Session 7.1: What Is Entity Reconciliation?

Course → Module 7: Entity Reconciliation

Session 1 of 7

The Problem: Scattered Signals

Your business exists in dozens of places online. Your website says one thing. Your Google Business Profile says another. Directory listings, social profiles, review sites, press mentions, and government registrations each hold fragments of your identity. Google sees all of these fragments. Its job is to figure out which fragments belong to the same entity.

That process is entity reconciliation.

When reconciliation succeeds, all those scattered signals combine into a single, strong entity. When it fails, Google treats your business as multiple weak, disconnected fragments. You end up competing with yourself.

Entity reconciliation is the process by which search engines take signals from multiple independent sources and determine whether they all refer to the same real-world thing.

How Reconciliation Works

Google's reconciliation engine operates in stages. It collects data from crawled sources, extracts entity mentions, compares attributes across sources, and then clusters matches into a single entity node in the Knowledge Graph.

The engine relies on matching signals. Name, address, phone number, URL, logo, description, associated people, founding date, industry category. The more signals that match across sources, the higher Google's confidence that all sources describe one entity.

graph LR A["Website"] --> R["Reconciliation Engine"] B["Google Business Profile"] --> R C["Directory Listings"] --> R D["Social Profiles"] --> R E["Press Mentions"] --> R R --> G["Single Entity Node"] R --> F["Fragmented Entries"] style G fill:#222221,stroke:#6b8f71,color:#ede9e3 style F fill:#222221,stroke:#c47a5a,color:#ede9e3

The reconciliation engine is not a single algorithm. It is a pipeline. Google's Enterprise Knowledge Graph documentation describes the process: data ingestion, knowledge extraction (turning relational data into RDF triples), entity resolution (clustering records that refer to the same entity), and storage in the Knowledge Graph.

Reconciliation Signals

Not all signals carry equal weight. Some are strong identifiers. Others are weak corroborators.

Signal	Strength	Why It Matters
Business name (exact match)	High	Primary identifier across all sources
Website URL	High	Unique canonical reference point
Phone number	High	Hard to fake, easy to cross-reference
Physical address	High	Geo-anchors the entity to a location
sameAs links (schema.org)	High	Explicit declaration of identity equivalence
Logo / brand imagery	Medium	Visual matching across platforms
Business description	Medium	Semantic similarity check
Associated people	Medium	Person entities linked to organization
Industry / category	Low	Contextual corroboration only
Founding date	Low	Disambiguation aid for common names

What Successful Reconciliation Looks Like

When Google successfully reconciles your entity, the results are visible. Your Knowledge Panel pulls information from multiple sources and presents it as one unified profile. Your Google Business Profile, website, social links, and reviews all appear connected. AI search tools like Perplexity and ChatGPT can answer questions about your company with accurate, consolidated facts.

When reconciliation fails, the symptoms are equally visible. No Knowledge Panel. Branded searches return unrelated results. AI tools give conflicting or incomplete information. Your Google Business Profile exists in isolation, disconnected from your website's structured data.

graph TD subgraph Success["Successful Reconciliation"] S1["Website schema"] --- SE["Unified Entity"] S2["GBP data"] --- SE S3["LinkedIn profile"] --- SE S4["Wikidata entry"] --- SE end subgraph Failure["Failed Reconciliation"] F1["Website schema"] --- FE1["Entity Fragment A"] F2["GBP data"] --- FE2["Entity Fragment B"] F3["LinkedIn profile"] --- FE3["Entity Fragment C"] F4["Wikidata entry"] --- FE4["Entity Fragment D"] end style SE fill:#222221,stroke:#6b8f71,color:#ede9e3 style FE1 fill:#222221,stroke:#c47a5a,color:#ede9e3 style FE2 fill:#222221,stroke:#c47a5a,color:#ede9e3 style FE3 fill:#222221,stroke:#c47a5a,color:#ede9e3 style FE4 fill:#222221,stroke:#c47a5a,color:#ede9e3

The Compounding Effect

Reconciliation is not just about Google knowing you exist. It is about signal compounding. A single directory listing with your NAP carries a small amount of trust. Twenty directory listings, all reconciled to the same entity, carry twenty times the trust. But only if Google connects them.

If those twenty listings have slightly different names, different phone formats, or inconsistent addresses, Google may treat them as evidence of multiple entities rather than corroboration of one. The signals cancel out instead of compounding.

This is why entity reconciliation is the bridge between having an online presence and having entity authority. You can do everything else right (build schema, optimize GBP, create citations) and still fail if the reconciliation layer breaks.

Entity infrastructure without reconciliation is a collection of disconnected signals. Reconciliation is what turns individual signals into compounding authority.

Assignment

List every place your business exists online: website, GBP, directories, social profiles, review sites, press mentions. For each, write the exact business name and URL as it appears there. Circle any inconsistencies in name, address, phone, or URL. Each inconsistency is a potential reconciliation failure point. Count the total number of discrepancies.

What Is Entity Reconciliation?