As AI search grows, 'which brand does the AI mention first' is becoming a new competitive metric. According to data recently published by Semrush, topical authority still matters — a lot — in AI search, and once a brand earns category ownership, it tends to hold onto it.
Semrush analyzed 1,094 US categories, five prompts per category, using monthly ChatGPT snapshots from January through June 2026 — covering more than 220,000 domains, 50,000+ brands, and 600,000+ citations. The study found that only 15.2% of categories have a clear owner, while 53.7% remain open fields with multiple contenders. More strikingly, 89.3% of estimated AI-search demand sits in categories that still have no clear owner. A category owner is defined as the brand with the highest share of mentions, appearing in at least four of five prompts, with at least a 5-percentage-point lead over the runner-up.
One counterintuitive finding: the most-cited domain was rarely the most-mentioned brand — only 20.8% of the time — while the most-mentioned brand was cited at least once 69.9% of the time. In other words, brand mentions drive AI recommendations more than citation frequency does. Category owners also showed 55.7% higher brand search volume, 48.4% higher organic traffic, and 52.5% higher authority scores than their nearest competitor, and retained the top spot in 90.4% of month-over-month comparisons. Leadership is more fragile, though, when the gap is thin — switching leaders had a median lead of just 1.3 points, versus 2.9 points for leaders that held their position.
Rather than chasing broad content coverage, Semrush recommends owning narrow, specific categories. Its example: instead of targeting "finance" broadly, a payroll software company should build content around W-2 deadlines, contractor classification, payroll-tax errors, overtime rules, and state registration — the exact questions it wants to be known for. Ownership concentration also varies sharply by industry: retail leads at 61.6%, while legal has no clear owners at all.
Practically, this means defining the 10-20 categories most tied to your business, and periodically testing five representative prompts per category — definitions, comparisons, alternatives, use cases, and buying questions — to see where you show up and where competitors do but you don't. Closing those specific gaps with better positioning, comparison content, and credible coverage is the next step. To build an AI-search strategy for your brand, explore Best Partner's services or get in touch.
A brand with the highest share of mentions in a category, appearing in at least four of five representative prompts, with at least a 5-percentage-point lead over the next brand.
Not necessarily. Only 20.8% of the time was the most-cited domain also the most-mentioned brand, while the most-mentioned brand was cited at least once 69.9% of the time — mentions matter more than citation frequency.
Usually. Category owners retained the top spot in 90.4% of month-over-month comparisons, though leadership is more fragile when the gap over the runner-up is thin.
Define the 10-20 categories most relevant to your business, test five prompt types per category (definitions, comparisons, alternatives, use cases, buying questions), and close the gaps where competitors appear but you don't with targeted content and positioning.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.
AI-era SEO is the Wild West again. After a decade inside Microsoft Bing, Duane Forrester says stop chasing the leading edge and work the chunky onions. What practitioners should drop, and what still holds.
The Google Search favicon bug is a week old and Google has confirmed it is on their end. Here is how to tell whether it is your site, and why this week's CTR drop should not enter your reporting as a result.
2027 SEO strategy shifts from ranking to being understood, trusted, and cited. Here are the priorities: AI answer engines, E-E-A-T, brand mentions, zero-click, and structured data.