gaash.ai

On-Page Optimization

On-page optimization is the work done inside a page so that a machine can tell what it is about, who published it and what it answers: the title, the headings, the body text, the internal links, the structured data and the metadata. It is the half of optimization a business controls entirely.

What a page has to carry

A title that names the subject. One h1, and a heading hierarchy underneath it that follows the questions a reader arrives with, each heading followed by the answer rather than by an introduction to it. Body text that states facts a machine can attribute: a price, an area served, a name, a time. Structured data that describes the business, not the website. One name, one address and one phone number, identical on every page. Internal links whose text says where they go. A meta description written from a sentence already on the page.

What the audit found on the page

This site fetched the homepage, contact page and about page of 9,483 small-business sites. The pages were not thin: the median homepage ran 880 words and only 3.9% were under 150. But 61.5% had no question-and-answer content of any kind, 50.7% had a heading phrased as a question with no answer isolated beneath it, and 43.1% were inconsistent about their own name, address or phone number across three pages. Only 17.8% had no schema.org markup, and most of the rest had markup that described the website, put there by the page builder. The words were there. The form was not. The full table is in what 9,483 small business websites serve to an AI crawler.

On-page work for a generative engine

Google’s documentation on AI features in Search says there are no additional requirements to appear in AI Overviews or AI Mode beyond the fundamentals that apply to Search. Aggarwal et al., “GEO: Generative Engine Optimization” (KDD 2024), tested source-side changes alone, adding statistics, quotations and citations to a page with the rest of the web held constant, and measured a change in how visible the source became in generated answers. Both point the same way: the page is where the work is, and the work is making each fact findable, attributable and stated once.

A common question

The audit found pages were not thin, so what was actually missing from them?

Form rather than words. The median homepage ran 880 words and only 3.9% came in under 150, yet 61.5% of the 9,483 sites carried no question-and-answer content at all, 50.7% had a heading phrased as a question with no answer set out beneath it, and 43.1% disagreed with themselves about name, address or phone number across three pages. Most of the markup that did exist described the website instead of the business. The facts sat on the page, and nothing about the page made any one of them easy to lift and attribute.

Related terms

Sources

No special requirements: Google’s documentation on AI features in Search. Source-side changes: Aggarwal et al., “GEO: Generative Engine Optimization” (KDD 2024). The audit figures are this site’s own; sample drawn per metro and trade, percentages describe the sample.