The $750 billion shift
You weren't asked. You didn't choose. And right now, 60% of AI-assisted discovery is putting brands in categories they never targeted.
The question is: do you know which ones?
$750B
US revenue through AI search by 2028
50%
of consumers already use AI search
84%
of brands don't track it
Source: McKinsey, October 2025
Where CategoryRank sits vs. traditional tools
Keyword rankings, backlinks, search visibility. Backward-looking. Inferred from search behavior.
Survey-based awareness, consideration, preference. Self-reported data, not AI perception.
Count how often AI mentions your brand. Measures frequency, not categorical understanding.
Measures categorical assignment + distance from canonical ontology. Detects perception drift before market impact.
The Fundamental Difference
Most tools ask: "Are you mentioned? How often?"
CategoryRank asks: "What does the model think you ARE?"
LLMs don't reason in keywords—they reason in latent concept space. Distance from canonical category centroids reveals misalignment, arbitrage opportunities, and emerging demand before the market prices it in.
When a human searches Google, they scan 10 blue links and might explore.
When someone asks AI "What's the best power management IC?", something different happens:
AI's decision process:
Category assignment happens before ranking.
If AI classifies your brand into the wrong category—or a category you never targeted— you will never be recommended. No SEO fixes this.
When someone asks AI "What's the best cross-training shoe?", they get instant answers about price, functionality, style, features, and reviews—synthesized from dozens of sources.
No scanning 10 blue links. No clicking through review sites. AI decides before the click happens.
From the McKinsey research:
In credit cards, hotels, electronics, and apparel—top brands are absent from AI answers, including Google AI Overview.
Your own website accounts for only 5-10% of what AI references. The rest? Third-party content, reviews, communities you don't control.
The result: brands may see 20-50% traffic decline from traditional search.
This is the new reality. The question is what you do about it.
10
AI Models
20K+
Categories
Weekly
Collection
KIM
Ontology
Find categories AI invented where you already have authority. Low-competition, high-intent opportunities.
See which categories you're #1 in—and which competitors are one position away from taking your spot.
When AI categorizes you differently than your marketing says, that's either a problem to fix or an opportunity to embrace.
SEO happens after the click. CategoryRank shows you what happens before—the invisible layer that determines if you're even considered.
Knowing where AI sees you is step one. The value is in what you do next.
Your keyword strategy was built for Google. But AI thinks in categories, not keywords. See which categories AI already associates with your brand—then align your content to match.
AI is creating categories that don't appear in any keyword tool or analyst report. Some of them already feature your brand. These are opportunities you're not marketing—yet.
Categories shift. Competitors move. AI models update their knowledge. Weekly tracking shows you where you're gaining ground, where you're slipping, and where new opportunities are emerging.
Right now, AI is telling millions of people which brands to consider.
It's making that decision based on categories.
Categories you didn't choose. Categories you might not even know about.
Shouldn't you know which ones?
About the Founder
Biology major. Plays cello. MBA from UC San Diego.
I've spent the last two decades working on systems that had to work: payment infrastructure at Microsoft, TN3270 emulator support at IBM, defense analytics for the Pentagon, and supply-chain intelligence at Arrow.
I was fortunate to be part of teams that exited in search and data infrastructure. I wasn't the central figure—just someone close enough to learn how durable systems and real companies are built.
CategoryRank came from a simple observation: AI doesn't just answer questions. It remembers and classifies the world. Yet very little effort is spent measuring what models actually know, how they group concepts, or how that memory changes over time.
AI Perception = f(category placement, recall strength, time)
We don't track people. We don't touch PII. We measure the function, not the user.
sam@categoryrank.ai
Find out which categories AI has assigned to your brand.