Ranking Factor

A ranking factor is any signal a search engine reads from a page, a site or the web around it and uses to order results for a query. The term names an input to ranking; it says nothing about the weight the engine gives it.

What Google says about its own

Google describes systems rather than factors. Its guide names BERT and neural matching for understanding what a query and a page mean, PageRank for reading how pages link to each other, passage ranking for finding the relevant section of a long page, and a freshness system, a reviews system, a site diversity system, spam and deduplication systems and several more. That is the disclosure. Lists of individual factors published elsewhere are inferences from what correlates with rank, not statements from the engine.

Ranking factors and generated answers

An assistant with retrieval does not rank pages. It searches, so a search engine’s ranking decides which pages it sees. Then it reads, and what it cites depends on whether a passage answers the question on its own. Then it writes, and which business it names depends on whether the pages identify one clearly. A ranking factor acts on the first step only. The second and third have their own conditions, and this site measures their outcome rather than guessing at weights; the methodology page describes how.

A common question

If an assistant does not rank pages, do ranking factors still matter at all?

They matter at one step. An assistant with retrieval searches first, so a search engine's ordering decides which pages it ever sees, and a ranking factor acts there and nowhere else. Then it reads, and what gets cited turns on whether a passage answers the question on its own; then it writes, and which business it names turns on whether the pages identify one clearly. Those later conditions belong to themselves, which is why this site measures the outcome instead of guessing at weights.

Sources

Named systems: Google, a guide to Search ranking systems.