Customer Communication When Clicks Disappear in Web4

For more than two decades, digital customer communication was built around a simple, stable logic. Users found information through search, landed on a site, evaluated the offer in the business’s controlled environment, and took action — purchase, sign-up, request.

Traffic was not just a metric. It was the mechanism through which intent turned into revenue.

Today that mechanism is breaking down.

With the spread of LLM search, AI assistants, and agentic interfaces, users increasingly get an answer, recommendation, or decision without visiting the brand’s site. As a result, it is not only the volume of clicks that changes. What changes is where the customer conversation happens — and who controls it.

This is the essence of Web4: an environment where interaction increasingly happens between systems, not through traditional interfaces.

Zero-click is no longer a forecast but a fact, backed by data

The “zero-click future” thesis often sounds like a hypothesis. In reality it is already confirmed by the numbers.

According to SparkToro, SimilarWeb, and Datos, over 65% of Google search queries in 2024 ended without a visit to any website. After the launch of Google AI Overviews, that share began to grow even faster. AI-native platforms such as ChatGPT, Perplexity, or Claude do not operate on click logic at all, providing synthesised answers without sending users to sources.

For the user this means less friction. For business — loss of transparency and attribution.

Demand has not disappeared. What has disappeared is the path by which business used to see how that demand forms, transforms, and converts.

How LLM search changes the logic of intent

Traditional search systems acted as routers: they sent the user to pages. LLM search works differently. It interprets intent and tries to resolve it inside the model before deciding whether external interaction is needed.

That means competition no longer happens at the level of SERP positions. It moves to the level of how the brand is represented in the model’s synthesis.

As Andrej Karpathy noted about AI-native interfaces, “the interface collapses into the model”. In practice, the web becomes latent: it is used but not visited.

Business remains responsible for the outcome — sales, service, retention — while losing control over the initial phase of interaction.

Where business loses control — and why it matters

The most serious problem of a zero-click environment is not falling traffic but loss of control over the context of the conversation.

On owned channels, business shapes the narrative, explains complex points, addresses objections, and sees behavioural signals. In an AI-intermediary environment, that context is compressed. The model decides how to present the offer, which attributes to highlight, what to compare with.

Operationally this shows up as a paradox: demand exists, but conversion worsens. Marketing sees organic decline without an obvious drop in interest. Operations cannot pinpoint where the customer journey “breaks”.

The conversation still happens. But not where business can see or correct it.

Why direct communication channels are strategic again

In this reality, direct channels — above all voice and authenticated AI communication — regain a strategic role.

Voice interaction has properties that LLM search does not: customer identification, two-way context, and the ability to execute an action immediately. Questions are clarified in real time, objections become visible, decisions are made on the spot.

This is not a return to old call centres. It is the realisation that direct communication becomes the only fully observable layer of the customer journey.

McKinsey 2024 research shows that companies with developed direct communication channels retain conversion much better during shifts in digital channels. The reason is simple — they control execution.

AI moves from channel to execution layer

Leading companies are not trying to fight LLM search. They are rebuilding architecture so that any intent — regardless of source — lands in a controlled execution environment.

In this model, AI stops being a channel. It becomes an execution layer: it interprets intent, orchestrates actions across CRM, ERP, billing, logistics, telephony, records the outcome, and triggers the optimisation cycle.

This shifts focus from discovery to execution. In a Web4 world, brand discovery happens outside. Control must stay inside. For more on our approach to execution control in e-commerce and retail operations, see our page.

Measurable risks of inaction

Gartner estimates that by 2026 over 30% of B2C companies will lose significant visibility into customer journey due to AI intermediaries. Without compensatory mechanisms — direct channels and execution-oriented AI systems — business will not be able to analyse conversion decline or justify investment properly.

Relying on third-party AI intermediaries without your own execution layer effectively means handing over control of customer relationships.

Customer communication after the age of clicks

Web4 does not destroy customer communication. It redistributes it.

Clicks disappear. Pages lose their central role. Intent increasingly flows through AI systems. Value lies not in visibility but in the ability to execute.

Companies that adapt will stop optimising traffic and start optimising speed of response, quality of execution, and stability in a fragmented environment. Others will keep refining channels that no longer carry the conversation.

In a Web4 world, customer communication does not disappear. It moves to where action is possible.

And increasingly, that is no longer the website.