Here's the number that should rearrange your Q3 priorities: 89.3% of estimated AI search demand currently sits in categories with no clear brand owner. That's from Kevin Indig's analysis of six months of ChatGPT answers data via Semrush, covering 1,094 U.S. categories, five prompt types per category, more than 50,000 brands, and over 600,000 citations.
The kicker is the second number. Once a brand owns a category, it held first place in 90.4% of month-over-month comparisons. Topical authority in AI search is durable, meaning the current land grab is the last easy one.
The Highest-Demand Categories Are the Least Claimed
Indig split the 1,094 categories into two halves by estimated AI search volume. The higher-volume half accounts for 98% of all AI search demand in the sample, with a clear-owner rate of just 11.3%. The lower-volume half had a 19% owner rate. The most valuable categories are currently the least decided.
His definition of "clear owner" is strict: one brand must appear in at least four of five test prompts and beat the runner-up by five percentage points or more. This high bar means some categories have multiple credible contenders without a dominant player. If you're a B2B SaaS brand needing to be the default answer in your category, the odds are good that nobody has locked it down yet.
53.7% of all categories were open fields with no brand close to locking the door. Only 15.2% had a clear owner.
Citations Are the Door, Not the Room
Here's where many GEO strategies falter. Indig found that citations and brand mentions are only weakly related, with a slightly negative correlation of -0.229. The most frequently cited domain in a category matched the most frequently mentioned brand only 20.8% of the time.
This poses a problem if your AI search strategy is solely focused on "getting more citations." Citations may get you into the answer set, but they don't ensure that your brand is named. A 2026 analysis found brands are 6.5x more likely to be cited through third-party sources than their own domains. AI engines treat earned media and independent validation as less biased evidence, aligning with Okara's analysis on the same point.
The real play isn't the volume of citations but being the brand that shows up when the model answers five types of questions about your category: definition, comparison, alternatives, use case, and buying decision. This requires breadth of coverage across prompt types, not just depth on one keyword.
The 3-Point Rule and Where Turnover Happens
Indig's durability data reveals a sharp edge. Categories where the leader changed hands had a median lead of just 1.3 percentage points going into the switch. In contrast, categories where the leader held on had a median lead of 2.9 points. Growth trajectory alone told him little; brands trending upward still got overtaken when the gap was thin.
The practical threshold: anything under about 3 points is contested. If you're leading a category by 1.5 points, you haven't won anything. You're in a fight, and the other side may not even know it yet.
What to Actually Do This Quarter
Identify 10 to 20 categories where your brand needs to be the default answer. For each, track five prompt types (definition, comparison, alternatives, use case, buying question). Sort by lead size, not by current appearance. Anything under 3 points requires real resourcing.
Shift budget from pure citation-chasing toward content types that determine whether you're the named brand: comparison pages, proof points, third-party coverage, and clear entity positioning. Brands appearing in AI Overviews tend to have clear topical authority and strong third-party validation, not necessarily the most backlinks, according to Google-guidance-based analysis from 2026.
Operationalize earned media as an AI search input, not a nice-to-have. If brands are 6.5x more likely to be cited via third-party sources, then PR and analyst relations directly feed your AI visibility pipeline. Track AI citations, branded search lift, and mention share alongside traditional SEO metrics. Click-based reporting alone won't reveal whether you're winning or losing in answer-first experiences.
The Trade-Off You're Accepting
This approach narrows focus. You're selecting a small number of categories and going deep instead of spreading content across everything your product touches. Volume may drop before quality improves. The alternative is spreading thin across categories that are about to be locked down by whoever concentrates first.
The 89% number won't hold. Indig's durability data makes that clear. Brands that act in the next two to three quarters will still be the default answer when this study is rerun in 2027. Those that bookmark the report and wait will be trying to unseat an entrenched leader with a 3-point head start and a compounding advantage.
That's the real constraint here: not budget, tooling, or headcount, but time.
