Why Most SaaS Homepages Score Low on AI Visibility

I wanted to know how "AI-visible" typical SaaS homepages actually are — not in theory, but measured. So I ran a batch of them through the same AEO/SEO/GEO checker we run on this site and looked at where they consistently fell short.

Methodology

The checker looks at a fixed set of rule-based signals: presence and validity of Schema.org JSON-LD, whether robots.txt allows known AI crawlers (GPTBot, ClaudeBot, PerplexityBot), heading structure (single H1, no skipped levels), and whether pages expose freshness signals like a dateModified field.

The Common Failure Pattern

The single most frequent issue wasn't missing content — it was missing structured data. Plenty of homepages had good copy and clear value props, but no Organization or Article schema at all, which means there's nothing machine-readable for an AI system to lift a clean answer from.

The second most common gap: robots.txt files that either explicitly blocked AI crawlers or simply never mentioned them — which functionally means "unspecified" rather than "allowed."

The Worst Case Pattern

The lowest-scoring sites shared three traits: no JSON-LD schema anywhere on the page, a robots.txt that blocked GPTBot outright, and no Open Graph tags — so even social shares of the page rendered with no title or preview image.

What the Best-Scoring Sites Did Differently

The highest scorers had valid Article and Organization schema, visible dateModified fields that were actually kept current, an llms.txt file at the root, and a robots.txt that explicitly named and allowed the major AI crawlers rather than leaving it ambiguous.

Why This Matters

There's a reasonable, increasingly common theory that AI crawlers weight freshness signals when deciding what to cite — content with no update date, or a stale one, gets treated as less trustworthy than actively maintained pages. It isn't a confirmed, published ranking rule the way something like a canonical tag is, but it lines up with what we saw: sites that hadn't touched their schema in over a year skewed toward the bottom of the results.

The practical takeaway isn't complicated: if a homepage has no structured data and no crawler access, it's not that it ranks poorly with AI search — it's that it's largely invisible to it. That's a fixable problem, and usually a fast one.