I Tested My Product in ChatGPT and It Didn't Recommend Us. Now What?

You typed your product name into ChatGPT. It didn't mention you. Maybe a competitor showed up instead. Maybe the model listed your whole category but skipped you entirely.

Now you're wondering: is this a real problem, or am I panicking about nothing?

Here's the honest answer: it depends on whether AI search is sending buyers to your category. If it is, and you're not in the results, that's a fixable gap. If buyers in your category aren't using AI search yet, you have lead time to get structured correctly before they do.

This post walks you through three things: how to test whether your product is actually being skipped, why it happens, and the 80/20 fix you can run in the next 15 minutes.

First: Is this actually hurting you right now?

Before you fix anything, run the test. Three prompts, three models, 10 minutes.

Prompt 1: The category query
Open ChatGPT (GPT-4o), Perplexity, and Claude. In each one, type:

"What are the best tools for [the job your product does]?"

Use your buyer's language for the job. Not your product's marketing language. If you sell project management software to agencies, type: "What are the best tools for agencies to manage client projects?" not "What are the best project management platforms?"

Prompt 2: The ICP scenario
In each model, type:

"I'm a [your buyer's role] at a [company type] looking for [specific outcome]. What do you recommend?"

Example: "I'm a Head of Marketing at a B2B SaaS company looking for a way to know if our brand shows up in AI search. What do you recommend?"

Prompt 3: The brand check
In each model, type your product name directly:

"What do users say about [your product name]?"

Track three things for each response: (1) does your brand appear at all, (2) if it does appear, is the description accurate and positive, (3) which competitor gets the recommendation instead of you.

If your brand doesn't appear in a single response across all three prompts and three models, you have a confirmed citation gap.

Why does this happen?

AI models don't pull from a registry. They generate responses based on what they learned during training and, for real-time models like Perplexity, what they can retrieve from the web right now.

If your product pages don't clearly signal what problem you solve for whom, AI systems have weak evidence to cite you. The most common reasons a product gets skipped:

Missing structured data. FAQPage schema and Product schema are the clearest signal you can give an AI retrieval system. Without it, your page looks like a wall of text with no clear question-answer relationship.

Language mismatch. Your marketing copy uses product language ("AI-powered," "enterprise-grade," "seamless"). Your buyers use job language ("track client deliverables," "know if our brand shows up," "reduce manual reporting"). If your page doesn't use the language from the second column, AI systems can't match it to buyer queries.

No third-party signal. AI systems weight pages that are mentioned by other credible sources in the same category. If three respected sites in your space have mentioned your competitor but not you, the model has stronger evidence for the competitor.

Canonical URL gaps. If you have multiple pages competing for the same topic, AI systems can split signal across them and end up citing neither clearly.

The 80/20 fix: structured data on your top three pages

The single highest-leverage fix for most B2B SaaS companies is adding FAQPage structured data to their top three buyer-facing pages.

Here's exactly what that looks like.

Before (no structured data):

When a buyer typed "What are the best AI visibility tools for B2B SaaS?" into Perplexity in January, one of our test accounts got this response: [competitor named], [competitor named], and [competitor named]. No mention of the test brand despite having a relevant product page.

After (FAQPage schema added, 7 days later):

Same prompt, same model: "For B2B SaaS companies specifically, [test brand] offers an AI visibility audit that tests your brand across 40+ buyer query variations..." Citation appeared in two of three model responses within 14 days.

The schema that drove the lift looked like this (simplified):

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [{
    "@type": "Question",
    "name": "How do I know if my B2B SaaS brand shows up in AI search?",
    "acceptedAnswer": {
      "@type": "Answer",
      "text": "Run three buyer prompts in ChatGPT, Perplexity, and Claude using your buyer's job language, not your product's marketing language. If your brand doesn't appear in any response across all three models, you have a confirmed citation gap..."
    }
  }]
}

The questions in your FAQPage schema should use the exact language your buyers type when searching. Not the keywords you rank for in Google. The language they use in Slack, Reddit, and when they describe the problem to a colleague.

How to write the right questions:

Go to Reddit, IndieHackers, or a Slack community where your buyers hang out. Search for the problem your product solves. Copy the exact phrases people use when describing the problem, not when describing the solution. Those phrases become your FAQ questions.

What to do right now (15 minutes)

If you have confirmed you're being skipped in AI search, here's the 15-minute action:

1. Pick your three most important buyer-facing pages (usually: your homepage, your main product page, and your highest-traffic blog post about the problem you solve).
2. For each page, write three questions your buyer would actually ask when searching for a product like yours. Use their language, not yours.
3. Add a FAQPage JSON-LD block to each page's head section. Most CMSs let you add this in a custom code field without a developer.
4. Come back in 7 days and run the same three prompts from the test section above.

If you're not sure which pages have the biggest citation gaps, or which questions to write for each one, the AI Visibility Self-Audit runs this diagnostic for your specific site in about 5 minutes. It shows you which pages are getting cited, which are being ignored, and where the structured data gaps are. It's free. No email required.

The part nobody talks about

Most of the advice about "AI SEO" focuses on creating new content. More blog posts. More topic clusters. More content velocity.

The brands getting cited in AI search right now are not the ones with the most content. They're the ones whose existing pages most clearly signal what they answer and for whom.

You probably already have pages that could be getting cited. The gap is structural, not volume. Adding FAQPage schema to your three best pages and using buyer language in the questions takes less time than writing a new piece of content. And it compounds: once an AI retrieval system has clear evidence to cite you, it uses that evidence across multiple query types.

Fix the structure first. Then add volume.

If you already know your pages have structured data gaps and want a comprehensive audit across 40+ buyer query variations on ChatGPT, Perplexity, Claude, and Gemini, the LLMRadar Audit ($197) delivers the full report within 48 hours.