What happened
Semrush, working with Kevin Indig, ran 1,094 U.S. topic categories through ChatGPT between January and June 2026 — five buyer-style prompts per category, covering definitions, comparisons, alternatives, use cases, and buying decisions. The dataset behind it: over 50,000 brands, 220,000 domains, and 600,000 citations. The question: does any single brand actually own the answers for a topic?
Mostly, no. A brand counted as the category owner if it held the highest share of mentions, showed up in at least four of the five prompts, and led the runner-up by five points or more. Just 15.2% of categories cleared that bar. Another 31.2% had an emerging leader. The majority — 53.7% — were unsettled, with no brand appearing consistently at all.
Ownership is scarcest where it's worth the most: only 11.3% of high-demand topics had a clear owner, against 19% in the bottom half — and the top half carried 98% of the AI search volume in the sample. And once a brand does own a topic, it keeps it: owners held first place in 90.4% of month-over-month comparisons.
Why this matters
The finding kills the screenshot as a measurement instrument. A single prompt where ChatGPT recommends you proves almost nothing — and the same is true in reverse when it recommends a competitor. With most categories unsettled, one answer is closer to a coin flip than a verdict. Presence across a spread of related buyer questions is the only reading that means anything.
The unsettled majority is the real headline for stores. In 53.7% of categories nobody shows up consistently — which means the seat is empty, not taken. Pair that with the stability number and the strategy writes itself: these positions are winnable now and sticky once won. Brands that establish consistent presence while categories are still open inherit an incumbency that held first place nine times out of ten.
Of the SEO metrics Semrush tested, branded search volume was the one that statistically separated owners from everyone else — organic traffic and Authority Score didn't. Read that as directional rather than gospel, but it points at brand demand, not backlink muscle, as the correlate of AI-answer ownership.
What to do about it
Test five prompts, not one
Pick your money category and write five real buyer questions — a definition, a comparison, an alternatives ask, a use case, a buying decision. Run them through ChatGPT and note every brand mentioned. That mirrors Semrush's methodology, and it tells you whether your category has an owner, a contender, or an empty seat.
Score yourself against the ownership bar
Ownership meant four-of-five prompt presence with a five-point lead over the runner-up. Appear once and you're noise; three times and you're the emerging leader. Track the grid monthly — the 90.4% stability figure means movement is slow, so the trend matters more than any single check.
Prioritize open high-demand categories
Top-half topics carry 98% of the volume and are owned only 11.3% of the time. That's where early, consistent presence buys the most. Map which of your product categories sit in that high-demand, no-owner quadrant and aim the content and PR work there first.
Invest in branded demand
If branded search volume is what separates owners, the work is making people search for you by name: distinctive positioning, consistent category association, content other sites cite with your brand attached. Generic keyword rankings didn't predict ownership in this data. Being known did.