Glossary · Measurement
Source Diversity
The variety of distinct domains and publishers an AI engine draws on when citing sources across a set of answers, as opposed to repeatedly citing the same handful of sites.
How it works
Source diversity is measured by counting unique cited domains across a batch of answers on a topic, rather than counting citations themselves. Low diversity means a small number of sites — a specific review platform, a couple of major publishers — are doing most of the work of shaping what the model says.
Why it matters
When diversity is low, getting coverage on the handful of sources the model actually relies on matters far more than broad-based publishing elsewhere. High diversity spreads the opportunity — and the risk — across a wider set of sites.
See more terms in the full glossary, or read the AI Visibility 101 guide for the full picture.