Nearly 9 in 10 categories ranked by AI search demand volume have no dominant brand yet. That is the headline finding from six months of ChatGPT answer data spanning 1,094 U.S. categories, more than 50,000 brands, and over 600,000 citations (links or references to a domain appearing in ChatGPT’s answers). Analyst Kevin Indig published the breakdown in his Growth Memo Substack on July 20, 2026, drawing on data provided by Semrush and covering January through June 2026.
What Does “Owning” a Category in AI Search Actually Mean?
Indig defined category ownership with a specific threshold. A brand qualifies as a clear owner when it appears in at least four of five test prompts for a category and leads the nearest competitor by five percentage points or more. An emerging leader shows up in three or more prompts without hitting that gap. Everything else counts as unsettled: no brand consistently dominates even three of five prompts.
Applied across all 1,094 categories in the Semrush dataset as of June 2026, only 15.2% of categories had a clear owner. Another 53.7% were open fields with multiple credible contenders and no brand close to locking the door, per Indig’s Growth Memo.
Why Are the Highest-Value Categories the Least Settled?
Indig split the 1,094 categories into two even groups by estimated AI search volume. The higher-volume half accounts for 98% of all AI search demand in the sample and has an owner rate of just 11.3%. The lower-volume half has a higher owner rate of 19%. Stack those together and 89.3% of estimated AI search demand sits in categories with no clear owner as of June 2026, according to Indig’s analysis of Semrush data.
In practical terms, the categories representing the most commercial intent inside AI search are also the ones where no brand has locked in a default position.
How Durable Is a Category Lead Once a Brand Earns It?
Indig tracked month-over-month leadership changes and found that a clear owner held first place in 90.4% of comparisons. That looks stable until you examine when switches did happen.
Categories where a lead switched had a median lead of just 1.3 percentage points going into the switch. Categories where the leader held on had a median lead of 2.9 points, more than double. Indig also found that growth trajectory alone was not a reliable signal: three examples in the data showed a leading brand trending upward month over month and still getting overtaken because a competitor simply grew faster or closed a narrow gap. Per the Growth Memo analysis, a lead under roughly three points is best treated as a contested position, not a secure one.
Why Getting Cited a Lot Doesn’t Make You the Brand ChatGPT Names
One of the more counterintuitive findings concerns the relationship between citations and brand mentions (the model actually naming a brand by name). The two turned out to be only weakly related, with a slightly negative correlation of -0.229 in Indig’s Semrush data.
The most frequently cited domain in a category matched the most mentioned brand in only 20.8% of cases. However, the most mentioned brand was cited at least once 69.9% of the time. Racking up citations is not the same as becoming the brand ChatGPT reaches for when it names a default answer. Indig’s data suggests citations are a threshold condition, not the deciding factor.
On what clear owners had in common compared to their nearest competitor, Indig found: higher branded search volume in 55.7% of category pairs, higher organic traffic in 48.4%, and a higher Semrush Authority Score in 52.5%. He explicitly labels these as correlations, not a causal formula.
What This Means for AI-Search Visibility
The gap most brands are missing right now is the difference between being in the answer and being the answer. Here is a concrete way to picture it: if someone asks ChatGPT “what are the best project management tools,” your brand might appear as item four in a list of seven options. That counts as a citation. But “owning the category” means your brand is the one ChatGPT names unprompted when a user asks the broader question. Those are two different outcomes, and Indig’s data suggests they require two different strategies.
The five prompt types Indig used to test each category map closely to a buyer’s journey: the definition question, the comparison, the alternatives list, the use case, and the buying decision. A brand optimizing only for the buying-intent prompt is leaving four of five category footprint positions unaddressed. In Indig’s framework, ChatGPT’s default-answer status is earned by showing up consistently across all five variations. A brand present in only two or three is still in a contested position even if its individual citation count looks healthy.
The source data points toward something most GEO (generative engine optimization, the practice of making a brand visible in AI-generated answers rather than traditional search results) strategies currently underweight. Most GEO dashboards track citation volume because it is the easiest number to pull into a report. Indig’s -0.229 correlation says citation volume and brand-naming frequency are almost unrelated. A brand can top the citation leaderboard in its category and still lose the naming race to a competitor with fewer citations but broader prompt-type coverage. That means Share of Model (the share of AI answer slots where a brand actually gets named, compared to competitors) is a more relevant metric than citation count alone, yet it is the one most teams are not yet tracking.
In Hingewise’s assessment, this creates a specific misread risk: brands that benchmark their GEO progress against citation counts may believe they are winning a category they are, in practice, still losing. The 89.3% figure is a real opportunity, but it will be captured by teams measuring the right thing, not just the convenient thing.
Before You Target an AI Search Category: A Quick Checklist
- List the 10 to 20 categories where your brand needs to be ChatGPT’s default answer, not just visible.
- For each category, test all five prompt types used in Indig’s framework: definition, comparison, alternatives, use case, and buying decision.
- Check your lead size, not just whether you are currently in first place. A lead under roughly 3 percentage points should be treated as contested, based on Indig’s turnover data.
- Separate your citation count from your brand-mention rate. Being in the list and being the named answer are different outcomes requiring different content investments.
- Map content gaps to the specific prompt types where your brand is absent or underrepresented, not just the overall category.
- Track month-over-month movement. A competitor who closed 1.5 points in 30 days is a real risk regardless of who holds the lead today.
- Prioritize higher-volume categories first: per Indig’s Semrush data, those are where 98% of AI search demand sits and where the owner rate is lowest at 11.3%.
The figure worth watching over the next two quarters is whether the category owner rate in high-volume categories moves materially above 11.3%. If it does, the window Indig describes in his July 2026 analysis is narrowing faster than the current snapshot suggests.
Author: Lam Nguyen, Hingewise
Source: Kevin Indig, Growth Memo Substack, July 20, 2026, via Search Engine Journal. Data: Semrush, January-June 2026, 1,094 U.S. categories, 5 prompts per category, 50,000+ brands, 600,000+ citations.
Sources
This article was drafted with AI assistance from the sources cited above and checked against them before publishing. The "What this means for AI-search visibility" section is Hingewise's editorial analysis, built on the frameworks and positions we use in our GEO work. How we work: Editorial policy.