In the era of traditional SEO, "topical authority" was a relatively straightforward concept: publish enough high-quality content on a specific subject, and search engines would eventually view your domain as a primary authority. But as Large Language Models (LLMs) like ChatGPT fundamentally alter the search landscape, the rules of the game are shifting. A central question for content strategists today is whether to double down on core topics or pivot to a broader "funnel" approach to capture wider AI visibility. New research, utilizing the Semrush AI Visibility Toolkit, suggests that the answer is not a simple binary. Instead, it reveals a nuanced hierarchy where "citations" and "mentions" function under different logic, and where the industry you operate in dictates your path to success. The Core Investigation: Mentions vs. Citations To understand if topical authority truly transfers from a brand’s core expertise to adjacent or distant categories, a recent study analyzed six months of U.S. ChatGPT data. The research examined over 1,000 categories, testing whether a brand’s established reputation in one domain carries weight in another. The data reveals a critical distinction: AI models treat mentions and citations as two separate signals. The Citation-First Strategy The findings indicate that brands have a high probability of being cited as a source even in categories outside their established core expertise. When a brand expands into distant categories, 50% of its appearances are as a citation, while only 25% are as a named mention. Only 9% of brands achieve the "gold standard" of being both cited and mentioned in these distant spaces. In contrast, within categories closely related to a brand’s core expertise, the numbers improve significantly: 74% are cited, 44% are named, and 34% achieve both. The takeaway for content teams is clear: You can be a reliable source for information on almost any topic if your content meets the basic criteria for utility and relevance. However, becoming a recommended brand—a named entity that the LLM actively suggests—requires a much higher threshold of perceived authority that is deeply tied to your core category. Chronology of the Research The study, which tracked data from January through June 2026, was designed to test the "spread thin" hypothesis—the fear that covering too many topics dilutes a brand’s authority. Phase One (January–March): The research team established baseline authority scores by tracking 1,094 categories. They mapped these against the core expertise of the brands involved using semantic similarity measurements. Phase Two (April–May): The team tested the "depth" of presence. They measured how often a domain appeared across the five prompt variants provided for each category. Phase Three (June): The study synthesized these data points to determine if expansion into new categories resulted in a net gain or loss in total visibility, controlling for variables like branded search demand and organic traffic. Supporting Data: Depth Over Breadth One of the most persistent myths in SEO is that a site should strictly limit its topical scope to avoid being seen as a "jack-of-all-trades." The data provides a more sophisticated view. The study found that simply appearing in many categories does not inherently penalize a domain. In fact, being cited in many categories shows no "spread thin" effect. Conversely, the "penalty" often associated with breadth is actually a penalty for shallow presence. The Depth Factor: When a domain appeared in only one out of five prompt variants, it often failed to gain traction. However, as depth increased—specifically, when a brand showed up in all five prompts—the association with citation share rose from +0.012 to +0.062. The Recommendation Gap: While citation-only presence remains stable across categories, brand mentions tell a different story. Being mentioned in only one prompt variant is associated with a lower mention share (-0.051). However, at five-out-of-five, the association turns slightly positive. This suggests that "owning" a category is a prerequisite for successful expansion. Brands that consistently dominate their primary space possess the momentum required to be recognized as authorities in adjacent spaces. Industry-Specific Variations Not all markets behave the same way under the gaze of an LLM. The research highlighted two distinct clusters: 1. Finance and Real Estate (High-Citation Sensitivity) These sectors exhibit a strong repeat-citation pattern. In these markets, broader, repeated coverage is positively correlated with stronger source visibility. For these brands, category expansion is a viable strategy because the LLMs prioritize consistent, high-volume sourcing. There is no obvious "penalty" for being broad here; the algorithm rewards the ubiquity of the source. 2. Legal and Healthcare (High-Mention Sensitivity) These industries face much tighter constraints. While consistent coverage can improve a domain’s citation position, it does not reliably lead to more named-brand visibility. The study posits that because these topics involve high-stakes, life-altering decisions, the AI’s threshold for "recommendation" is significantly higher. A brand must be more than just a source; it must be an established, authoritative entity to be named in an answer. Implications for Content Strategy For CMOs and SEO leads, these findings necessitate a pivot in how we measure success in the AI era. Shift from "Ranking" to "Recommending" If you are operating in the legal or healthcare space, do not chase broad categories with thin content. You are unlikely to be "recommended" unless you have already established profound topical depth. Focus on building brand equity that the AI can associate with expertise. Use Citations as an Early Warning System For those in finance, real estate, or retail, use your citation share as a leading indicator. If you are being cited across an adjacent category but are not yet being "mentioned," you are on the right track. Your goal should be to deepen your content density in that category until the AI shifts from citing your link to naming your brand. The Death of the "Universal Limit" The data does not support the existence of a hard, universal limit on how many categories a domain can serve. The limitation is not on the number of topics, but on the quality of engagement within those topics. If your team is capable of producing deep, consistent, and expert-level content across multiple verticals, the LLM will follow. The Role of Implementation Kits As content operations become more complex, the need for standardization grows. The research emphasizes that when multiple writers contribute to a brand’s output, the quality of the "source" material must be uniform. Tools such as AI Writing Implementation Kits are becoming essential, as they allow teams to audit content against the specific criteria that LLMs use to determine authority. Conclusion The transition from keyword-based search to AI-driven discovery does not mean that topical authority is dead; it means it has been recalibrated. The new "Authority" is defined by depth of coverage and the ability to distinguish between being a mere source of information and a trusted, recommended brand. While the "spread thin" penalty is largely a myth, the "shallow content" penalty is very real. Brands that intend to expand must do so with the understanding that the AI is watching not just what they talk about, but how consistently they occupy the conversation. For businesses, the path forward is clear: define your core, prove your expertise through deep, repeated presence, and use your citation footprint as a compass to navigate into new, adjacent territories. Post navigation The Accessibility Tree: Why AI Agents Are Changing the Future of SEO
