, the more likely you are to show up somewhere in that response.
What Google Is Now Rewarding
The May 2026 core update made a few things clearer. Google significantly strengthened E-E-A-T signals (Experience, Expertise, Authoritativeness, and Trustworthiness), but it also started actively penalizing content that lacks real depth or unique insight. This is why low-quality (read: straight AI-generated) content is not great.
The other thing that was very apparent in everything I read was surrounding non-commodity content. If your copy could belong to any competitor, i.e., content with vague value propositions, uses generic language (in my world it’s a lot of the same “passionate about helping small businesses” language you’ve seen a thousand times), Google’s AI will summarize it rather than send someone to you. You become a data point, not a destination, which made sense to me. Think about looking at hotels in a new vacation spot. If they all tout a pool or the same amenity, the perk is no longer special. People want to know what makes you stand apart from the rest!
Reviews, case studies, documented outcomes, specific proof. Those matter more right now than they ever have. (This is also something that I’m still working on for myself. I’m terrible at asking for reviews, especially ones through Google but apparently I need to get on board– eep!)
What I Actually Changed on My Own Site
Okay, here’s the practical part. After going through my own site with all of this in mind, these are the areas I focused on… and what I’d encourage you to look at too.
I Made Sure the Basics Were Solid
Fast load times, compressed images, a clean plugin list. None of that is new advice, but it matters more now because AI agents crawling your site for citable content don’t wait around for slow pages. If you haven’t done a performance check lately, that’s the place to start. I’ve written about image optimization and the plugins I actually rely on if you need somewhere to begin.
I Audited My Logistical Information
Pricing, location, services, availability. I went through my site asking: if an AI agent landed here, could it quickly determine what I do, what it costs, and where I’m based? Napa Valley WordPress designer is something an agent can work with and pass along to a human. Creative digital solutions is not. Though still a work in progress, I am trying to tighten up the specifics wherever I find vague language.
I Strengthened My FAQ Content
This is probably the highest-impact change you can make. FAQ content is highly citable because it’s structured, direct, and answers the exact questions people are already asking. I went through mine and made sure I was answering the real questions clients ask before they hire me: about process, timeline, what’s included, what happens after launch. I tried not to use too much marketing language because I wanted AI to find readable answers here, while also explaining it for any human who is actually reading it.
In my case, if a potential client’s AI agent is searching for “what does a WordPress website redesign include” I want my FAQ to answer that question clearly so I have a shot at being the cited source.
I Verified My Structured Data Was in Place
Structured data sounds more intimidating than it is. It’s a way of tagging information like business name, location, services, reviews, so that search engines and AI can read it without having to dig. If you’re using Yoast SEO or another SEO plugin, most of this can be configured without touching any code, but it’s worth checking that it’s set up and accurate.
I Updated My Google Business Profile
For local businesses especially, your Google Business Profile is doing a lot of heavy lifting right now. I made sure mine was current, with accurate location, updated services and I added some of my blog posts to it. I also made sure the info matched what’s on my site. Consistency between these two things matters for how Google understands and presents your business.
And if you haven’t connected Google Search Console yet, do that now. Free, quick to set up, and it starts collecting data immediately. I walked through how to do it in The WordPress Must-Do List for Small Business Owners.
Revisiting Page Content and Meta Data
One thing I haven’t fully tackled yet (but am ticking away on) is leading with answers earlier on key pages. AI systems pull from early content when generating summaries, so if your most important information is buried three paragraphs in, it may not make the cut. This is a challenge for me and my long-winded ways! But it’s on my to-do list and I’d encourage you to think about it too.
What I did do in the meantime: I pulled up Google Search Console and looked at pages with strong impressions but low click-through rates. That tells you Google thinks your content is relevant enough to surface, but something about how it’s presenting isn’t compelling enough to get the click. In most of those cases, I revised the meta title and meta description to be more direct and specific. It’s a smaller lift than rewriting a whole page, and it’s a worthwhile place to start while you work through bigger content updates.
One Thing I’d Caution Against
Don’t start chasing “GEO hacks.” Google called these out directly in their official guidance. Inauthentic brand mentions, keyword-stuffing for AI, and other shortcuts are more likely to work against you than for you. The goal is to be genuinely useful and clearly structured, not to game a system that’s specifically being built to detect that kind of thing.
A Note on Brand Voice
With all the suggested edits for AI optimization, that does not mean stripping the personality out of your website. You don’t have to choose between brand voice and machine-readable content. They serve different jobs, and a well-built site does both.
Think of it this way: the AI’s job is to figure out what you do, who you serve, where you’re located, and whether you’re a credible source worth sending someone to. That part of your site (your services page, your FAQ, your logistical details) should be clear, specific, and direct. A good example of where this goes sideways: about 10 years ago, I remember all these service-based businesses wanted to shift and call their pricing an “investment.” Psychologically clever, but if someone’s AI agent is searching for “how much does a WordPress website cost,” a page that only uses the word “investment” is essentially invisible to that query. The AI can’t connect those dots. Calling it what it is — pricing, cost, rates — is what gets you found. Your personality lives in how you talk about it, not in what you call it.
The approachability, humor (I try!), and voice woven through the rest of your site is what makes a human actually pick up the phone after the AI points them your way. Both matter. They just matter at different moments in the process.
The Bottom Line
Going through my own site with a GEO lens wasn’t a dramatic overhaul… it was more like a focused tune-up. The fundamentals of good web content haven’t really changed. What did change was the importance of being specific, structured, and substantive enough that both humans and AI can quickly understand who you are and why you’re the right choice.
One caveat: it’s only been a few weeks since I made these changes, so I don’t have results to report yet . I’ll be watching for jumps in traffic or click-throughs to see if it’s working. But based on what Google announced, it feels like the right approach. SEO and GEO both take time to show their effects so I’ll likely revisit this in a few months with an update on what actually seemed to work!
If you want to do this audit on your own site and aren’t sure where to start, or if you’d rather hand it off entirely, let’s talk. →