People also ask
In practice
Selection is where the strategy stops being generic. The candidate pool comes from the site’s own economics: pages where a buyer converts (pricing, integrations, alternatives), pages close to ranking (position 5–20 on searches that matter, where authority is the visible gap), pages whose absence from shortlist searches costs deals. Each candidate is scored on three questions: value (what is this page worth winning), distance (how far from page one, and what gap explains it), readiness (indexable, complete, current — a page failing readiness is disqualified no matter how valuable).
The output is a short, ranked list — usually a handful of pages per quarter, not twenty — because concentration is what makes a programme legible: a few pages moving visibly beats a dozen moving imperceptibly. The common failure is the opposite default: everything to the homepage, because it’s easy to pitch and easy to place — and easy placements to the homepage are the profile equivalent of calories without nutrition: numbers up, pages unchanged.
Selection also sets the placement mix per page (insertions for pages that need precision, guest posts for pages that need a story, mentions for pages whose category the AI layer reads) and the anchor approach — both of which inherit the page’s intent. The review rhythm: quarterly, against the gap data — pages that moved graduate out, new candidates enter from Search Console’s emerging queries, and the list stays a decision rather than a habit.
See also: One page or one hundred: why links point at URLs, Keyword-to-page mapping, Target URL readiness.
Checklist: Link Requirements.
