What Is an Entity Gap in SEO and How to Close One

ENTITY GAP

An entity gap is an entity that the pages ranking for a query and the knowledge base around its subject associate with that query, and that the target page does not mention. It differs from a keyword gap, which compares strings, and from a content gap, which compares topics or pages; an entity gap compares the things a page talks about. Entity gaps come from four sources: ranking pages, knowledge base attributes, related questions and the query network. Finding them takes seven steps of extraction and counting, and reading the list matters more than producing it, because many missing entities belong to another page, to a competitor’s brand or to boilerplate. A gap closes with a sentence that relates the entity to the page’s subject, not with a repeated mention, and the entities a page needs depend on its page type. Closing every gap brings a page to the consensus and no further, so information gain still has to come from elsewhere. Holistic Radar found an entity gap on its own site, and the questions at the end cover tools and rankings.

What an entity gap is

An entity gap is a missing entity: a person, organization, concept, place or product that a search engine expects a page about a subject to mention, and that the page leaves out. The expectation comes from consensus. When eight of the ten pages ranking for “topical map” mention source context and the target page does not, source context is an entity gap of the target page, and closing it raises the page’s topical relevance to the query.

Target pageRanking consensustopic clusterpillar pagesource contextcentral entityattributetopical mapsharedEntities in the consensus and not on the target are its entity gap
Figure 1 · An entity gap is consensus the page lacks

The unit is the entity, not the word. “Source context”, “the publisher’s context” and “the site’s identity and monetization model” are three strings for one entity, and a page that uses any of them has no gap for it. A string comparison counts three gaps where an entity comparison counts none, which is why entity gaps need entity extraction and, ideally, entity linking to a knowledge base.

An entity gap is measured per page and per query. A page can have no gap for its primary query and several for a secondary query it also targets, and the same entity can be a gap on one page and correctly absent from another.

A page can also mention an entity and miss its attributes. A page about Core Web Vitals that names Largest Contentful Paint but never states its threshold has the entity and lacks the attribute value the consensus states. That partial case is an attribute gap inside an entity, and it closes the same way: with the triple the page omitted.

3 GAPSString comparisonsource contextthe publisher’s contextidentity and monetization model0 GAPSEntity comparisonOne entity: source contextAny variant counts as presentRequires entity linkingvs
Figure 2 · Strings differ; the entity is one

Entity gap, content gap and keyword gap

Three gap analyses compare a site with what searchers or search engines expect, and each compares a different unit.

AnalysisUnit comparedTypical findingFix
Keyword gapQuery strings a competitor ranks for“Competitor ranks for X, we do not”Target the query on a page
Content gapTopics, tuples or pages the topical map requires“No URL covers this attribute”Publish or deepen a page
Entity gapEntities inside one page’s main content“Page omits an entity the consensus includes”Add a sentence or a link
Keyword gapquery strings a competitor ranks forContent gaptuples and pages the map requiresEntity gapentities inside one pageThe map finds causes; page-level entity gaps find symptoms
Figure 3 · Three gap analyses, three units

The analyses nest. A content gap analysis runs against the topical map and finds five gap types, from missing tuples to missing information gain. An entity gap is what several of those types look like from inside one page: a missing tuple leaves the site without the entity entirely, and a question gap often leaves a page without the entity the question asks about. Running entity gaps page by page without the map finds symptoms; the map finds causes.

Where entity gaps come from

An entity gap list draws on four sources, and each catches entities the others miss.

  • Ranking pages. The main content of the top ten results for the query shows which entities the current consensus includes.
  • Knowledge base attributes. Wikidata and the Knowledge Graph list properties of the central entity, such as founder, inventor or part of, which name entities ranking pages may all have skipped.
  • Related questions. People Also Ask and related searches name entities searchers connect to the query, such as a tool, a standard or a competing method.
  • The query network. The queries of one cluster in the site’s own query network name entities the cluster’s page must define to answer them.
Ranking pagesnoisiestKnowledge basecleanestRelated questionsPAA, relatedQuery networkown clustersGap listper page
Figure 4 · Four sources of an entity gap list

Ranking pages are the noisiest source, because they include each competitor’s brand, products and template text. Knowledge base attributes are the cleanest, because each entity there already carries an identity, but they are incomplete for new or niche subjects. A list built from one source only reflects that source’s blind spot.

How to find entity gaps

Finding entity gaps takes seven steps, and the counting steps are scriptable:

  1. Choose one query and the one page on the site that owns it in the topical map.
  2. Collect the top ten organic results and strip each to its main content.
  3. Extract named entities and noun phrases from every reference and from the target.
  4. Link the extracted spans to knowledge base entities, so variants merge into one entity.
  5. Count, for each entity, how many references mention it.
  6. Keep entities above a consensus threshold, such as three of ten references, and subtract the entities of the target.
  7. Add knowledge base attributes and question entities the references missed.
Query + ownerpageTop tenmain contentExtractentities, nounphrasesLinkmerge variantsCountreferences perentitySubtractabove thresholdAddKB + questionentities
Figure 5 · Seven steps from query to gap list

Steps 2 to 6 are the script in the Python for semantic SEO article, which runs spaCy over references and target and prints the consensus terms the target lacks. Entity linking in step 4 is the step most tools skip. Without it, “Google Knowledge Graph” and “the Knowledge Graph” are two entities, and the gap list doubles. The Google Cloud Natural Language API returns a knowledge base identifier for entities it recognizes and a salience score between 0 and 1, which estimates how central each entity is to the document.

The consensus threshold sets the length of the list. At two of ten references the list includes one-off framings and runs to hundreds of entities; at seven of ten it shrinks to the entities no author omits. Three of ten is a working default for a first pass, and the strategist raises it when the list is longer than the page could absorb. Search results change, so a gap list is dated and rerun when a page loses positions, because a new entity entering the consensus is a common reason a complete page stops looking complete.

How to read an entity gap list

An entity gap list is a list of candidates, and most candidates are rejected. Five filters decide which gaps to close:

  • Ownership. An entity that belongs to another page’s entity-attribute pair goes to that page, and the target page links to it instead of explaining it.
  • Brand noise. Competitor names, product names and authors of the reference pages are gaps only when the query is about them.
  • Boilerplate. Entities from navigation, newsletter boxes and footers survived cleaning; they are not consensus.
  • Consensus strength. An entity in nine of ten references is expected; an entity in three may be one author’s framing.
  • Intent fit. A definitional page needs the entities that define; a procedural page needs the tools, inputs and outputs of the procedure.
Candidatesevery missing entityOwnershipbelongs to another pageBrand noisecompetitor namesBoilerplatenav and footer residueConsensus + intentstrength and page roleGaps to closeone sentence each
Figure 6 · Five filters reject most candidates

The ownership filter rejects the most candidates on a site with a topical map. A page about topical maps should name internal linking and define it in a sentence, but the explanation of link types belongs to the internal linking page. Closing that gap in full on the topical map page would create two pages answering one question, which is cannibalization by entity rather than by keyword.

How to close an entity gap

An entity gap closes with a sentence that states a triple relating the missing entity to the page’s subject. “A topical map is filtered by source context, the identity and monetization model of the site” adds the entity, defines it and connects it to the central entity in one sentence. Mentioning “source context” in a list of related terms adds the string and none of the relation.

CLOSING SENTENCEA topical map is filtered by source context, the identity and monetization model of the site.subjectrelationgap entityThe appositive after the comma defines the entity. A mention in a term list addsthe string; this sentence adds the triple
Figure 7 · One sentence adds, defines and relates the entity

Placement follows the entity’s role. An entity that defines the subject goes in the definition section; an entity that is a step, tool or input goes in the procedure; an entity that is a neighbouring concept goes in the closing bridge with a link to its own page. The extractive summary names the entity only when the entity is central enough that its absence would change what the page is about.

Mention count does not close a gap. Salience models weigh an entity’s grammatical role and position as well as its mentions, so one sentence with the entity as subject under a relevant heading raises salience more than five passing mentions in other sentences. Repeating an entity to match a competitor’s frequency reads as stuffing to a reader and adds nothing an extractor did not already have.

Entity gaps by page type

The entities a page needs depend on what the page does, so the same query consensus produces different gap lists for different page types.

Page typeEntities the consensus expectsCommon false gap
DefinitionThe class, the distinguishing attributes, the confused neighbourTools and vendors
ProcedureInputs, tools, steps, outputs, failure statesHistory and origin
ComparisonBoth compared entities and the dimensions they differ onEntities related to only one side
Service pageThe deliverables, the method, the proof and the price modelEducational definitions
01Definitionclass,distinguishingattributes02Procedureinputs, tools,steps, outputs03Comparisonboth sides,dimensions04Service pagedeliverables,method, proof
Figure 8 · What each page type needs

A false gap is an entity the references mention for a reason the target page does not share. Ranking pages for “what is a topical map” often list software vendors because their authors sell software; a definitional page on a consultancy site does not need them. Reading the gap list against the page type removes these before the ownership filter runs.

Entity gaps and information gain

Closing every entity gap brings a page to the consensus of the ranking pages and stops there. A page that mentions exactly what the top ten mention offers a search engine nothing it cannot retrieve from the top ten, which is the opposite of information gain.

Information gainfacts no reference hasConsensusentities most references shareEntity gapswhat the page still lacks
Figure 9 · Gaps reach consensus; gain goes above it

The two analyses do different jobs in order. The entity gap list removes the reasons a page looks incomplete; information gain adds the reasons a page deserves a position above the pages it now matches. Information gain comes from entities and facts the references do not contain: first-party data, a tested procedure, a case with figures, or a distinction the consensus blurs. An entity that appears in no reference and that the page can support with evidence is the reverse of an entity gap, and it is the more valuable finding.

The order matters because a page that is missing consensus entities rarely gets evaluated on what it adds. A page that is complete and adds nothing gets evaluated and loses on equal terms.

An entity gap on holisticradar.com

Holistic Radar found an entity gap on its own site in November 2026. Google Search Console showed impressions for “what is entity gap” at an average position near 69, and for “entity gap” near 81, while no page on holisticradar.com used the phrase “entity gap”. All 16 impressions over 90 days went to the content gap analysis page, which covered gaps without naming the entity gap as a measurement.

Impressionswhat is entitygap, ~69Site check0 pages use thetermOwnershipnew owner URLLinksgap analysis,Python
Figure 10 · How the site closed its own entity gap

The fix followed the filters above. Ownership placed the definition on a new page, this one, because the content gap analysis page owns the five gap types and the entity SEO page owns entities, attributes and relations. The content gap analysis page gained a sentence relating entity gaps to its gap types with a link here, and the Python article carries the extraction script. Each page now states one part of the subject and links to the others, and the query “what is entity gap” has one owner URL.

Entity gap questions

What is an entity gap in SEO?

An entity gap in SEO is an entity that most pages ranking for a query mention and the target page omits. Closing it means adding a sentence that relates the entity to the page’s subject, when the entity belongs on that page.

Which tools find entity gaps?

Entity gaps are found by any tool that extracts entities from the ranking pages and the target page and compares the sets. spaCy and the Google Cloud Natural Language API perform the extraction; a short Python script performs the comparison. Commercial content editors run the same comparison on terms, often without entity linking.

Does closing entity gaps improve rankings?

Closing entity gaps removes a reason a page looks incomplete for its query, and it improves rankings when incompleteness was the reason a page ranked below the consensus. It does not lift a complete page above its competitors; information gain does that.

Entity Gap in the semantic SEO topical map

Entity Gap is the page-level gap attribute of the content gap analysis entity inside the Holistic Radar topical map, and it sits in the outer section under the Know predicate.

The outer section holds the page because it defines a measurement without selling one, and its bridge passes session signals toward Semantic SEO Audits, where entity gaps are listed for every page of a client site.

→ Semantic SEO audits: entity gaps listed page by page against the topical map.

→ Content gap analysis: the five gap types an entity gap sits inside.

→ What is entity SEO: entities, attributes and relations across a site.

→ Python for semantic SEO: the extraction and comparison script.

Get my free Radar ScanFree Radar Scan