Online shops live from product pages that Google can reach, index and rank, and most shops do not know how many of theirs it cannot. Knowledge-Graph-SEO analyses and optimizes a shop with agents. It crawls the shop the way Googlebot does and stores every page, link, zone and finding as a knowledge graph, so the whole shop structure becomes something an agent can query. It grades that graph against Google's own rules. Agents with memory then read it to find the causes, plan and prioritise the optimizations, carry them through with the shop owner and measure the effect on the next crawl: more products that reach Google, a stronger internal link structure, and a plan whose success is proven, not assumed.
What follows is one example: a single category page of a shop with its products. Shop, brand and product names are hidden; every number is real.
form / JavaScript · pages 2–107 noindex, canonical → page 1 · zone summary: slider 11.9 unique links per page206 reached · 1,245 at depth 0 · 688 with ≤ 1 source · 856 links → 4xx/5xx<a href> · each pagination page its own canonical · link every product that should be indexed · 4xx targets count as goneEvery count on this page comes from one stored analysis of Category A (snapshot 2026-10-27, compared with 2026-04-28). Shop, brand and product names are hidden. Values after the plan are projections and are labelled as such.