Zero-Click and AI Cannibalization: The Traffic Paradox

Why the traffic paradox matters for DACH B2B now

Marketing and SEO teams in the DACH region are currently experiencing two opposing movements. Their brand's presence in AI answers from ChatGPT, Perplexity and Google AI Overviews is growing. At the same time, the number of clicks reaching the website from organic search is falling. This is not a measurement error. It is a structural effect: AI search systems deliver the answer directly in the interface and thereby replace the click on the source.

The scale is considerable. In the first four months of 2026, 68.01 percent of all Google searches ended without a click, compared with 60.45 percent in 2024 (data from SparkToro and Similarweb, US only). In February 2024, Gartner even predicted that traditional search engine volume would drop by 25 percent by 2026 because generative AI would become a substitute answer engine.

For B2B companies in Austria and the rest of the DACH region, this means: if you keep measuring visibility only in sessions and clicks, you see a falling curve and conclude that you have failed. In reality, the value is shifting from the click to the mention and to the conversion quality of the remaining traffic. In 2026, resolving the paradox is less a technical task than a matter of measurement and argument.

The zero-click mechanics in 2026: how AI answers replace the click

The core of the effect is well documented. A Pew Research Center study based on 68,879 Google searches by 900 US adults in March 2025 shows the mechanism clearly: users who saw an AI summary clicked on a traditional search result in only 8 percent of cases. Without an AI summary, the figure was 15 percent, almost twice as high. Even more striking: users clicked a link inside the AI summary itself in only 1 percent of visits.

This halving of clicks hits harder the more search queries trigger an AI answer in the first place. AI Overviews coverage, however, fluctuates strongly. A Semrush analysis of more than 10 million keywords shows that AI Overviews rose from 6.49 percent of keywords in January 2025 to a peak of 24.61 percent in July and then fell back to 15.69 percent in November 2025. The visibility surface is therefore not a fixed value but a moving target that depends on how Google decides to show AI Overviews.

Important for context: the figures above come from the US. They show the direction and the mechanics, but they cannot be transferred one to one to the DACH market. The AI traffic share there is still much smaller (see the section on the DACH reality).

From cannibalization to conversion: why fewer clicks do not mean less value

The decisive point of the paradox: the remaining AI traffic is often higher quality. Someone who reads a concrete answer in ChatGPT or Perplexity and then clicks anyway arrives with stronger purchase intent. The AI has already pre-qualified them.

The data supports this. A Semrush study concludes that an AI search visitor is 4.4 times as valuable as an average visitor from classic organic search, measured by conversion rate. The same study projects that, for marketing and SEO topics, AI search could send more visitors to websites than classic search starting in early 2028.

Behavior on the site differs as well. An Adobe analysis of US retail shows that visitors from generative AI sources have 8 percent more engagement, 12 percent more pages per visit and a 23 percent lower bounce rate than visitors from non-AI sources. Over the same period, referral traffic from generative AI sources to US retail sites rose by 1,200 percent (February 2025 compared with July 2024). What emerges is a small channel that grows fast and brings better visitors.

For B2B, this means the right metric is no longer the raw click count but the value per session and the AI channel's contribution to the pipeline. Falling clicks alongside rising conversion are not a contradiction. They are two sides of the same paradox.

Strategy: optimizing for citations instead of clicks

When clicks become rarer, citations and mentions move to the center. The study GEO: Generative Engine Optimization by Aggarwal et al. (Princeton University and others, KDD 2024) tested nine text methods meant to make a source more visible in AI answers. Three content additions performed best: quotations, statistics and source citations.

• Add statistics: Concrete numbers instead of qualitative statements make a source much more visible in the AI answer.
• Quotations and source citations: Direct quotations and clear source references make a paragraph easier for an AI system to use and more trustworthy.

Additional practices from the field, not part of the Princeton study:

• Front-loading and answer islands: The answer belongs at the beginning of the paragraph. A self-contained answer block that makes sense without surrounding context (an answer island) is easier for an AI system to extract than a statement spread across several paragraphs.
• Clear structure: Precise subheadings, short sentences and a question-and-answer logic help the model find the relevant section and reproduce it correctly.

These tactics aim to get your own statement into the AI answer, ideally with your brand named. Even without a click, this builds visibility and trust with your target audience.

Brand mentions instead of backlinks: what really correlates with AI visibility

One of the most important shifts concerns the off-page signal. In classic SEO, the backlink was the central currency. For AI visibility, the weight shifts to the plain brand mention.

An Ahrefs study of 75,000 brands (Spearman correlation) shows that brand mentions on the web correlate with AI visibility at 0.664 for ChatGPT, 0.709 for AI Mode and 0.656 for AI Overviews. The strongest correlation comes from YouTube mentions, at about 0.737. According to the study, classic link metrics such as the number of backlinks show only very weak correlations. Ahrefs sums it up this way: what matters most is being mentioned across a broad range of sites.

In practice, for DACH B2B this means that presence in trade media, industry portals, on YouTube and in communities such as Reddit feeds directly into AI visibility, even when these mentions contain no link. Digital PR and thought leadership thus become GEO measures. The link graph remains relevant for classic SEO, but it is no longer a reliable proxy for how AI systems perceive you.

The technical GEO foundation: llms.txt, schema and crawler control

For AI systems to capture and correctly interpret your content at all, you need a technical foundation. Three building blocks are central.

• AI crawler control in robots.txt: AI crawlers such as GPTBot (OpenAI), ClaudeBot (Anthropic) and PerplexityBot can be explicitly allowed or blocked in robots.txt. This control is a deliberate decision: if you want to appear in AI answers, you have to allow the relevant bots. Adoption is increasing. According to HTTP Archive, robots.txt directives for GPTBot rose from 2.6 percent (2024) to around 4.5 percent (2025).
• Structured data and schema: Schema markup helps AI systems capture entities, facts and relationships unambiguously. Clean structured data increases the chance that statements are reproduced correctly and with reference to your brand.
• llms.txt: This proposed file is meant to give AI systems a curated overview of a website's most important content. Adoption is still low: according to HTTP Archive, only 2.13 percent of desktop pages have a valid llms.txt (2.1 percent on mobile). The standard is not yet established, but the effort is small, so early adoption can make sense as a low-risk experiment.

The technical foundation makes content accessible and machine-readable. It does not replace the content levers from the previous section, though. The two work together.

Common mistakes when assessing the traffic paradox

• Treating falling clicks as a failure across the board: If you only measure sessions, you miss the value of increased AI visibility and better conversion. The channel is shifting, not disappearing.
• Not separating AI-induced loss from seasonality: A drop in clicks can be caused by seasonality, a Google update or AI Overviews. Without a clean year-over-year comparison and without matching against the AIO coverage of your own keywords, you cannot attribute the causes.
• Applying US figures to the DACH market: Values such as the 8 percent CTR or the 68 percent zero-click rate come from the US. The AI traffic share in the DACH region is much smaller. Adopting them uncritically leads to the wrong priorities.
• Betting on backlinks instead of mentions: If you put your entire off-page budget into link building, you ignore the fact that brand mentions correlate more strongly with AI visibility.
• Not tracking brand mentions in AI answers: If you only measure clicks and ignore share of voice in AI answers, you have no indicator of actual GEO success.

The DACH reality: how large AI traffic really is

A sober view is needed here. In the DACH region, direct AI referral traffic is still very small. An SE Ranking analysis of 63,987 websites from Germany, Austria and Switzerland (January to April 2025) puts the AI traffic share at only 0.13 to 0.15 percent. Among the engines, ChatGPT clearly dominates with 76.48 percent of AI traffic, followed by Perplexity with 20.54 percent.

The user base, by contrast, is growing fast. According to Bitkom, 67 percent of people in Germany aged 16 and over use generative AI at least occasionally (2025), compared with 40 percent in the summer of 2024. Among the tools, ChatGPT leads with 43 percent, followed by Microsoft Copilot with 39 percent, Gemini with 28 percent and Perplexity with 1 percent. The gap between high usage and still low measurable referral traffic is the real finding: the users are already there, while the attributable stream of clicks lags behind.

The conversion side supports the investment. A Conductor analysis reports a conversion rate of 14.2 percent for AI traffic, compared with 2.8 percent for Google traffic. Important: this conversion figure is not limited geographically to the DACH region, while the traffic shares (0.15 percent, ChatGPT, Perplexity) come explicitly from the DACH analysis. For Austrian B2B providers, the conclusion is: small in volume, high in intent, with a clear growth trajectory.

Agentic commerce: the next wave

The next stage goes beyond the answer to action. AI agents such as Perplexity Comet do more than research. They act as a purchasing interface: they compare, select and initiate transactions. This turns the AI itself into the decision node between vendor and customer.

For B2B visibility, this shifts the requirements once again. When an agent pre-qualifies products and vendors, product data, specifications and terms must be available in machine-readable, complete and structured form. Whatever the agent cannot capture cleanly does not make the shortlist. Structured data and a consistent entity profile thus become a sales factor, not just a visibility factor. This trend is still young in 2026, but the direction is clear: if agents cannot see you, you drop out of the preselection.

Metrics and measurement: making the traffic paradox visible

To manage the paradox, you need metrics beyond the click count.

• Share of voice in AI answers: How often does your brand appear in the answers of ChatGPT, Perplexity, Google AI Mode and AI Overviews for relevant prompts, relative to the competition? This is the central GEO visibility indicator.
• Citation and mention tracking: Record whether the brand is cited (with a source) or merely mentioned. Also watch the mention-source divide: which sources (trade media, YouTube, communities) feed the mentions that show up in AI answers?
• AI referral attribution in GA4: Segment traffic from AI sources and measure its conversion rate and value per session separately from classic organic traffic. This lets you prove the AI channel's higher conversion value.
• Isolate click loss cleanly: Compare CTR and clicks year over year and match the decline against the AIO coverage of your own keywords. Only then can you separate AI-induced loss from seasonality and algorithm updates.

The ROI argument for senior management follows directly: clicks can fall while visibility (share of voice) and value per session rise. If you measure both sides, you can prove the AI channel's contribution to the pipeline instead of just presenting a falling click curve.

Action framework for DACH B2B

The zero-click paradox is not a reason to worry but a call to adapt. Three priorities follow from the data.

First, content: prepare it so that AI systems can extract and cite answers from it. According to the Princeton study, concrete numbers, quotations and source citations are the most effective levers. In practice, front-loading and self-contained answer blocks help as well. Second, the off-page strategy: build brand mentions across a broad range of trade media, YouTube and communities, because they correlate more strongly with AI visibility than backlinks. Third, measurement: track share of voice in AI answers and value per session, report AI traffic separately in GA4 and check click loss against AIO coverage.

For the DACH region, a realistic expectation applies: AI traffic is small today, but it is growing fast and converts above average. If you lay the technical foundation now and prepare your content for answer extraction, you build visibility in a channel that will foreseeably gain weight. You resolve the paradox not by winning back the clicks but by measuring success correctly.