The notebookInside the ranking system
Note 11 of 203 min read

What are Google ranking signals?

Google can use information it already has about pages and sites. A technical field name doesn't reveal its ranking effect.

Previously: Does domain age affect SEO? Some information can be stored before the next search happens.

Worked example

Before the search and during the search

Distinguish stored information from the query being processed now.

The system does not start from nothing

  1. Before the query

    Information about backlinks, topics and past searches is already stored.

    Stored priors
  2. A new query arrives

    Words, location and time identify possible answers.

    Query understanding
  3. Pages are compared

    The system can use both page information and query-specific information.

    Ranking

What to noticeExisting links and search history can be available before a new query. The page still needs to match and compete for that query.

The categories follow the reference model. Private weights and exact field use remain unknown.

Ranking signals are pieces of information a search system can use to compare pages. They can describe the page, its links, the site, the query or previous searches and clicks.

Some information is collected or calculated before someone searches. The model calls this stored priors. Here, that means information already available to the system.

Check the site and the page

A search system may already have information about a site’s backlinks, topics and search history. It can use that information when comparing a page with other possible results.

Each page still has its own subject and links. A site that ranks well for bike repairs doesn’t automatically rank every new page it publishes.

When investigating a ranking change, check whether it affects several pages or only one. Then check the exact queries for each affected page in Search Console.

What the technical labels tell you

The model includes labels such as NSR, site-pr and pnav. These name possible stored information discussed in the reference. The names alone don’t tell us how Google uses them today or what weight they have.

For your own site, work on what you can check: pages that clearly target a query, relevant backlinks, internal links and organic clicks from related searches.

An SEO tool’s score is its own estimate

Google Search Console reports average position, impressions and clicks. A backlink tool reports its own estimates of links and authority. These measures answer different questions.

Save the tool name and definition alongside a score. Combining several estimates into a planning score doesn’t make it Google’s ranking formula.

Choose an edit for a clear reason

An author bio tells readers who wrote the page. Structured data labels information for search features. Neither creates third-party authority, and this model assigns structured data no ranking weight.

Before making an edit, write down what you expect it to help. For example: “This repair guide earns clicks for chain-repair queries. I’ll link from it to the chain-replacement page and compare that page’s rankings afterward.”

Keep this in your notes

Google can use information it already has about pages and sites. A technical field name doesn't reveal its ranking effect.

The next question

Spam checks can also affect whether a page appears and where it ranks. Continue with how does google detect spam?