Analytics Are Not Always Perfect

Measuring internet traffic is not yet an exact science. Although as an industry we are getting closer, there are still discrepancies between certain analytics sources when calculating visitors for the same website.

In his article Facebook Vs. YouTube: Which Site Has More Visitors?, Martin Beck shows that when using two different data sources, Compete and ComScore, to answer the simple question, “Does Facebook or YouTube generate more unique monthly visitors?” he gets two differing answers.

Compete’s data showed YouTube received more unique U.S. visitors in June – 167.8 million compared to Facebook’s 166.4 million.

On the other hand, when looking at the numbers provided by ComScore, Facebook generated more unique monthly visitors in June, with 200.4 million compared to YouTube’s 198.7 million.

Beck believes this discrepancy in data is caused as a result of two factors:

  1. The data is taken from a smaller sample size and then averaged across the population. Thus, the data collected from that sample is not always going to be an exact representation of the entire country.
  2. Mobile visitors – if measuring total monthly visitors is not an exact science, calculating mobile visitors is even less of one. Those on mobile devices could be impacting the data.

The main takeaway from this article is to take search analytics with a grain of salt. We’ve gotten to the point where most of the time the numbers are going to be extremely close to the real story, but sometimes, such as this case, the data is not always perfect.