Knowledge graph
A structured database of entities and the relationships between them, used by search and AI engines to answer factual questions and to resolve which thing a name refers to.
In depth
What it really means
A knowledge graph stores facts as connections: this company sells this product, this product belongs to this category, this person founded this company. That structure lets an engine answer questions no single page states outright, by traversing the connections.
Google’s is the best known and powers knowledge panels. Wikidata is the open equivalent and gets read by nearly everything, which makes it an unusually cheap win for a free property most B2B brands have never touched.
How it works
- Entities are extracted from structured data, trusted databases and page text.
- Each is disambiguated against existing graph nodes.
- Relationships are recorded as typed edges between nodes.
- Confidence is scored from how many independent sources agree, then used at answer time.
Pros & cons
Pros
- Accurate graph representation makes engines describe you correctly by default.
- It works across every engine, since they draw on overlapping public sources.
- Wikidata and schema markup are free and directly editable.
Cons
- You submit signals, you do not write entries. Google decides what it accepts.
- Corrections take weeks to months to propagate.
- Notability thresholds keep smaller brands out of some graphs entirely.
Common mistakes
- Assuming schema markup alone creates a graph entry. It is one corroborating signal among several.
- Different founding years, headcount or descriptions across LinkedIn, Crunchbase and your about page.
- Ignoring Wikidata, which is editable, widely read, and empty for most B2B brands.
- Leaving sameAs out of your Organization schema, which is the link that ties your properties together.
Best practices
- Publish Organization schema with sameAs pointing to every profile you control.
- Make one canonical set of facts and apply it identically everywhere.
- Create or correct your Wikidata item with proper source references.
- Keep third-party databases current, especially after funding or a rename.
- Get your core facts stated on independent sources, since corroboration drives confidence.
FAQs
What is a knowledge graph?
A structured database of entities and their relationships that search and AI engines use to answer factual questions and work out which thing a name refers to.
How does the knowledge graph affect AI answers?
It supplies the confident factual layer. How you are represented shapes how engines describe you and which topics they associate you with.
How do I influence the knowledge graph?
Consistent facts everywhere, Organization schema with sameAs links, an accurate Wikidata item, current third-party profiles, and independent sources that corroborate the same details.
What is a knowledge panel?
The box of facts about an entity shown beside some Google results, populated from the knowledge graph. You can claim it and suggest edits, but Google controls what appears.
Keep reading
Related on LymLyt
Beyond LymLyt
Further reading
Want this working on your site?
We build the content behind the term, ranked in search and cited by AI.
Book a 30-min call →