On-page SEO

Technical SEO handles infrastructure: whether bots can access, render, and index the site (crawling, rendering, speed, schema). On-page SEO handles each page's content: which search intent it covers, how the text is structured, its title/meta, how it's linked internally. Technical makes you accessible; on-page makes you relevant. You need both — a technically perfect site whose pages answer nothing stays uncited.

Cannibalization happens when two or more of your pages target the same search intent. Engines don't know which to show, so they split the signals between them and neither ranks well. The fix: map each page to a distinct intent, then consolidate (301 or canonical) the redundant pages into the strongest one.

RAG systems in AI engines don't extract whole pages, they extract passages. A descriptive H2 followed by a self-contained 2-4 sentence answer is a unit the model can lift and cite directly. A wall of unstructured text forces the model to guess where the answer starts and ends — and it usually picks another source instead.

Google rewrites them often, but not always — and when it keeps yours, that text decides the click. Meta descriptions are also used verbatim by other surfaces: share previews, aggregators, some AI interfaces. The cost of writing one well is small; the cost of letting an engine pick an unflattering fragment is a lost click.