A page targeting a topic lacks many of the multi-word phrases, representing real subtopics, that recur across the pages currently ranking for that topic. Google's phrase-based indexing patents describe indexing and ranking documents by related phrases that predict one another. Google has not confirmed use, so this is a proxy; missing subtopic phrases usually also means missing subtopics for readers.
Why does topic page omits the related phrases that recur across top-ranking pages (phrase-based indexing) matter for AI search?
Google's phrase-based indexing patents describe indexing and ranking documents by related phrases that predict one another. Google has not confirmed use, so this is a proxy; missing subtopic phrases usually also means missing subtopics for readers.
How do you fix it?
Cover the missing subtopics with real content, or link to supporting pages that cover them. Do not insert phrase lists; the companion patent treats excessive related phrases as a spam signal.
How do you find it on your site?
- Collect the top 10 ranking URLs for the page's main query and extract their main content.
- Extract two to four word phrases and keep those appearing on at least half of the ranking pages.
- Remove navigational and generic phrases, leaving subtopic phrases.
- Check which remain absent from your page and flag it if a meaningful share of subtopics is missing.
Which ranking and citation factors does it touch?
Written for the Phrase-based indexing page (evidence: Patent only). Google has not documented this system in a way a site can test directly, so this check measures a proxy for it, not the system itself.
What is the impact?
Indirect. Estimate only.
What is the evidence?
US7536408B2 and US7603345B2 (Google Patents), inventor Anna Lynn Patterson.
Sources
Frequently asked questions about topic page omits the related phrases that recur across top-ranking pages (phrase-based indexing)
What does "Topic page omits the related phrases that recur across top-ranking pages (phrase-based indexing)" mean?
A page targeting a topic lacks many of the multi-word phrases, representing real subtopics, that recur across the pages currently ranking for that topic.
How do you fix topic page omits the related phrases that recur across top-ranking pages (phrase-based indexing)?
Cover the missing subtopics with real content, or link to supporting pages that cover them. Do not insert phrase lists; the companion patent treats excessive related phrases as a spam signal.