TL;DR Summary:
Depth Wins: New research says AI tools are more likely to cite your brand across many topics, but they recommend you by name mainly where you already have strong, repeated presence.Citations vs Mentions: In distant categories, brands were cited 50% of the time but recommended only 25%, while close-to-core categories rose to 74% citations and 44% mentions.Industry Differences: Finance and real estate benefit more from broader topic expansion, but legal and healthcare still struggle to turn citations into mentions even with deeper coverage.Should you write about one topic or many if you want AI tools to recommend your brand? New research from Semrush’s AI Visibility Toolkit answers this with real numbers, and the answer changes depending on whether you want a citation or a mention. Topical authority in AI Search works differently than most people assume, and the difference matters for anyone planning content for the second half of 2026.
Topical Authority in AI Search Splits Citations From Recommendations
The study, covering 1,094 categories and 1,458 mapped brand entities from January through June 2026, found a clear gap between getting cited and getting recommended. In categories far from a brand’s core expertise, ChatGPT cited the brand 50% of the time but named it as a recommendation only 25% of the time. In categories close to a brand’s expertise, citations rose to 74% and mentions rose to 44%. Citation-only presence barely moved between distant and close categories, at 41% versus 40%. This tells you AI tools will pull your site as a source on almost any subject, but they only recommend you by name where you have already shown up consistently. Because this gap between being cited and being recommended is exactly the kind of thing brands struggle to see on their own, platforms like SnowSEO are built to track whether AI tools are actually citing and recommending a brand across categories, and how that visibility share stacks up against competitors over time.
Depth Matters More Than Spread
Kevin Indig’s analysis measured depth as how many of five prompt variants a brand appeared in in each category. For citations, showing up across many categories carried no penalty. The association rose from +0.012 at minimal coverage to +0.062 at full five-out-of-five coverage. Brand mentions told a different story. At one out of five prompt variants, the mention association dropped to -0.051. At full coverage, it turned slightly positive. This means the “spread too thin” idea in topical authority in AI Search is really about shallow presence, not about covering many categories.
Finance and Real Estate Reward Repeat Citations
The topical authority effect in AI Search does not apply evenly across industries. In finance, the citation-breadth association climbed from +0.054 at minimal coverage to +0.139 at full coverage. Real estate moved from +0.021 to +0.135. Both industries showed no drop-off from covering more categories. If you work in finance or real estate, testing expansion into related topics makes sense, with citation share as your early signal of progress.
Legal and Healthcare Set a Higher Bar for Mentions
Legal and healthcare showed the opposite pattern. In legal, the mention-breadth association stayed negative even at full coverage, at -0.058. Healthcare followed the same path at -0.037. Wider coverage still helped citations in these fields, but it did not turn into more brand mentions. The likely reason is that legal and healthcare involve higher-stakes decisions, where being a source is not enough for AI tools to recommend a specific brand.
What the Data Does Not Prove
Every number in this study is an association after controls for factors like Authority Score, organic traffic, and branded search demand. None of it proves that publishing more content or expanding into new categories caused the results. The researchers also found no fixed limit on how many categories a brand can serve well. Treat these findings as a guide for testing, not a formula.
If you take one thing from this research, focus on depth before breadth. Build repeat presence in your core categories first, since that is what turns citations into recommendations. Once you see consistent mention share in your main topics, testing adjacent categories carries less risk, especially if you work in finance or real estate. As you test depth-first content strategies in your core categories, tools like SnowSEO’s AI visibility tracking can help you monitor whether those repeat citations are actually converting into AI-generated recommendations over time.
