As AI-driven search becomes a default starting point for each of us, a new visibility divide is opening up. Traditional SEO and content optimisation are no longer enough to guarantee visibility. AI systems increasingly generate direct answers, and the sources an AI retrieves and trusts determine which brands are surfaced at the moment intent is expressed. Many organisations are now discovering that despite having a digital presence, they are not always visible to AI.
A different kind of animal – AI search
AI assistants such as ChatGPT, Gemini and Perplexity do not surface results in the way search engines do. Instead, they compile answers from sources they trust. These sources, referred to as AI citations, play a critical role in determining which businesses are referenced in AI search results, recommendations and explanations.
New research analysing 6.8 million AI citations shows how concentrated those sources really are. The findings reveal that 86 per cent of citations come from places brands already control, including websites, listings and reviews. Public debate often frames platforms such as Reddit as dominant, yet the data shows that forums contributed roughly only 2 per cent of citations once location and query intent were factored in.
The research also found that when analysing unbranded objective queries, such as searching for the closest restaurant, concert or a nearby service, first-party websites and local pages accounted for nearly 60 per cent of citations. Branded or subjective queries leaned more heavily on structured listings and review platforms.
The implication is clear – visibility is shaped by structured information that AI systems can interpret and trust. Brand presence needs to be consistently strong across all digital environments where consumers might search in local, specific or task-driven ways.
How AI search varies across each industry
This landscape can vary per industry, as different sectors rely on different types of machine-readable data. The research found that retail citations were driven mostly by first-party websites (47.6%), while financial services leaned even more on authoritative, brand-owned domains (48.2%), reflecting the sector’s need for verified information.
Meanwhile, healthcare diverged, with industry directories such as WebMD dominating at 52.6%, and food service showed a stronger pull toward reputation signals like reviews and social content (13.3%), though listings still led at 41.6%.
There is clearly a consistent pattern across all sectors: AI systems prioritise structured, trusted sources, and the greatest influence continues to come from brand-managed ecosystems.
No time to cry when the data goes wrong
Most organisations lose visibility in AI search not because their content is weak, but because their data is inconsistent or difficult for AI systems to interpret. AI desires structured data and understands linked-entity relationships, which is why using models such as a Knowledge Graph is critical.
Search queries are increasingly long-tail and conversational. Brands need entity-based knowledge structures that can not only answer the initial question but also continue to support additional requests with accurate responses in the same session.
If a brand’s information is incomplete, inconsistent or poorly structured across key sources, AI models will cite other entities instead, and it might seem that some brands do not exist at all.
AI citations as a visibility metric – how brands can respond
Understanding AI citations provides a measurable way to assess how often brands appear in AI-generated answers. It reveals which sources AI models prefer, where brands are consistently referenced and where they are missing entirely.
This matters because more than half of consumers already use AI assistants weekly, according to recent survey data. The same survey found that 72 per cent of marketing leaders believe AI search will have a greater impact on customer acquisition than traditional SEO within three years.
To help improve AI visibility in a structured way, I would suggest that brands follow these three actions:
-
Measure citation-based visibility across all major sources
Brands should assess where they are frequently cited and where they are not referenced at all. This should be measured at the level where consumers actually search, which is location and service level, in order to provide a clear picture of visibility gaps. -
Structure and centralise factual data
Information from locations, services, products, faqs, help articles, menu items & allergens through to professionals needs to be machine-readable and consistent across all digital properties. Schema markup aligned to each relevant entity type is critical to this. Consider breaking content out into dedicated, entity-based content to allow AI systems to easily understand and relate your brand content. -
Distribute accurate information to the sources AI relies on
Brands should treat data as an operational asset by ensuring that websites, listings, social and review platforms contain aligned and complete information, as this all aids in increasing citation likelihood. AI prioritises precision, clarity, and consistency as signals of trustworthy, answerable content.
A future defined by data quality
AI search adoption is accelerating, and the organisations that understand how citations work will set the pace of digital visibility. Future visibility will depend less on who publishes the most content and more on who provides the strongest structured data.
Brands need to prepare their ecosystems now, ensuring they appear in the places AI looks for answers, and will gain an early advantage. Those who delay will continue to lose ground despite stable customer demand.
