Statistical Obstacles Part I: Spam!

          During the initial days of launching the website I logged into our Google Analytics account and I got really happy to see visits from China. The day after that, I saw many visits from Russia. My reaction was: “Okay ... this looks fishy”. Apparently we were getting spammed.

          Accurate statistics are crucial to us and having spammed statistics made me question the effectiveness of our approach.

          I read many explanations and finally understood what’s wrong. I did some optimization to our Google Analytics account but we were still getting spammed.

          Being too busy to try again; I decided to postpone fixing the statistics until after graduation. Once I got the time to do so, I searched for a solution and I found a great article in which the author shares comprehensive and well explained steps to solve all the statistical issues we were facing. He also offers his services to do it all for a reasonable price.

          That was, probably, the fifth attempt to fix the statistics. I was overwhelmed enough to take the shortcut of hiring someone to do it; however, I decided to try one last time before hiring someone to do it.

          It worked like a charm, without hiring someone to do it, and we were so happy that we can finally get accurate statistics. Little did we know, the second statistical obstacle (the topic of the next blog) was even more difficult to solve.

          Here is a quick look at an overview of our statistics with and without spam:

Spam & actual worldwide sessions for the duration (24/4/2015 – 12/8/2015)

Spam & actual worldwide sessions for the duration (24/4/2015 – 12/8/2015)

Actual worldwide sessions for the duration (24/4/2015 – 12/8/2015)

Actual worldwide sessions for the duration (24/4/2015 – 12/8/2015)

Orange: Sessions with spam. Blue: sessions without spam.

Orange: Sessions with spam. Blue: sessions without spam.

From August 8 spam was blocked (all sessions = optimized sessions)

From August 8 spam was blocked (all sessions = optimized sessions)

 

Author: Yazan Aljohani