Appearing on page one of Google is no longer the coveted prize it once was, forcing a fundamental rethink of how businesses market themselves online. The rise of artificial intelligence (AI) tools is reshaping how people search. Instead of typing a question into a search engine and combing the results, now they can ask a large language model (LLM) such as ChatGPT or Claude, which delivers a direct answer at speed.
It’s less than three years since ChatGPT’s public launch introduced people to the concept of LLMs, but its maker OpenAI says the tool is already used by 10% of the world’s population. With rival LLMs produced by the likes of Google and Meta also popular, it’s no surprise that AI tools now account for 5.6% of desktop search traffic in the US – more than double their share a year earlier. In the UK, AI search drives less than 1% of traffic, a fourfold increase in the past year. While small, this has already reshaped traditional search. Google now augments its results with AI Overviews occupying up to half the screen on some devices, pushing links down the page.
The consequences are already noticeable. Publishers have seen their search visibility fall by as much as 80% since 2019, with a related decline in readership and advertising revenues. E-commerce brands are beginning to feel the same squeeze, because when fewer people reach their websites, fewer buy from them.
For businesses that have invested in search engine optimisation (SEO) to attract customers, or who try to raise their profile by using PR companies to get themselves published in media outlets, this is a wake-up call. Producing content and hoping Google finds it is no longer enough. Staying relevant means completely rethinking how you present your business in an era where AI controls online discovery.
The path to zero-click answers
SEO works by setting out your website pages and the underlying computer code to closely match what search engines are looking for. AI tools, however, do not necessarily drive traffic to business websites. ChatGPT and Claude offer a ‘zero-click’ experience, either searching the web and picking the best result or pulling an answer from the vast amounts of information stored in their memory.
For marketers, the strategic implications are profound. If AI searches become the norm, then visibility is no longer about position but about probability: the likelihood that an AI model references your content when generating an answer.
“Everything becomes a statistical likelihood, not a guarantee,” says Tim Burley, of digital agency Organic. “Whereas rankings in the old Google model were at least predictable, LLMs are designed to make it up as they go along.”
This new way of searching is known as Generative Engine Optimisation, or GEO, and it could completely change the game for everyone.
Solving the ‘messy middle’
Burley has followed this new search trend closely for his clients. He sees search engines evolving into “answer engines” that summarise information and keep users within their ecosystem.
Before this point, the basic SEO bargain had held for around two decades: Google, which has around 90% of the search market, and its rivals used automated tools to ‘crawl’ websites and add them to their index. In exchange, they sent users to relevant websites. That deal is now breaking down.
“Google doesn’t want to be a search engine anymore,” Burley observed. “It wants to solve the ‘messy middle’ – the part of the customer journey where people bounce around comparing options. And it wants to keep those people within its own ecosystem.”
Industry analysts Gartner predict that organic search traffic will decline by up to 25% by 2026 as AI systems resolve more user queries directly, without sending to them to a relevant website.
Adapting to LLM search
First described in academic research in late 2023, GEO focuses on making content more discoverable and usable by generative models. Answer Engine Optimisation (AEO), its close cousin, prioritises concise, intent-driven content.
The economic stakes are high. Generative AI investment reached $33.9 billion globally in 2024, according to Stanford’s AI Index, and the AI market overall is forecast to hit $1.8 trillion by 2030. Even if this investment is a sign of a bubble, as critics argue, for Burley, “a seismic shift” would be required to halt the momentum of AI-powered search.
LLMs complicate search optimisation because they combine traditional web search with other sources, such as their training data, to generate the best answer. Publishers cannot influence their access to many of these sources; they were either part of the training data or they weren’t.
They can, however, adapt. Burley says: “Your content needs to be structured so it’s easy for LLMs to cite – often they’re answering the question first and looking for evidence after.” That could mean structuring more of your website as frequently asked questions (FAQ), for example, or posing a question, then giving the straightforward answer initially, followed by supporting information.
Five foundations for a GEO strategy
If the tactical landscape is shifting, businesses need to understand what a strategic response looks like. Burley outlined five principles that he believes remain robust, even as algorithms evolve:
- Map your entire ‘search experience’ landscape.
“Every business should know where their traffic comes from, whether that’s Google, LinkedIn, Reddit, or niche communities,” Burley explained. “Understand those touchpoints and assess where you’re strong and where you have gaps.” This is partly organic, based on where your customers are, and partly the result of where you choose to publish. A company that regularly interacts with customers on the product’s Reddit page, for example, will almost certainly attract more traffic from there. Analysing your traffic sources will help you understand where to focus. - Build brand authority signals.
SEO once rewarded gaming the system; LLMs reward trust. Brands should ensure consistent messaging across platforms, reinforcing a coherent identity. For example, if the tone of your communication is typically knowledgeable and authoritative, then it’s better to maintain this tone for every channel, rather than adding a jokey social media account into the mix. “LLMs love to put you in a box,” Burley said. “The clearer your story, the easier it is for them to know what you stand for.” - Answer questions directly.
Long narrative intros may need a rethink. If you’re explaining, for example, why your cast-iron pans are better than stainless steel ones, then answer that question at the beginning of the article. Don’t spend the first four paragraphs on a scene-setting story about your grandma’s cooking. “Cut to the chase,” Burley advised. “Be specific and purposeful. If answer engines want answers, make sure your content delivers them.” - Invest in structured data.
The technical backend elements of how content is actually published will be more important than ever. For example, coding your product pages so AI and search engines can easily understand what the item is, how much it costs, and whether it’s in stock.. “If everything is free-flowing text, you’re making it harder,” he noted. “Structured data makes it easier for LLMs to connect the dots.” That means thinking about descriptive page titles, well-organised navigation, sub-headings that explain what content is coming up and so on. - Adopt an experimental mindset.
With rules in flux, agility beats certainty. “Everything we’re doing is a test,” said Burley. “Validated today, but it may not be tomorrow. So, keep trying, keep tracking and learn fast.”
These principles do more than protect company visibility online; they align with user expectations. Faster, clearer answers help not only machines, but also the humans behind the queries.
What this means for content marketers
For content creators, the fundamentals of storytelling endure, but delivery needs to adapt. Brands can no longer rely solely on populating their owned channels, such as their mobile app or blog. “We can’t just build it and hope they come,” Burley warned. “We must go where our audiences are – LinkedIn, Reddit, Discord – and provide the right format for each space. Then join the dots across all those stories.”
Measurement will also become more complex. Traditional SEO offered tangible metrics, such as page rankings, click-through rates and referral traffic, that clients could see on a dashboard. Those signals are fading because AI platforms don’t provide explanations of where they got their information or how. Indeed, often the people who built them cannot explain how these tools arrive at their answers.
“Attribution is broken,” Burley admitted. “The web is messier than ever, and marketing data is at an all-time low for confidence. But the goal hasn’t changed: connect meaningfully with people. That means stronger brands, clearer stories and smarter data use.”
Preparing for an answer-first future
AI-powered answer engines are redefining discovery. Google’s AI Overviews and tools such as ChatGPT may still be imperfect – like the one that claimed that tripe’s kosher status “depends on the religion of the cow” – but their influence is growing fast. Businesses cannot afford to be complacent.
Structured content, coherent brand narratives and agile experimentation will be critical. So will a mindset shift: away from chasing rankings and towards building trust, clarity and authority.
As Burley put it: “SEO has always changed. The difference now is the pace and the fact that the old certainties are gone. The brands that adapt early will have the advantage.” For those willing to act now, the opportunity is as significant as the challenge.
Leave a comment