How Bad is Similarweb Data?
A genuine question
I recently tried out similarweb as a potential data source for my dissertation. My experience was troubling.
The title of this post is a genuine question: just how bad is similarweb data? Who trusts data from Similarweb, and why?
As a brief example to illustrate my concern, consider the “top 20” list. With any kind of representative sample of data, it would be fairly simple to generate an accurate list of the most popular websites in the United States.
According to Similarweb, the most popular websites in the United States in March 2022 were:
- google.com
- youtube.com
- facebook.com
- yahoo.com
- amazon.com
According to Comscore, which uses web browing data from a random panel of internet users in the U.S., the most popular webistes in March 2022 were:
- youtube.com
- google.com
- facebook.com
- amazon.com
- yahoo.com
So far, not so bad: Similarweb and Comscore report the same set of five websites, just in a different order.