Audit 17: Knowledge Graph Gap Filler

Surgically identify what Google's Knowledge Graph does NOT know about an entity, then generate the exact structured data, content, and cross-platform actions needed to fill those gaps.

Title & Description

What it is: A systematic audit that reverse-engineers Google's self-updating Knowledge Graph process to identify missing entity attributes and prescribe targeted fixes.

When to run it: When a Knowledge Panel is incomplete, missing attributes, or not showing at all. Also run proactively for any entity that should have a Knowledge Panel.

Patent Foundation: Google's Self-Updating Knowledge Graph patent (2018) — Google actively audits its own Knowledge Graph, comparing entity completeness against type norms, generating questions about gaps, searching for answers across tiered sources, and updating when confidence is high.

Patent & Research Foundation

Core Patent: Self-Updating Knowledge Graph (2018)

Google's Knowledge Graph does not passively wait for data. It actively:

Audits entity completeness against type norms
Generates questions about attribute gaps
Searches for answers across tiered source hierarchy
Updates the KG when confidence threshold is met

Source Trust Hierarchy

Tier	Sources	Trust Level
Tier 1	Official website (schema.org markup)	Highest
Tier 2	Wikipedia, Wikidata	Very High
Tier 3	Crunchbase, LinkedIn, BBB, industry databases	High
Tier 4	News articles, blog mentions	Medium
Tier 5	Social profiles, directories	Lower

Strategic Implication: Fixing a gap at Tier 4 is nearly worthless if the gap exists at Tier 1. Always work top-down through the trust hierarchy.

The 4 Principles of KG Gap Analysis

1. Completeness Over Presence

Having a Knowledge Panel is not the goal. Having a COMPLETE Knowledge Panel with every standard attribute populated is the goal.

2. Source Hierarchy Dictates Strategy

Fix the highest-tier gap first. Official website schema → Wikipedia → Wikidata → industry databases.

3. Reconciliation Before Attributes

If Google cannot confirm all references point to the same entity, adding attributes is wasted effort. Fix identity consistency first.

4. Machine Questions, Human Answers

The KG generates questions to fill its gaps. Anticipate those questions and answer them in structured, extractable formats.

Audit Methodology

Phase 1: Entity Identification

Determine what entity type Google currently recognizes:

[ ] Person (name, job title, employer, notable work)
[ ] Organization (name, industry, headquarters, founders)
[ ] LocalBusiness (name, address, phone, hours, category)
[ ] Place (name, location, geo coordinates)
[ ] CreativeWork (title, author, publisher, genre)
[ ] Product (name, brand, SKU, offers)

Check if entity is reconciled:

Search entity name in Google — does a Knowledge Panel appear?
If Knowledge Panel shows: entity is recognized
If no Knowledge Panel: entity may not be reconciled yet

Phase 2: Attribute Gap Analysis

Standard attributes for the entity type:

For Organization:

Name, website, headquarters, founded, founders, employees
Products/services, industry, description
Social profiles (LinkedIn, Twitter, Facebook)
Wikipedia page, Wikidata QID

For Person:

Full name, date of birth (if public), nationality
Occupation, employer, notable works
Education, awards, social profiles
Wikipedia page (if notable enough)

For LocalBusiness:

Name, address, phone, website
Hours, category, price range
Coordinates, GMB listing
Reviews, images

Gap identification:

[ ] List every standard attribute for entity type
[ ] Check which attributes appear in Knowledge Panel
[ ] Missing attributes = gaps to fill
[ ] Partial attributes (low confidence) = strengthen sources

Phase 3: Source Audit

For each missing attribute, identify:

Does the official website have schema markup with this attribute?
Does Wikipedia mention this attribute?
Does Wikidata have this property populated?
Do third-party databases (Crunchbase, LinkedIn) have this?

Priority fix order:

Gap at Tier 1 (website): Add JSON-LD schema
Gap at Tier 2 (Wikipedia): Edit Wikipedia article
Gap at Tier 3 (Wikidata): Add Wikidata property
Gap at Tier 4 (databases): Update Crunchbase/LinkedIn
Gap at Tier 5 (directories): Update citation listings

How to Run This Audit

Step 1: Knowledge Panel Audit

Open Google, search for the entity name.

If Knowledge Panel shows:
- Screenshot it
- List every visible attribute
- Note missing attributes vs. standard for this entity type

If no Knowledge Panel:
- Entity is not reconciled (or not notable enough)
- Focus on reconciliation first (Tier 1 + Tier 2 coverage)

Step 2: Schema Audit (Tier 1)

Check the official website for entity schema:

bash

# Check for existing schema
# Look in page source for: application/ld+json
# Or use Google's Rich Results Test

If Organization schema is missing or incomplete:

json

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "@id": "https://yourdomain.com/#organization",
  "name": "Full Legal Entity Name",
  "url": "https://yourdomain.com",
  "logo": "https://yourdomain.com/logo.png",
  "description": "Clear entity description",
  "foundingDate": "YYYY",
  "numberOfEmployees": {"@type": "QuantitativeValue", "value": "N"},
  "address": {
    "@type": "PostalAddress",
    "addressLocality": "City",
    "addressRegion": "State",
    "addressCountry": "US"
  },
  "sameAs": [
    "https://en.wikipedia.org/wiki/EntityName",
    "https://www.wikidata.org/wiki/QXXXXXX",
    "https://www.linkedin.com/company/name"
  ]
}

Step 3: Wikipedia & Wikidata (Tier 2)

Wikipedia:

Does a Wikipedia article exist for this entity?
If yes: Does it include all standard attributes?
If no: Is the entity notable enough for a Wikipedia article?

Wikidata:

Does a Wikidata item (Q-number) exist?
If yes: Are all standard properties populated?
Add missing properties directly in Wikidata

Wikidata Key Properties by Entity Type:

Property	ID	Description
instance of	P31	What type of entity is this?
official website	P856	Website URL
founded	P571	Date entity was founded
headquarters	P159	Physical location
CEO / leader	P169	Named executive
number of employees	P1128	Company size
LinkedIn	P4264	LinkedIn profile ID

Step 4: Cross-Platform Consistency Check

Identity consistency (reconciliation):

[ ] Entity name identical across: website, Wikipedia, Wikidata, LinkedIn, GMB
[ ] sameAs links connect all profiles to each other
[ ] @id on website schema creates a canonical entity URL
[ ] No conflicting information across platforms

Scoring

Score	Interpretation
90-100	Complete entity — Knowledge Panel should be fully populated
70-89	Minor gaps — Knowledge Panel present but incomplete
50-69	Significant gaps — Knowledge Panel likely missing key attributes
Below 50	Major work needed — Entity may not have Knowledge Panel

Entity Extraction Audit — Entity signals in content
Agent Rank Author Audit — Person entity signals
Local Geo-Relevance Audit — LocalBusiness entity signals

Audit 17: Knowledge Graph Gap Filler ​

Title & Description ​

Patent & Research Foundation ​

Core Patent: Self-Updating Knowledge Graph (2018) ​

Source Trust Hierarchy ​

The 4 Principles of KG Gap Analysis ​

1. Completeness Over Presence ​

2. Source Hierarchy Dictates Strategy ​

3. Reconciliation Before Attributes ​

4. Machine Questions, Human Answers ​

Audit Methodology ​

Phase 1: Entity Identification ​

Phase 2: Attribute Gap Analysis ​

Phase 3: Source Audit ​

How to Run This Audit ​

Step 1: Knowledge Panel Audit ​

Step 2: Schema Audit (Tier 1) ​

Step 3: Wikipedia & Wikidata (Tier 2) ​

Step 4: Cross-Platform Consistency Check ​

Scoring ​

Related Audits ​