The gap between looking good and reading well
Your site passes visual QA. The typography is sharp, the layout is polished, and the performance scores are green. None of that tells you whether a search crawler, an LLM scraper, or an AI agent can actually read and retrieve your content.
Machine legibility is a structural property, not a visual one. It lives in the HTML below the design layer — in heading hierarchy, semantic element selection, DOM depth, and the presence or absence of structured data that makes your content machine-attributable.
Why heading hierarchy is not a style decision
Headings are the primary structural signal AI systems use to build a document outline. When a heading level is skipped — H2 to H4, H1 to H3 — the document outline contains phantom sections. Content after the skip is placed under the wrong conceptual parent in the AI’s structural model.
This is not an edge case. It is the most common single failure mode across every site we have diagnosed.
What structured data actually does
Schema.org JSON-LD is the explicit machine-readable semantic layer. Without it, AI systems must infer page topic from heading structure, meta description, and body content alone. That inference is less reliable than an explicit declaration.
A page without Organization schema is a page where your brand entity is not definitively linked to your domain. A blog post without datePublished cannot be placed in temporal context. These are not compliance checkboxes. They are the signals AI systems use to decide whether your content is worth retrieving.
The one question worth asking
When a buyer’s AI agent is asked to find vendors for a problem you solve — can your site answer that query from its HTML alone, without requiring the agent to execute JavaScript, wait for hydration, or guess at heading context?
If the answer is no, the structural layer needs work before anything else.