89% of AI Search Demand Has No Clear Owner — Here Is the Category Playbook

Nine out of ten AI search categories have no clear brand owner. That is the headline finding from Kevin Indig's analysis of six months of ChatGPT citation data, published July 20, 2026, using Semrush data across 1,094 US categories, more than 50,000 brands, and over 600,000 citations. The biggest categories by volume are the least decided — and the first brand to lock one down is extremely hard to dislodge. Here is what that means for your next quarter.

The Market Is Wide Open, but Not for Long

Indig built three buckets based on how consistently a brand appears across five standardized prompts per category. A category has a clear owner when one brand shows up in at least four of five prompts and leads the runner-up by five percentage points or more. An emerging leader hits three prompts without that gap. Everything else counts as unsettled.

In June 2026, only 15.2% of categories had a clear owner. Over half — 53.7% — were open fields with multiple credible contenders and no brand close to locking the door.

The kicker: when Indig ranked all 1,094 categories by estimated AI search volume and split them in half, the higher-volume group accounted for 98% of all demand in the sample but had the lower owner rate — 11.3% versus 19% for smaller categories. Stack that up and 89.3% of estimated AI search demand sits in categories with no clear owner.

The categories worth the most money are the least decided.

Early Leads Are Extremely Durable

This is the part that should change your planning. Indig tracked month-over-month leadership changes across the same categories and found that a clear owner held first place in 90.4% of comparisons.

The switches that did happen clustered in categories where the lead was already thin. Categories where the leader changed had a median lead of just 1.3 percentage points going in. Categories where the leader held had a median lead of 2.9 points. Growth trajectory alone told almost nothing — brands trending upward still got overtaken when a competitor grew faster from a close-enough starting point.

The lesson: a one- or two-point lead is not a lead. A brand sitting in that range should assume it is still in a fight. But a brand that opens a five-point gap tends to keep it.

What Actually Gets Cited (It Is Not Your Homepage)

Indig's data also shows which page types ChatGPT pulls from. The most cited are product or service landing pages, followed by editorial content. Homepages accounted for just 4% of citations.

If your AI visibility strategy is built around brand-level SEO — domain authority, homepage optimization, link profiles — you are optimizing for the wrong surface. ChatGPT cites specific pages that answer specific questions. That means your category play is a content architecture decision, not a domain-level one.

This matters more than it used to because the traffic that does arrive from AI platforms converts at 14.2% versus Google organic's 2.8% — a 5x quality premium, according to AI Business Weekly's analysis of 2026 market data. The visitor who reaches your site through an AI citation already completed the consideration phase inside the answer. They are not browsing. They are buying.

The Playbook: How to Claim a Category Before It Closes

Based on Indig's durability data and what I am seeing across the brands we work with, here is the execution path:

1. Identify your category. Use Semrush's AI Visibility Toolkit or a comparable monitor to find the category where your brand already appears — even inconsistently. If you show up in two of five prompt types, you are an emerging contender. If you show up in none, you are invisible.

2. Map the five prompt types. Indig used five standardized prompts per category. Your category likely has equivalents: "best X," "how does X work," "X vs Y," "what should I know about X," and a recommendation prompt. Build a page that directly answers each one.

3. Prioritize product and service pages over editorial. The citation data says product pages get cited most. Make sure your core offering pages are structured for extraction — clear answers, specific claims, comparison data.

4. Build depth, not breadth. The entire dataset confirms that topical authority matters in AI search. Brands that go deep on their category hold their lead. Brands that go wide get outpaced by specialists.

5. Measure monthly. A one-point lead means nothing. Track your share of mentions across prompt types monthly. The goal is a five-point gap over the next competitor — that is where durability kicks in.

Why the Skeptics Are Asking the Wrong Question

A MediaPost column published July 28 asked whether AI search visibility is "fool's gold or the future." The author's concern: citations do not carry impression value, consistency is low, and the ROI math does not close on informational queries.

That framing treats AI visibility as a performance channel. It is not — and as Search Engine Land reported in June, AI visibility is increasingly tied to brand authority rather than traditional SEO metrics. Indig's data shows it is a category ownership race. The question is not "what is my CPA on a ChatGPT citation?" The question is "will I own this category in six months, or will my competitor?" Once a brand locks a five-point lead, the 90.4% retention rate means displacement requires a sustained, expensive campaign by the competitor — or it does not happen at all.

The brands that treat this as an ROI-per-citation exercise will keep debating the math. The brands that treat it as a land grab will own the categories. Both cannot be right for long.

Where Machine Relations Fits

This data maps directly to what we track through Machine Relations — the discipline of earning AI citations, recommendations, and visibility. The reason category ownership is durable is the same reason MR works: AI systems develop source preferences based on consistent, authoritative coverage of a topic. That is not SEO. That is relationship-building with the retrieval layer.

If your team is still treating AI visibility as a content SEO extension, Indig's data is the clearest signal yet that the game has changed. The window is open. Eighty-nine percent of the demand is unclaimed. But the durability data says that window will not stay open indefinitely.

FAQ

How long does it take to become a category owner in AI search?

Indig's data covers January through June 2026 — six months. Some brands established clear ownership within that window, and once they did, they held it in over 90% of month-over-month comparisons. The timeline depends on category competitiveness, but the data suggests consistent coverage across prompt types over three to six months is the minimum commitment.

Does traditional SEO ranking affect AI search category ownership?

Indig's study explicitly did not include organic ranking data, so there is no direct correlation in this dataset. However, the page types that get cited most — product pages and editorial content — are the same assets that perform in traditional search. The overlap is in content quality and topical depth, not in domain authority signals.

Is AI search visibility worth the investment for mid-market brands?

The MediaPost skeptic argues the math only works for Fortune 500 brands. Indig's data suggests the opposite: the biggest categories are the least decided, which means a focused mid-market brand that goes deep on its niche can own a category that a distracted enterprise competitor has not bothered to claim. The cost is content architecture and consistency, not a six-figure tooling budget.