Open ChatGPT right now. Ask it: "What's the best [your category] tool?"
Whose name comes back?
If it's not yours, you have a problem most SEO tools won't catch — and one that's quietly eroding your top of funnel. Here's why it happens and what to do about it.
Why ChatGPT picks who it picks
LLMs don't rank like Google. They synthesize. When asked for a recommendation, ChatGPT pulls from three sources:
- Pretraining data — what was on the open web when the model was trained
- Retrieval at query time — live web search results when the model decides to look something up
- Brand associations — entity-level signals from Wikipedia, Crunchbase, and structured data
Your competitors are probably winning at least two of those three. The good news: every one of them is fixable.
The four signals that actually move the needle
1. Entity presence on the sources LLMs trust
- A Wikipedia page (where you're notable enough to qualify)
- A complete Crunchbase profile with funding, team, and category tags
- G2 / Capterra / Trustpilot profiles with real reviews
- Consistent schema.org markup pointing back to your canonical entity
2. Quotable content blocks
LLMs love definition lists, FAQs, and TL;DRs because they're easy to extract and cite. If your homepage is a hero image and three feature bullets, the model has nothing to quote.
Free audit
Grade your site free
See how you score across all 9 categories — in roughly 60 seconds. No signup.
Add a clean "What is [your product]?" block above the fold. Use FAQ schema. Write a one-sentence answer to the most-asked question in your category and put it where a crawler can find it in the first 500 bytes.
3. Third-party citations
Get mentioned by name on the publications and YouTube channels your buyers read. Even unlinked mentions count — LLMs ingest text, not just hyperlinks.
4. Bot access
Check your robots.txt right now. If it blocks GPTBot, ClaudeBot, PerplexityBot, or Google-Extended, you've taken yourself out of the running on purpose.
User-agent: GPTBot
Allow: /
How to measure it
Pick 20 prompts a real buyer in your category would ask. Run them through ChatGPT, Claude, Perplexity, and Google AI Overviews. Score how often you appear, in what position, and with what sentiment.
That's your AI visibility baseline. Re-run it monthly and watch the numbers move as you fix the four signals above.
If you'd rather automate the whole thing — measurement, baseline, and the gap analysis — that's exactly what SEOGrade's AI Citability category was built for. Run a free audit and see where you're losing the prompt war.