You Have Months, Not Years, to Become the AI’s Default Answer in Your Niche
Quick summary: an analysis of more than 50,000 brands and 1,094 categories on ChatGPT found a clear owner in just 15.2% of them. The categories with the highest AI search volume — the same ones that account for 98% of estimated demand — are precisely the ones least likely to have an owner, with an ownership rate of 11.3%. And once a brand claims the owner position with a comfortable margin, it holds the lead in 90.4% of month-over-month comparisons. Topical authority in AI behaves like compound interest: whoever enters an open category early accumulates an advantage that gets more and more expensive for latecomers to close.
What does it mean to “own” a category in AI?
Picture a payroll software company deciding on its next batch of 20 articles. It can chase a broad, high-volume finance topic, or build around the specific payroll questions it wants to be known for answering: W-2 deadlines, contractor classification, payroll tax errors, overtime rules, state registration. Betting on topical authority means choosing the second path, even knowing ChatGPT already answers some of those questions without generating a click. The goal isn’t glossary-page traffic — it’s becoming the brand that shows up every time someone asks something related, including the question that decides the purchase: “which payroll software should I use?”
The study, conducted by Kevin Indig with data from the Semrush AI Visibility Toolkit, defined that status objectively. An “owner” brand appears in at least 4 of the 5 representative questions for a category and holds at least a 5-percentage-point lead over the runner-up. An “emerging leader” brand appears in at least 3 questions without reaching that margin. An “undecided” category is one where no brand appears in 3 or more questions.
Why do only 15.2% of categories have an owner?
The analysis covered 1,094 U.S. categories, with five representative prompts each, tracked monthly on ChatGPT between January and June 2026, adding up to more than 220,000 domains, 50,000 brands, and 600,000 citations. As of June, only 15.2% of categories had a clear owner. Another 53.7% were wide open, with multiple competitors fighting for the top spot and nobody consolidated.
That’s quite different from what most people assume about AI search. The common expectation is that a large, established brand would already dominate any relevant category, the way it dominates traditional search. The data doesn’t back that expectation up.
Why are the highest-volume categories the ones least likely to have an owner?
This is the most counterintuitive finding in the study. The normal expectation is that a higher-demand category attracts more competition and, as a result, settles around a winner faster, the way it happens on Google. On ChatGPT, it’s the opposite.
Splitting the 1,094 categories into two groups by AI search volume, the higher-demand group had an ownership rate of just 11.3%, versus 19.0% in the lower-demand group. And that higher-demand group accounts for 98% of the sample’s entire estimated volume. Put the two together: 89.3% of AI search demand sits in categories with no defined owner. Demand concentrates; ownership doesn’t follow.
What decides who becomes the owner of a category?
The study measured ownership by brand mention within the answer, not by source citation, and that choice isn’t arbitrary. Earlier research from Growth Memo itself on how users navigate purchase decisions in AI showed that 74% of participants picked as their final choice the brand mentioned first in the answer, and 64% never clicked a single source during the entire task. The user reads the AI’s text, maybe scrolls through an embedded product card, and already declares a finalist, without building their own comparison against an external source.
That’s why citation and mention aren’t the same thing, and the correlation between the two is slightly negative. The most-cited domain is rarely (20.8% of cases) also the most-mentioned brand. But the most-mentioned brand almost always (69.9% of cases) shows up cited at least once. Citation helps form the answer. Mention is what the user actually sees and uses to decide.
As for traditional SEO metrics, they help, but they don’t decide on their own. Comparing the owner against the runner-up in each category, the owner had higher brand search volume in 55.7% of pairs, higher organic traffic in 48.4%, and a higher Authority Score in 52.5%. In other words, in nearly half of categories the runner-up had at least one of those advantages and still lost the owner position. Traditional SEO strength helps get you into the conversation, but topic-specific factors — source relevance, quality of the cited content, and reputation — carry more weight in deciding who becomes the reference.
Once you become the owner, can you lose the position?
You can, but it gets harder and harder as the margin grows, and that’s where the compound-interest comparison holds up. An owner with a comfortable margin held the lead in 90.4% of month-over-month comparisons. Leadership changes did happen — 1,950 out of 5,470 total comparisons — but they were concentrated in categories where the initial advantage was already thin: a median margin of 1.3 percentage points among the swaps, versus 2.9 points among the positions that held.
That’s what compounding advantage looks like: whoever enters an open category and builds a solid margin early tends to stay, because every month of mentions reinforces the position for the next month. Whoever comes in late, into a category that already has an owner with a wide margin, is competing against an advantage that’s already been compounding for months.
How do you put this into practice?
The study suggests five moves. First, build a list of 10 to 20 commercially important categories for your brand and track at least 5 representative prompts per category: definition, comparison, alternative, use case, and purchase question. Second, classify each category as clear owner, emerging leader, or undecided. Third, prioritize investment in commercially valuable categories with no clear owner, or with a small margin between leader and runner-up, defending categories your brand already leads and being more selective before attacking a category where a competitor already shows up in nearly every prompt.
Here’s an example of how that plays out in practice: the same payroll software company runs its five watchlist prompts for “overtime calculation” and finds no brand appearing in 3 of the 5 questions — an undecided category. Alongside that, it runs the same test for “best payroll software” and finds a competitor present in 4 of 5 prompts with a 12-point lead over the runner-up — a category that already has an owner with a wide margin. The investment decision changes a lot between the two: the first is worth an immediate push, the second is worth a long-term fight or repositioning around a more specific slice that still has no owner.
Fourth, close the gap on prompts where a competitor shows up and your brand doesn’t, which might mean clearer product positioning, a comparison page, documentation, concrete proof, and trustworthy third-party coverage. Fifth, measure with the right metric for each thing: brand mention measures category ownership, citation measures source authority, and referrals, pipeline, or revenue measure the business outcome. An isolated shift in which brand appears most in a given month is a reason to investigate, not proof that a specific action caused a win.
My take on this
This kind of data matters twice as much for anyone building content or a tool in a technical niche like GEO, which is still new enough to have a good shot at falling in the 84.8% that’s still up for grabs. The temptation at this stage is to try to cover every adjacent subtopic so as not to “miss out” on any conversation. The data suggests the opposite. Spread-out coverage dilutes the odds of any specific prompt turning into a consistent mention. Picking a handful of questions your brand genuinely wants to answer, and answering each one more specifically than any competitor, builds ownership faster than broad, shallow coverage.
Summary: the window is open, but not for long
- Only 15.2% of the 1,094 categories analyzed have a clear owner on ChatGPT
- 89.3% of AI search demand is concentrated in categories with no owner
- Ownership is decided by brand mention in the answer, not by source citation, and the two correlate negatively
- Traditional SEO metrics (traffic, authority, brand search) only match category ownership in about half of cases
- An owner with a comfortable margin holds the lead in 90.4% of the following months, an advantage that compounds like interest
The cost of picking the wrong next 5 categories to go after is lower today than it will be 12 months from now. The window to become a niche’s default answer in generative AI is open right now. It won’t stay that way. If you want to run this diagnosis on your own business’s categories before a competitor locks in the advantage, a strategic SEO audit maps out where to start.
Sources: Does topical authority matter in AI Search?, by Kevin Indig, Growth Memo; How Consumers Navigate High-Stakes Purchases in AI Mode, also from Growth Memo.