Normalized Discounted Cumulative Gain
NDCG evaluates a ranking by considering relevance and giving more weight to higher positions. It is normalized so queries with different candidate sets can be compared.
What NDCG means in GEO
These metrics come from experimental evaluation. They help when reading GEO papers and designing more rigorous tests.
Why it matters
Define the relevance scale and ranking cutoff before comparing systems.
Review the unit of analysis, sample, baseline, confidence interval and treatment of ties or missing items.
How it connects to other concepts
NDCG belongs to research metrics. It is best analyzed alongside related terms because AI visibility depends on several stages and signals rather than one isolated optimization.
Common mistake
Turning an improvement on a closed benchmark into a universal ranking promise for commercial products.
Turn these concepts into metrics
Bee LLM tracks mentions, citations, position, sentiment and competitors across major AI engines.