Until recently, client questions about SEO used to be fairly predictable. Why aren’t we on page one yet? Can we rank for this keyword? How do we get more website traffic?
Lately though, you might have noticed those questions changing.
Maybe it’s a client wanting to know how to get their business showing up in ChatGPT. Or forwarding you an “SEO is officially dead” article with the message: “should we be worried??” Or (and this one’s my personal favourite), a business owner wondering out loud whether they even need a website anymore, since everyone’s just getting their answers from AI.
If any of those sound familiar, you’ll probably also recognise the feeling that comes with them. Nodding along, buying yourself a second, quietly wondering whether you know enough about any of this to answer with a straight face.
You’re in good company. Over the past 10 years in the SEO industry I’ve watched clients panic about algorithm updates, mobile-first indexing, the death of keywords, the death of backlinks, and the death of SEO itself at least six separate times.
This is the first shift where the panic and the hype have turned up at the same volume, at the same time. Half the internet is telling your clients to burn their marketing plan down. The other half is telling them the secret to AI visibility is schema markup and a llm.txt file. Neither camp is being straight with them.
So let’s cut through it properly. But first, to understand what AI search is and why it matters, you need to see what it replaced. Because the shift makes a lot more sense, and feels a lot less scary, when you look at search before and search now, side by side.
How search used to work
For most of the last two decades, search followed one simple pattern. Someone typed a short string of keywords into Google. Something like “best accountant Brisbane” or “wedding photographer Sunshine Coast”. Google matched those words against pages across the web and returned a list of blue links. The searcher clicked one, landed on a website, and found their answer there (or didn’t and went back to the list to find another site).
Under that model, the job was clear: get your webpage into that list, as high as possible. Rankings were everything, because rankings drove clicks, and clicks drove customers. Every SEO tactic, every report, every client conversation was built on that assumption.
This is probably the mental model most of your clients are still carrying. So when they say “we want to do SEO,” what they usually picture is that list of blue links and their business sitting at the top of it. That picture is now incomplete.
What AI search actually is
AI search is any search experience where an AI system reads sources from across the web and generates the answer itself, instead of just pointing you at a list of websites. You’ll sometimes hear this technology called generative AI, or the models behind it called LLMs (large language models). Different names, same idea: instead of matching your words to a page, the AI actually reads and writes an answer.
You’ve probably already spotted this in the wild, even if the mechanics are still a bit fuzzy. Google’s AI Overviews, those AI-written summaries at the top of the results page, now appear on somewhere between a quarter and nearly half of all searches depending on which study you trust. Semrush and Conductor track it around 25%, BrightEdge closer to 48%. Then there’s Google’s AI Mode, a fully conversational version of search that passed 100 million monthly users within months of launching. And the standalone assistants: ChatGPT, Perplexity, Claude, Gemini, where people are asking the exact questions they used to type into Google and getting direct answers, often with specific businesses named.
The mechanics matter here, because they explain everything that follows. In the old model, the search engine matched your words to pages and handed you options. In the new model, the AI reads thousands of sources, forms its own synthesised answer, and delivers it in plain language. Sometimes it cites where the information came from. Sometimes it doesn’t.
Which changes your job. It’s no longer just to rank in a list. It’s to be one of the sources the AI trusts enough to draw from when it builds its answer.
Think of it as the difference between being listed in a directory and being recommended by a well-connected friend. The directory shows you as one option among many, and the searcher does the choosing. The friend gives a verdict. And if you’re not a business that friend knows and trusts, you simply don’t come up.
How people actually search now
The technology behind search isn’t the only thing that’s changed. The behaviour changed too, and it’s something you can probably see it in your own habits.
People search in more places than ever. Google is no longer the automatic starting point for every question. Recommendations happen on TikTok, honest opinions get hunted down on Reddit, research and comparison happen in ChatGPT and Perplexity, how-to answers live on YouTube. A single buying decision might touch four or five platforms before anyone contacts a business.
And in a growing number of cases, it’s not even a person making that buying decision anymore. Ask ChatGPT to find you a plumber, compare three project management tools, or book a restaurant for Friday, and the AI is the one weighing up the options and narrowing them down, sometimes acting on your behalf without you looking at a single website yourself. The decision is being made for someone, not just researched by them. If a business isn’t part of what that AI can see and trust, it’s not in the running, whether or not the human ever knew it existed.
What people are typing into search tools has also changed. Short keyword phrases are giving way to full questions asked in natural language, the way you’d ask someone knowledgeable that you trust. Not “small business accountant Brisbane” but “I run a small trades business in Brisbane, what should I look for in an accountant?” AI search actively encourages this, because unlike old-school Google, it can actually handle the nuance.
And increasingly, the search ends where it starts. AI Overviews and AI Mode deliver the answer right there on the results page. In fact, Seer Interactive found organic click-through rates drop by around 60% when an AI Overview appears. More and more, the results page is the destination, not the doorway. Which means being visible inside the answer now matters as much as earning the click, and sometimes more.
What’s changed beneath the surface
Put all of that together and there are three bigger shifts that underpin it all. And these are the ones your clients will actually feel.
You’re optimising a brand’s whole presence, not just a website. AI systems don’t form their picture of a business from its website alone. They read the whole web: reviews, forum threads, news mentions, social content, what other credible sources say. The website still matters enormously, it’s where a business proves what it does and why it’s worth trusting. But it’s shifted from being the whole game to being the foundation of a bigger one. Necessary, but no longer sufficient on its own.
Which, by the way, is the honest answer to that “do we even need a website anymore” question. Yes. More than ever. An AI can’t cite, summarise, or recommend a business that hasn’t given it anything solid to read.
Success looks different on a report. A business can be cited in dozens of AI-generated answers and watch its organic traffic dip at the same time, because the answer was delivered without a click. Under the old measurement model that reads as failure. In reality, that business was surfaced to a relevant audience as a credible source. More and more, the question is shifting from “how much traffic are we getting” to “how visible are we in the moments that matter”. It might be harder to measure, but it’s alot more meaningful.
Trust is assessed at the brand level. Old-school SEO judged credibility mostly at the website level, links pointing in, authority of the linking sites. AI systems judge it much more broadly. Do third-party sources mention this business? Do the reviews back up the claims? Does the content sound like it was written by someone with real experience? This is Google’s E-E-A-T thinking (experience, expertise, authoritativeness and trust) applied to an entire brand rather than a single page.
What hasn’t changed
This is the part the panic merchants skip, and it’s the part that should let you and your clients breathe out.
The reason behind every search is the same as it’s always been. Someone is trying to solve a problem, make a decision, learn something, or find someone to help. The intent hasn’t changed. Only the way it’s expressed, and where.
The fundamentals that earn visibility haven’t changed either. Helpful content. Clear structure. Genuine authority. The things that have made SEO work for two decades are the exact things AI systems look for when choosing their sources. If anything, they matter more now, because when a system is picking three or four sources out of millions, “pretty decent” doesn’t make the cut.
And the actual goal is untouched. Rankings and traffic were only ever proxies for what a business really wants: customers, revenue, growth. Those proxies are getting less reliable. The goal underneath them is exactly where it’s always been.
AI search hasn’t replaced SEO. It’s expanded it.
Why this matters for small businesses
Small businesses are already living this shift, whether they’ve clocked it yet or not. Their customers are asking ChatGPT for recommendations, reading AI Overviews instead of clicking through, making shortlists before they ever visit a website. A business that’s invisible in those moments isn’t losing rankings. It’s not in the conversation at all.
For the businesses that get their foundations right, though, this is a genuine opportunity. Most of their competitors are either ignoring the shift or panicking about it. The ones with clear websites, real authority, and a credible presence across the web are the ones AI systems keep reaching for. Showing up in the answer builds trust before a customer ever lands on the site.
And if you’re someone small businesses turn to for advice on this, understanding the shift is fast becoming part of the job. How to actually handle those conversations, what to say when a client asks the hard stuff, that’s a whole topic of its own and one I’ll cover properly in another post. For now, just focus on getting your own head around it. Once it stops feeling like a threat to you, it’s a lot easier to explain to someone else.
Search has expanded, not disappeared
If you Zoom out from all of it the — AI Overviews, the assistants, the shift in what gets measured — one thing becomes obvious. Search didn’t get replaced. It got bigger. More places, more question-shaped queries, more answers delivered on the spot, and more weight on whether a business is genuinely credible across the whole web.
Everything you already know about good SEO still holds. It’s the foundation the new layer sits on. Which means the ground under your feet is a lot more solid than the headlines want you to believe.
