Extension of our website sample set

Posted by Matthias Gelbmann on 31 March 2020 in News

All our statistics used to be based on the Alexa top 10 million sites. Starting with April 2020, we additionally include sites that are in the Tranco top 1 million list. Tranco combines four popularity lists of websites into one ranking: Alexa, Cisco Umbrella, Majestic and Quantcast. Since these providers use quite different methods to measure popularity, there are actually a non-negligible number of sites that are in the Tranco top 1m list, but not in the Alexa top 10m.

Why do we limit out sample set at all?

We are often asked, why we do not simply use all the domains on the web for our statistics. The reason is, that we are convinced that this would make our data less useful. There are millions and millions of domains registered, that do not show any useful content. The domains may be registered by domain hoarders, they may be used to be part of linking schemes, they may be parked domains, planned to be used at a later stage. These sites have in common that hardly anyone ever visits them. They mostly have very little content, if any, and that content is often just ads or just a plain page. The technology used by such sites does in no way reflect any trends in the usage of website creation technologies. If we counted them, they would just pollute our data.

With the inclusion of the Tranco list, we extend our scope somewhat, without the risk of deteriorating the value of our statistics.

Share this page