A traffic rank orders websites by how much traffic they receive, with rank 1 the most visited. Some sources publish an exact position, others publish a bucket, such as the top 10,000, which says only that a site falls inside that band. Because sources measure different visitors, ranks from two sources do not agree.
How ranks are produced
Sources estimate traffic from a sample: a browser panel, browser telemetry, data from internet providers, or a crawl. Each sample sees some visitors and not others, so a rank describes that population, not every visitor. A bucket is coarser still. It tells you the band a site is in, not its position, and a site that moves within a band does not show.
Using it, and misreading it
A rank is a rough proxy for the reach of a company's public website. It is used to segment prospects and to spot a site that climbs a band. It is a poor proxy for company size: a large business-to-business firm can run a quiet site, and a small consumer brand a busy one.
Three cautions apply. A rank is relative, so it moves when other sites move. A missing rank means too few visitors to measure, not zero visitors. And two sources should never be mixed in one comparison.
Traffic rank in Fokals data
The Web Traffic dataset gives the monthly traffic tier of each website: the smallest tier the site falls in, from 1,000 to 50,000,000, where a smaller tier means more traffic. It places each site in a band and tracks it from month to month. To list the sites in the top 10,000 for a month:
select domain, rank_bucket
from traffic_ranks
where month = :month
and rank_bucket <= 10000
order by rank_bucket, domain;In Market Series, the traffic up and traffic down families compare the latest month with the one before, or with six months earlier. The dataset sits in the announcements and scale dataset, and the data dictionary defines each field.
Related terms
A traffic rank is one attribute of firmographic data. Joining it to your own list takes domain matching, and its gaps are a question of data coverage. Trends in it are compared on a same-store cohort.
Frequently asked questions
What does a traffic rank of 10,000 mean?
In general, that the site is among the 10,000 most visited sites in the source's measured population, though a source that publishes buckets does not say where within the band. In the Web Traffic dataset, a tier of 10,000 means the site is outside the top 5,000 and inside the top 10,000, because each site gets the smallest tier it fits in. It is a position in one source's sample, not a count of visits.
Is traffic rank the same as traffic volume?
No. A rank orders sites against each other, while a volume counts visits. The same rank in two different months can stand for very different visit counts, because the whole field moves. Fokals delivers the monthly tier, which is the measure for segmenting sites and tracking their movement between bands, and it is read on its own scale, not as a visit count.
Why do traffic ranks differ between sources?
Each source measures a different sample of visitors in a different way. A browser-telemetry source, for example, counts only the users of one browser who have turned on usage reporting, so it describes that group, not all web users. Other sources use panels, internet-provider data or their own crawls. Compare ranks only within one source and one month, and treat a bucket as a band, not a position.
The queries and code on this page are examples to adapt. Test them in your own environment before you rely on them.