If you've asked ChatGPT or Perplexity about your category and your brand wasn't named, you've already done the hardest diagnostic step. Most teams never even check.

The next question is why. There are five common reasons. Each one has a different fix.

Reason 1: AI assistants don't know you exist yet

The simplest reason. If you're a newer brand, or you operate in a category that's underrepresented in training data, the model genuinely doesn't recognize you when it's asked who's in this space.

This is a "brand awareness" problem in the most literal sense. The model has never been trained to associate your brand with the category.

The fix: generate the awareness signals AI training models pick up.

• Press mentions in industry publications (not your own blog)
• Citations from comparison pages on other sites
• Mentions on Reddit, Hacker News, niche industry forums
• Coverage in newsletters that the model's training corpus likely included
• Wikipedia or schema.org entity entries if your brand qualifies

This is the slowest fix. Press takes months to influence training data. But it's also the most durable.

The faster workaround: lean hard on Perplexity-style retrieval-driven assistants. They search the live web, so they don't need your brand to be in training data — just retrievable from a high-authority page right now.

Reason 2: You have content but it's the wrong shape

Most "I have a lot of content and still don't get cited" cases come down to content shape, not content volume.

The classic failure pattern:

• 200+ blog posts written for SEO
• Most are 2,500–4,000 words
• Heavy keyword optimization, lots of intro fluff
• Few comparison pages, few use-case pages, few clear buyer's guides
• Schema is inconsistent or missing

This content ranks fine on Google for top-of-funnel queries but rarely gets cited by AI assistants because the answers aren't extractable.

The fix: replace volume thinking with shape thinking.

• Audit your top 50 pages. Of those, how many are comparison pages, alternatives pages, or buyer's guides? If under 10, you have a content-shape gap.
• For each high-traffic page that's just "what is X?", consider whether it should be replaced or supplemented by a "best X for Y" page.
• Make sure the answer to the page's title appears in the first paragraph. Not paragraph 4.
• Add FAQPage, Article, Product, or ComparisonPage schema as appropriate.

We've covered the shape patterns in detail in How to write content that AI assistants actually cite.

Reason 3: Your competitors are winning the comparison game

This is the most common reason for mid-sized brands that should be appearing but aren't.

Symptoms:

• You appear sometimes, but always after a competitor is named first
• Your competitor has a "X vs Y" page targeting you, and AI cites their page
• "Alternatives to [your competitor]" prompts return their alternative list — which doesn't include you

The fix: own your own comparison and alternative pages.

• Write the "[Your brand] vs [Competitor]" page yourself. Make it honest. AI loves honest comparison pages.
• Write the "Alternatives to [Your competitor]" page yourself. List yourself first with a real reason.
• Add ComparisonPage or Article schema with the comparison entities named.
• Internal-link these pages from your top product pages so they get crawl priority.

You'd be surprised how often the competitor's "vs you" page exists and yours doesn't. Fix that first.

Reason 4: Your structured data is missing or wrong

Schema is the cheapest, most-overlooked GEO lever. Pages without correct schema get extracted less reliably. Pages with broken schema get downweighted.

Symptoms:

• Comparison pages that have no Article or Product schema
• FAQ sections with no FAQPage markup
• Author bylines that aren't structured as Person objects
• Missing datePublished / dateModified

The fix: run Google's Rich Results Test on your top 20 pages. Anything that fails, anything that's missing relevant schema, fix it.

This is usually a single sprint of work — and it can move presence 5–15 points on its own.

We wrote the full schema playbook here: Schema markup for AI search visibility.

Reason 5: You're invisible to retrieval

The most diagnosable reason — and the one most teams miss.

AI assistants with web browsing (most consumer ChatGPT, Perplexity, Claude with web search) don't recommend brands they can't retrieve. If your pages aren't ranking in the underlying search index for relevant queries, they're not in the retrieval set, and they won't be cited regardless of how good the content is.

This is the most-fixable subset of GEO problems because the diagnostic is the same as SEO.

Diagnose:

• For each of your target prompts, type a slightly-rephrased version into Google.
• Are you on page one? On page two?
• If you're not in the top 30 organic results for relevant queries, you're probably not in the retrieval set.

Fix:

• Make sure your high-priority pages are technically crawlable (no noindex, no JS-only rendering)
• Check robots.txt doesn't block crawlers
• Make sure pages have unique titles and meta descriptions
• Build internal links to high-priority pages from your sitemap and home
• Earn backlinks (the slow but durable part)

Once you're in the retrieval set, the content-shape work above starts to compound.

Diagnose your own brand in 60 seconds

Submit your domain at /audit — Tracemetry will run three category-relevant prompts across ChatGPT, Claude, and Perplexity, parse the answers, and tell you exactly which of the five reasons above is your primary problem. Free. No signup.

If you want continuous tracking, the Pro plan runs 250 prompts every week, surfaces the specific gaps, generates source-grounded content briefs in your brand voice, and re-measures every week so you can see what's working.

The five reasons here are not mysterious — they're testable. Once you know which one is yours, the work is concrete.