GoogleAnalytics suspicious referral traffic in depth analysis

I am a large electricity supplier website operators, often to analyze the flow of data from different sources. As for statistical tools, it is the electricity supplier website favorite analysis weapon GoogleAnalytics!

When using

GoogleAnalytics tools, see "referral traffic",,, often see such sources, however, we can not do without e-mail advertising, recently EDM e-mail marketing, this group of mysterious flow from what is what


according to the author’s 9 years of experience in the electricity supplier website analysis, step by step to make in-depth analysis……

first, I through direct observation, found this batch of recommended address are some well-known examples include mailbox, QQ mail, 163, 126 etc..

the second step, through the deep layer, then use GoogleAnalytics directly on the jump function, also did not find what is special, jump to the e-mail address, just back into the sub pages, and with a string parameter, the QQ mailbox is /cgi-bin/mail_spam. A colleague of this address specifically checked Baidu, concluded that it is an e-mail address. But I think it is not so simple, because some data from behind, found with the usual referral sources is not like


third step, the flow of user data performance.

here, there is an amazing discovery, this source of visitors, every visit to the page is very high, reached 8.1, nearly 7 minutes long when browsing. See here, I can have judgment: This is very likely to come from direct traffic! Browser favorites or favorites of the network access, direct traffic. In the country to do a good job in the electricity supplier Xinhua bookstore website only direct flow, in order to achieve such a high browsing page!

sum up, we in the usual traffic analysis, to want to see more, careful analysis, often can see a lot of people’s attention to unusual traces.