Knowledge graph
Updated 2026-09-15
A knowledge graph is a structured store of entities and the relationships between them: this company builds this product, this person founded this company, this product belongs to this category. Answer systems lean on graph-like knowledge to keep facts consistent across a generated answer.
How it affects brand answers
When a brand is present in an engine's entity knowledge, answers about it become stable: the same description recurs, categories are assigned consistently, and related entities are named correctly. When it is absent or contradictory, answers fall back on whatever the retrieved documents say that day, which produces the shifting, sometimes wrong descriptions brands notice first. Publishing consistent structured data and consistent prose is how a brand contributes to that record.
Illustrative contrast (example, not a customer measurement)
Brand A is described in near-identical terms by three engines across ten samples. Brand B, same category, gets three different category labels in ten samples. The difference is rarely content volume; it is whether the entity facts agree wherever the engines look.
Common mistakes
- Treating graph inclusion as something you can buy or submit. It is earned through consistency and corroboration.
- Changing the product description with every campaign, which keeps the record unstable.
- Assuming markup alone builds the record. Independent sources describing you the same way carry at least as much weight.
Frequently asked questions
- Can we edit what an engine knows about us?
- Not directly. You control your own statements and your verifiable profiles; the record updates as engines re-read those and the sources that cite them.
- Why do two engines describe us differently?
- They draw on different sources and different entity records. Persistent disagreement usually points at an inconsistency in what the web says about you, not at one engine being broken.
