New discussion in the search marketing industry suggests businesses should not judge AI visibility only by referral traffic. An article published by Search Engine Journal argues that generative engine optimisation (GEO) can matter even when visits from AI tools such as ChatGPT appear limited in analytics platforms.

What happened

Search Engine Journal published an analysis saying that referral data may miss a significant part of how AI systems influence purchase decisions. The core argument is that users may discover a brand, product or supplier through an AI assistant, but complete their visit later through another channel such as direct traffic, branded search or a separate browser session.

The article also references research it says shows visitors from ChatGPT converting at a much higher rate than traditional search visitors. That kind of finding, if replicated across sectors, would suggest that AI-driven discovery can have commercial value disproportionate to its raw visit numbers. However, this should be understood as cited research discussed in an industry article, not as a universal benchmark for every business.

Why it matters for European businesses

For European SMEs, e-commerce companies and B2B service providers, this is relevant because many marketing teams still evaluate channel performance mainly through last-click attribution or direct referral reports. That approach can undercount the role of AI tools in early-stage research, vendor shortlisting and product comparison.

In practice, this means a company could conclude that AI search is unimportant because analytics shows little traffic from ChatGPT or similar platforms, while in reality those systems may already be shaping awareness and consideration. This is especially important for firms with longer sales cycles, complex services, higher-value purchases or research-heavy customer journeys.

The development also matters because AI interfaces are changing how users consume information. Instead of clicking through multiple search results, potential buyers may ask for recommendations, comparisons or summaries inside a chatbot. If a business is not visible in the sources these systems rely on, it may lose influence before a website visit even happens.

Who may be affected

• B2B companies with long consideration cycles where prospects research suppliers before making contact.
• E-commerce businesses whose products may be compared or recommended in AI-assisted shopping and discovery journeys.
• Marketing teams and agencies responsible for SEO, content performance and attribution models.
• Founders and SME owners making budget decisions based on channel reports that may not capture AI influence well.
• Digital teams responsible for website content structure, brand visibility and analytics setup.

What companies should consider

• Do not rely only on referral traffic when assessing AI visibility. Review branded search trends, direct traffic patterns, assisted conversions and lead quality as well.
• Monitor where AI may affect the buyer journey, especially for high-intent queries such as product comparisons, category questions, supplier research and problem-solution searches.
• Strengthen source content by publishing clear, structured and credible pages that explain products, services, use cases, pricing factors, specifications and trust signals.
• Review attribution limits in analytics and CRM reporting. Some AI-assisted journeys may not appear as a clean, trackable referral before a sale or enquiry.
• Test GEO alongside SEO rather than treating it as a replacement. Businesses should focus on being understandable, citable and easy to evaluate across search engines, AI assistants and website content.
• Measure commercial outcomes such as qualified leads, conversion rates and sales influence, not just sessions, if experimenting with AI-focused content strategies.

For European businesses, the practical takeaway is not that AI has replaced traditional search, but that reporting models may no longer capture the full path to conversion. Companies that look only at visible referral numbers risk underestimating how AI systems are already shaping discovery and purchase decisions.