The 10-Signal Test
AI search engines don't rank pages. They decide, in the moment, whether to include you in an answer. That decision depends on whether the AI can understand your page, trust your content, and confidently recommend you.
I've distilled this into 10 signals based on what we've observed testing hundreds of pages against ChatGPT, Perplexity, and Gemini. For each one, I'll show you what good and bad looks like so you can score yourself honestly.
Grab your landing page URL. Let's go through them.
Signals 1-5: Can the AI Understand You?
Signal 1: Clear value proposition in text
Open your landing page and look at the first 200 words of visible text content. Does it clearly state what you sell, who it's for, and what problem it solves?
GOOD
"Invoice automation for freelance designers. Create, send, and track invoices in under 60 seconds."
BAD
"Unleash your potential. The platform that works as hard as you do."
Signal 2: Product category stated explicitly
Does your page name what category you're in? Not implied through context clues — actually written in text. "Email marketing platform." "Project management tool." "CRM for real estate agents."
I've seen pages for genuinely good products where the AI described them as "some kind of business tool" because the page never said what category it belonged to. The founders assumed visitors would figure it out from context. AI doesn't work on vibes.
Signal 3: Semantic HTML structure
View your page source. Is there a clear <h1> followed by logical <h2> and <h3> hierarchy? Or is it a soup of <div> tags with styling classes doing the visual work?
AI crawlers rely on heading hierarchy to understand content structure. A page with proper semantic HTML gets parsed accurately far more often than a page where "headings" are just big, bold <div> elements.
Signal 4: Text-based content (not image-dependent)
Disable images in your browser and look at your page. Does the core message still come through? Or does your page become a skeleton of "Get Started" buttons with no substance between them?
This is one of the most common failures I see. Beautiful pages where all the value proposition, feature descriptions, and differentiation live inside designed graphics. To AI crawlers, those pages are almost empty.
Signal 5: Descriptive headings that match questions
Look at your <h2> tags. Do any of them match questions a buyer would ask? "How much does [Product] cost?" "What tools does [Product] integrate with?" "Who is [Product] built for?"
GOOD
"How TaskFlow Works With Your Existing Stack" — directly answers integration questions
BAD
"Integrations" — a label, not an answer
Signals 6-8: Does the AI Trust You?
Signal 6: Specific, verifiable claims
Count the claims on your landing page. Now count how many include a specific number, a named source, or a link to evidence. If that ratio is below 50%, you have a credibility problem — not just with AI, but with human visitors too.
"Saves teams an average of 12 hours per week (based on 200-customer survey)" is something an AI can cite confidently. "Saves tons of time!" is something it'll ignore.
Signal 7: Structured data present
Check your page source for JSON-LD. Do you have Organization schema? Product schema? FAQ schema? You can use Google's Rich Results Test to check quickly.
Structured data acts as a clean summary for AI crawlers. It's the difference between handing someone a well-organized brief and asking them to figure things out from a brochure. Pages with complete structured data get described more accurately by AI models — we've confirmed this across dozens of tests.
Signal 8: FAQ section with real questions
Does your page have an FAQ section? If so, are the questions ones a buyer would actually ask, or are they softball setup questions for marketing copy?
"How does [Product] compare to [Competitor]?" is a real question. "Why is [Product] the best choice?" is marketing wearing a question mark. AI models can tell the difference, and they heavily favor genuine Q&A content because it maps directly to how people query AI search.
Signals 9-10: Can the AI Find You?
Signal 9: AI crawlers not blocked
Check your robots.txt file. Are GPTBot, ClaudeBot, and PerplexityBot allowed to crawl your site? This is a binary yes-or-no signal, but it's surprising how many sites accidentally block AI crawlers — either through overly aggressive bot blocking or default settings in security plugins.
If the crawlers can't see your page, nothing else on this list matters.
Signal 10: Consistent information across the web
AI models don't just read your landing page. They synthesize information from multiple sources — your site, review platforms, social media, third-party articles. If your landing page says you're an "email marketing platform" but your G2 profile says "marketing automation suite" and your LinkedIn says "growth tool," the AI gets confused about what you actually are.
Google yourself. Check your profiles on G2, Capterra, Product Hunt, Crunchbase, LinkedIn. Is the description consistent? Are the features accurate? Outdated profiles with wrong information actively hurt your AI visibility.
Score Yourself
Give yourself one point for each signal where you honestly pass. Here's what your score means:
8-10: Your page is well-positioned for AI search. Focus on monitoring and iterating.
5-7: You have a solid foundation but gaps that likely keep you out of AI recommendations. The fixes are usually straightforward — structured data, an FAQ section, more specific copy.
3-4: Significant work needed. Start with Signals 1, 2, and 7 — they give the most impact for the least effort.
0-2: Your page is essentially invisible to AI search. But that also means there's massive upside from relatively simple improvements.
Most pages I evaluate land in the 4-6 range. The good news: every signal on this list is fixable, and most of the fixes also improve your conversion rate for human visitors.
Want a more detailed assessment? Run your page through our Landing Page Analyzer to check your content quality across all the dimensions that drive both conversions and AI visibility. The same factors that make humans trust your page are the factors that make AI cite it.