The rules for online visibility are undergoing a fundamental shift: during the era of traditional search engine optimization (SEO), the main objective was to attract clicks. In the age of AI-driven search, however, it matters more whether a brand is cited by artificial intelligence systems when they generate answers. This was the message delivered by János Varju, an expert at the INTREN digital agency, at the Digital-Media Hungary conference.
How search has transformed
Previously, users selected from lists of search results and website traffic was a direct indicator of visibility. Today, AI systems frequently provide direct answers on the search interface, meaning users often never visit the original websites. Varju described the shift as the search engine moving from the role of a "librarian" that points to information to a "consultant" that processes content and produces its own responses.
As a result, impressions may rise while click-throughs and traditional site traffic fall. Companies therefore need to redefine their goals: it is no longer sufficient to collect clicks; they must become sources the AI considers trustworthy enough to cite.
Credibility becomes central
Varju emphasized that AI does not simply accept what a brand claims about itself. Systems evaluate signals from across the internet — media coverage, social feedback, user reviews and independent platform ratings — to form an assessment of a brand. If external signals contradict a company's claims — for example, strong marketing claims but poor user reviews — the AI will not treat the brand as credible.
Content quality alone is not enough. AI also checks whether a topic matches real user queries, whether the information is up to date, and whether it contains concrete, measurable data. Varju noted that AI "responds particularly well to numbers and concrete data," so original research, statistics and measurable outcomes significantly raise the chances of a brand being included in AI-generated answers.
How AI decides whom to cite
AI-based search works in stages: systems crawl the web, identify brands, then seek relevant content linked to specific queries — a process often referred to as retrieval augmented generation (RAG). In the final step the AI decides whether a given source should be presented as a citation in the generated response.
Varju also pointed out that AI models operate differently: some rely more on pre-trained knowledge, while others prioritize fresh online content or traditional search indexes. Consequently, brands cannot optimize for a single platform; they need a comprehensive, consistent digital presence.
What companies should focus on
- Building external credibility: press mentions, independent reviews and positive user feedback increase the likelihood of being cited by AI.
- Data-driven content: proprietary research, statistics and measurable results improve trustworthiness.
- Timely, relevant information: AI favors fresh content tied to user searches.
- Multi-channel presence: because AI systems differ, companies should not rely on a single platform.
Varju concluded that in the AI era successful companies will be those that are not only visible online but are genuinely citable — supported by external evidence and measurable expertise — so that AI systems will reference them in generated answers.
