Small Business Owner: Here's How You Adapt To AI SEO

Search engines feel like they’ve been around forever. But we all know this isn’t so.

Back in the late ’90s, you didn’t “Google it.” You asked Jeeves.

People typed out full questions like “How do I boil an egg?” or “What’s the best car for snow?”

And with formal aplomb, Jeeves answered your questions—until Ask.com popped with the dotcom bubble.

Then Google taught us to speak its language: “boil egg,” “best car snow.” For twenty years, we adapted to search like computers, not like people.

Now the pendulum is swinging back. Since ChatGPT launched in 2022, people are typing questions again. Long ones. With nuance.

Unlike Jeeves, this machine talks back, in natural(ish) language, no less. And with as many follow-up questions as you like.

Google responded by pushing generative AI directly into Search. What began as Search Generative Experience is now AI Overviews and AI Mode: systems that can synthesize information, answer follow-up questions, and show links without forcing the user through the old ten-blue-links workflow.

It’s the original promise of Jeeves, fulfilled at last.

But there’s a huge problem: a lot of businesses are seeing their traffic drop as a direct result. That’s a problem. And now marketers and business owners have to figure out what to do about it.

And to do that, we need to talk about how…

Answer engines are tools that pull from multiple sources, condense the findings, and give a single polished response without requiring a click.

Now if you’re a small- or mid-size business that relies on organic search, this is, frankly, scary as all hell. It means your pages can lose traffic—and leads—even while your pages “rank” in the algorithm.

We now have better evidence than the early forecasts. In a Pew Research Center analysis of U.S. browsing behavior from March 2025, users clicked a traditional Google result on 8% of visits with an AI summary, versus 15% of visits without one. That does not tell us what will happen to every site, but it is strong evidence that AI answers can reduce clicks even when search visibility remains high.

That means people are getting answers straight from Google AI Overviews, to say nothing of ChatGPT, Perplexity, Claude, Bing Chat, DeepSeek, or others. More and more people are making decisions without visiting company websites.

But there is also a consolation, as Search Engine Land also reported a 527% increase in LLM referral traffic across 19 websites—proof that these visitors are still finding websites, but through new means.

There’s now an entire vocabulary around this—AEO, GEO, AI SEO, LLM optimization—but I would be careful about treating it as a separate discipline with a bag of secret tricks. For Google, at least, the company’s own guidance is blunt: the same foundational SEO practices still apply to AI Overviews and AI Mode, and there is no special AI schema or machine-readable markup required.

Clear structure still helps readers, search engines, and AI systems understand a page. But I would not treat FAQ blocks, schema, or any particular formatting pattern as a magic AI-selection lever. The safer rule is much less exciting: make the page easy to crawl, easy to understand, factually accurate, internally linked, and useful to the person who lands on it.

This is all very important because it will help business owners continue to win traditional search engine traffic and boost their visibility in various AI systems. But I think that before I start rattling off solutions, we need to talk about a deeper question first.

People did not start using search engines because they wanted web pages. They used them because they wanted answers. And silly and nostalgic as my Jeeves metaphor may have seemed, I think that old avatar of the dotcom era is proof of this fact.

Case in point: search behavior is changing fast, and it’s not just a shift in where people click. More users are getting what they need without ever leaving the page. These are called zero-click searches, and they’re on the rise. The old SEO model—rank high, get the click—doesn’t hold up when the answer shows up before your link does.

That doesn’t mean traffic is gone. Like I said, search engines are still a huge source of traffic for websites and LLM-driven traffic is on the rise too.

It’s also worth pointing out that if someone wants what you’re selling, they still have to interact with you in some way, be it a phone call, email, purchase, or visit to your website. And visibility in search engines, even if it often results in a zero-click outcome, still makes all of these things more likely to happen (even if you can’t measure it).

I cannot stress this enough: there are lots of scenarios where your brand can get more impressions, fewer clicks, and the same (or more) leads or sales.

Consider this: queries are getting conversational. So if I’m looking for an email marketing vendor, for example, then instead of typing in keywords, I can type in a multi-sentence question that’s really specific. And rather than getting a list of “10 best email tools” and high-ranking vendors like Mailchimp and Klaviyo, I can ask 5 follow-up questions to make sure I’m really getting what I need.

Now, sure, I might very well end up going with Mailchimp or Klaviyo in the end after my AI-fueled research. But one of two things will happen as a result: I either go to one of the big vendors with less trepidation and uncertainty, or I find something off the beaten path because I’ve articulated a really specific problem that needs solving, and an AI happens to think this smaller vendor has the answer.

So if you’re running a small business, it might help to see the opportunity wrapped in the threat that is AI search. Instead of trying to outrank the big players, you need to become the answer to a really specific problem in the market.

But how do you actually do that?

Google uses many systems to decide what to show users, and we cannot observe a simple “E-E-A-T score” that determines whether an AI answer cites you. I still like E-E-A-T as a quality checklist because it tells you to show real experience, support claims, make authorship clear, and keep facts current. But saying “I like these criteria because I think they make good content” is different from pretending I know the exact citation algorithm.

There are a lot of ways you can prepare your small business to take advantage of the changes AI is making to the way people search. Here are six ideas that come to mind:

Start with the data you can observe. Google Search Console now has dedicated generative-AI performance reports showing impressions from AI Overviews, AI Mode, and other generative Search experiences. Pair that with referral traffic from ChatGPT, Perplexity, Claude, and other tools. But even as you do this, remember that referral data captures only clicks, not all of the influence that happened before the click.

Don’t bury the answer to a simple question in paragraph four. Put a concise summary right at the top. Use clear headings, logical sections, and schema markup if relevant. Think more like a reference doc, less like a long narrative.

Do not assume AI visibility is independent of ordinary search visibility. Google says AI Overviews and AI Mode still rely on its core Search systems, even though the responses and links can vary. So I would optimize for selection in addition to rank: clear entities, accurate facts, strong sourcing, real experience, useful internal links, and content that answers the question better than the commodity alternatives.

AI might grab a snippet from your product page or FAQ, not just your blog. So every part of your site should be clear, accurate, and structured. If a support article has the best answer, it can become the citation.

AI systems can synthesize, retrieve, compare, and reason across existing information. What they cannot manufacture is your first-party evidence: your original data, real case studies, firsthand experience, proprietary processes, and specific facts about your business. Publish information that adds something to the corpus instead of merely restating it.

Fast load times, clean code, and mobile responsiveness are still extremely important, as are clear calls to action and good UX. AI might change how people arrive, but it hasn’t changed why they stay.

And if you read the above and you’ve noticed that this isn’t too far off from traditional white-hat SEO, well, that’s kind of the beauty of it. After all, search engines want to send people to useful sites and AIs want to draw from useful sites. Their missions are not that different.

But that’s not to say that AI won’t further change SEO rules over time.

If you feel nervous, I sure don’t blame you. This big change doesn’t show any sign of stopping.

Google is continuing to roll out AI overviews more broadly. Microsoft is going to keep working AI into Bing. ChatGPT, Claude, and Perplexity are going to keep tending to their products and growing market share and loyal user bases of their own. And perhaps some competitors will pop up as well, even if one or two of the OGs go defunct.

But make no mistake: people will still search for information one way or another.

Of course, your metrics will need to adapt. You won’t always be able to count clicks, since a lot of AI traffic will show up as “direct/none” in Google Analytics. Some of that traffic will disappear entirely due to zero-click searches, and you’ll probably want a way to track that too, be it Google Search Console or maybe SEMRush’s new AI visibility tools.

Over time, visibility and citations will increase in importance, so you will want to find ways to track that. You’ll want to know if your brand is showing up in summaries, getting mentioned, quoted, and linked.

And it’s important to not forget that other changes are happening simultaneously. Voice and visual search are continuing to get better too with tools like Google Lens and Siri. They’re not disruptive enough to make headlines these days, but their continued adoption is going to keep changing the way people search, little bit by little bit.

Finally, make no mistake about it: AI companies have been looking for better ways to monetize. And monetization is no longer hypothetical. Google has been working ads into AI-heavy search experiences, and OpenAI began testing ads in ChatGPT in 2026. That means paid placement is likely to become another layer of an already fragmented search journey. It’s not a replacement for organic visibility, but it is another thing businesses may eventually have to measure alongside it.

The future will probably hold a mix of old-school organic search, sponsored search, conversational AI research, AI overviews on top of traditional search, paid AI results, voice search, and visual search. And yes, this kind of fragmented search is a lot harder to track than Google traffic alone would have been in the halcyon days of 2018!

But even still, two golden rules will remain:

If you do these two things and you move forward in a spirit of bold experimentation, I think you’ll come out of the AI SEO upheaval stronger than you went into it.

SEO isn’t dead. And it won’t die unless the concept of searching for answers itself does.

But you’d be making a big mistake if you didn’t bank on big changes. AI is already dramatically changing the way people search for answers, whether they use new tools like ChatGPT, or stick to established ones like Google.

And look, I won’t lie to you—if you’re putting all your eggs in the “organic search traffic” basket, you should reconsider. Because there’s a tremendous amount of opportunity out there if you’re willing to seize it.

If you are fast, clear, genuinely helpful, and your business is positioned to solve real problems had by real people, then you are well-positioned to stand out. Not just in the search engines, but in the AI systems that are reshaping how things are found on the internet.

Remember: the AIs have to get their answers from somewhere. And that could be you. So keep publishing content that helps people, and make sure you format it so machines can understand it. The machines, in turn, will do their best to help the people you’re trying to help.

As ever, the best defense a business can have is to make something worthwhile. The other half is showing up where your customers are already looking.

If you want a more measurement-heavy follow-up, read Do You Need a Rank Tracker? I Read 200+ SEO Threads To See. It explains why both classic rankings and AI-visibility trackers are useful samples rather than perfect ground truth.