Stephen Downward.
Stephen Downward, Head of Digital Experience, Atomic 212°
Every few years, the industry declares SEO is dead. First, it was because of social media, then mobile, then voice search. Now generative AI is the latest reason. Except SEO is still here. What’s changing is the way in which people discover brands, and how AI learns about them
AI doesn’t break from search so much as complete a shift Google started years ago. Links, brand mentions, reviews and editorial authority have shaped results for as long as most marketers have worked in the channel. LLMs take that same principle to its conclusion. They don’t retrieve the best-optimised page but instead connect signals from across the web: editorial coverage, expert opinions, comparison sites, industry bodies, review platforms, research, communities and brand-owned content.
The difference lies in what LLMs do with these signals. Google uses them to rank a list you have to click through. AI uses them to write the answer. No single page wins anymore; instead the most trusted data source does.
If you take a prompt as simple as “What’s the best family SUV?”, neither ChatGPT nor Google’s AI Mode is likely to recommend a vehicle using a manufacturer’s website alone. The answer is synthesised from trusted auto publications, ANCAP safety ratings, comparison sites, owner reviews, editorial testing and even expert commentary, long before the brand’s own content enters the picture. It’s this ecosystem that builds the recommendation. And every industry has one: in financial services, it’s comparison sites, analysts and regulators. In travel, it’s publishers, booking platforms and review communities. In retail, it’s product reviewers, creators and marketplaces. These players have always shaped what consumers believe, and now they’re shaping what AI believes too.
Long-valued SEO playbooks are starting to lose their edge. The web is filled with articles that answer the same questions in almost identical words. Google’s helpful content systems – and its long emphasis on experience, expertise, authoritativeness and trust – exist precisely to reward information that’s “genuinely useful” rather than content built to rank.
At the same time, click-through rates are falling as AI answers resolve more searches without a visit to a website. Visibility is becoming less about winning the click and more about being the source an AI trusts enough to use for its results: editorial coverage, original research, expert commentary and independent mentions. These sources build the knowledge graph AI draws on to provide recommendations. As a brand, it’s no longer about publishing the most, but being the one earning the most authority.
A good place to start is to map who shapes opinion in your category. Not just competitors, but the publishers, experts, creators, analysts and review platforms that consistently move both consumer decisions and AI recommendations.
Then, change what you measure. You need to know where AI platforms cite you, which publishers dominate the recommendations, where competitors are earning authority, and where the gaps are.
After that, invest where trust is created. Technical SEO remains essential and a strong website still matters, but both are becoming the foundation rather than the advantage. The edge now comes from original insight, credible expertise, editorial recognition and independent validation.
Search isn’t dead, nor is it becoming less important. It’s more connected than ever. To show up at the right time, and in the right place, brands need to focus on building stronger ecosystems of authority around their business and the products or services they offer. Visibility is earned one trusted mention at a time.