There is a version of this conversation that focuses on tactics. Which schema markup to use. How to structure FAQ content for AI Overviews. Whether your robots.txt is blocking Googlebot-Extended. Those things matter - but they are not the conversation most UK brands need to be having right now.
The more urgent question is structural: can Google or an AI engine do what your website does, without your website? If the answer is yes, or even maybe, then you have a business model problem. Not a visibility problem. Not a content problem. A model problem. And that sits above anything a marketing team can fix on its own.
Tactical Excellence Has a Ceiling Now
Rand Fishkin's analysis at SparkToro makes a point that deserves to be said plainly: being brilliant at SEO, social content, or short-form video no longer guarantees survival if the underlying business is one that AI can absorb. The best keyword researcher in the room cannot outrank a Google AI Overview that answers the query directly. The most disciplined content operation cannot win traffic that never leaves the search results page.
This is not a counsel of despair. It is a prompt to think at a higher level. The brands that will sustain organic and AI-driven visibility are those whose value cannot be fully replicated by a language model reading their content. That distinction is worth sitting with. What do you offer that requires a human to choose you specifically, rather than accept a synthesised answer?
The Disintermediation Test Every Brand Should Run
Put your core product or service into ChatGPT, Perplexity, and a Google AI Overview. Ask the question your customer would ask. If the AI resolves the query completely - gives a recommendation, an answer, a price range, a how-to - without naming you or sending anyone to your site, you are already being disintermediated. This is not hypothetical. It is happening now, across categories from financial comparison to travel to HR software.
The test is uncomfortable because the results are often worse than brands expect. Informational content that took months to produce gets summarised in two sentences. Product category pages get replaced by AI-generated comparison tables. Directories that built their entire model on aggregating information find that the aggregator is now the AI itself. Recognising this early gives you options. Recognising it late gives you a crisis.
What Actually Survives AI Disintermediation
The SparkToro analysis points to strategic features that predict survival - not tactical ones. Proprietary data is one. If your content or product is built on information that only you hold - original research, first-party customer data, a dataset no AI can scrape - you have something to protect. AI engines can cite you, but they cannot replace you. That is a meaningful distinction.
Community is another. Platforms and brands that have built genuine, active communities around their product or content hold something an AI cannot synthesise: the social proof and peer interaction that comes from real people in dialogue. Forums, membership communities, and networks where the value is the people - not the content alone - are structurally harder to disintermediate. For UK brands, this might mean reconsidering how much effort goes into publishing versus convening.
Transactional necessity is a third. If customers must visit your site to complete an action - book, purchase, configure, sign - you retain a floor of traffic regardless of how much the discovery phase shifts to AI. The challenge is that discovery is precisely where AI is eating first. Brands that depend on long organic funnels starting with informational queries are most exposed. Brands where the AI can only recommend and the user must still act on your platform are in a better position.
What This Means for AI Visibility Strategy
GEO and AEO work - optimising for citations in Google AI Overviews, ChatGPT, and Perplexity - is still worth doing. If AI is going to answer your customer's question with or without you, you would rather be cited than ignored. Being named as the source, the expert, the recommended brand matters for what happens next: the follow-up search, the direct navigation, the brand recall that eventually converts.
But AI visibility optimisation layered on top of a disintermediable business model is a short-term play. You are optimising for citations in answers that replace your traffic. The smarter sequence is to audit the business model first, identify where you are genuinely differentiated, then build your GEO strategy around those differentiators. Structure your content around the things only you can say - proprietary insight, specific expertise, original data - rather than trying to out-optimise commodity content.
Paid Search Becomes More Important, Not Less
One immediate consequence of zero-click growth is that paid channels gain relative importance for any brand whose organic model is under pressure. If the top of the search results page is increasingly owned by AI Overviews and the middle by AI-generated content summaries, paid placements - particularly those that appear within or adjacent to AI results - become a more reliable route to maintaining visibility during a transition period.
Performance Max, Demand Gen, and the emerging AI MAX format are all relevant here. None of them solve the structural problem - a disintermediable business is still disintermediable regardless of how well your PMax campaigns perform - but they provide a floor. They keep traffic moving while the harder strategic work happens. For UK brands facing declining organic click volume, treating paid search as a bridge rather than a long-term substitute is the practical framing.
The Strategic Audit Most Brands Haven't Done
Most marketing teams are still optimising within their existing model rather than questioning the model itself. That is understandable - it is what the job description asks for. But the zero-click shift is moving fast enough that brands which wait for a traffic collapse to prompt strategic review are likely to find themselves with fewer options when that moment arrives.
The audit does not need to be elaborate. Map your top traffic-driving queries. Run them through AI engines. Categorise the results: queries where AI resolves the need completely, queries where AI cites sources including you, and queries where AI cannot fully answer without directing users to a specific provider. That third category is your moat. Build from there.