Coverage — what the rank means at every depth

webrank tracks ~293 million domains and ranks 146 million of them (v3). But no traffic source on earth measures 146M domains — so the meaning of the rank changes with depth, and we'd rather you know exactly how.

Per-source depth

The three depth zones

• Head (top ~1M): every signal type is present — real-user data, DNS data, links. This

is where webrank is a true traffic rank and where it beats equal weighting by ~46% (methodology).

• Torso (~1M–10M): traffic evidence thins to the broader sources (host.io, Open

PageRank, Tranco's tail). The rank is popularity-driven but with wider error bars.

• Tail (10M+): only the link graph reaches here. webrank and pagerank necessarily

converge — both are link-derived, because links are the only signal that exists at this depth. A tail rank means "how visible is this domain on the web at all", not "how many visitors does it get".

The depth-normalized formula does this automatically: each source only contributes where it has coverage, so the composite degrades gracefully instead of pretending. The sources field in every API response tells you how many sources saw the domain — treat it as a confidence indicator.

Freshness

Daily sources refresh daily (rank updated by ~04:00 UTC); CrUX, Open PageRank, host.io and the Common Crawl graph refresh monthly. Frozen citable snapshots are cut on the 15th — see updates.

Related: sources · methodology · FAQ