Why authority, not clicks, decides who conquers Web 4.0
Opinion
Brands that win won’t be the ones with the nicest websites. They’ll be the ones the AI systems understand most clearly, trust most, and feel most confident recommending, says OMG UK’s chief AI and innovation officer.
For twenty years, digital marketing has been built on a simple assumption: humans search, humans click, humans visit websites.
That’s the logic that shaped SEO, UX, performance marketing, content strategy, and pretty much our entire digital measurement stack. But the internet is rapidly evolving, and we’re seeing the emergence of Web 4.0: an internet increasingly mediated by AI platforms and agents. Instead of people doing the browsing, comparing, and deciding, that work is shifting to AI platforms acting on their behalf.
This shift is already visible in the numbers. ChatGPT now has more than 800m weekly users and is the fifth largest website in the world. Google’s AI Overviews have triggered traffic drops for publishers. And OpenAI’s new purchasing capabilities mean ChatGPT has quickly become the fifth-largest retail referrer in the US.
Zero-click behaviours
The internet is being rewired around three emerging behaviours: zero-click search, zero-click browsing, and zero-click purchasing.
In all three cases, the user asks; the platform does the work. When the platform does the work, it becomes the gatekeeper. And when the platform becomes the gatekeeper, brand authority becomes the only metric that matters to marketers.
Authority is the new battleground
In a human-operated web, you could influence brand visibility with SEO, paid media budgets, and clever UX. In an AI-mediated web, visibility is determined by something far more fundamental: what AI models think of your brand.
For years, we have shaped digital content around the incentives of a few dominant algorithms, such as Google Search, YouTube recommendations, and Amazon product rankings. Those optimisation rules do not translate cleanly into Web 4.0. Content that performs well on traditional platforms is not always treated as authoritative by AI models.
Large Language Models reward clarity, specificity, and question-answer structure rather than polished design, keyword density or high-production assets. This is why highly targeted blogs, long-tail YouTube explainers, niche subreddit threads, and tightly named tutorials appear more often as data sources in AI platforms. Their structure maps far more naturally to how LLMs ingest, parse, and retrieve information.
This shift does not replace everything we have learned over the last twenty years. It requires us to architect for two worlds simultaneously.
The human web still matters
Design, UX, storytelling, brand craft, and classic SEO remain essential. But alongside that, we now need a second layer optimised for AI systems: model-friendly site architecture, text that is easy for crawlers to parse, well-structured product data, and active authority building in the earned spaces that models trust.
These capabilities must complement, not displace, the tactics that worked in a human-operated internet. In Web 4.0, authority is earned by serving both audiences: people and the AI platforms acting on their behalf.
The challenge is that each AI platform interprets the web differently. Profound’s analysis shows ChatGPT leans heavily on Wikipedia, Gemini pulls a large share of its signals from Reddit, YouTube and LinkedIn, and Perplexity is highly influenced by Reddit and YouTube. This balance also shifts over time.
This is why the old idea of optimising for a single algorithm no longer holds. Brands are now trying to influence a fragmented, constantly shifting ecosystem of AI models, each with its own data diet, trust hierarchy, and logic for determining authority.
Why ‘authority’ requires a multidisciplinary approach
Authority in Web 4.0 isn’t something a single team can manufacture. It isn’t a search problem, or a content problem, or a PR problem, or a website problem — it’s all of them at once.
AI platforms form their understanding of a brand by stitching together signals from across the entire digital ecosystem. That means your technical setup, your content architecture, your earned presence, and your product data all blend into a single picture inside the model.
This is why authority can’t be “owned” by one discipline. Engineering choices affect what AI crawlers can parse. Content choices affect whether the model can understand your brand. PR and comms shape the earned signals models that are treated as high-trust.
When you pull these threads together, you start to see authority less as a tactic and more as an organisational capability. It’s the combined effect of how your brand shows up across the open web, in structured data, in trusted communities, and on your own website.
Getting this right requires alignment across disciplines that historically haven’t always had to work together. But in an AI-mediated world, these silos need to come together because AI models collapse their output.
What brands need to do next
Brands now need to architect for two worlds: the human-operated web that still exists, and the AI-mediated web that is rapidly evolving. This requires new capabilities: model-friendly site design, an earned authority strategy, product feed optimisation, and emerging agentic API layers such as the Model Context Protocol.
The brands that win won’t be the ones with the nicest websites. They’ll be the ones the AI systems understand most clearly, trust most, and feel most confident recommending.
Authority is quickly becoming the core currency of Web 4.0. The question every CMO now needs to ask is simple: if an AI platform was choosing on behalf of your customer tomorrow, would it choose you?
Sean Betts is the chief AI and innovation officer at Omnicom Media Group UK