Use Google Analytics to identify false traffic (II.)

This is the "use of Google Analytics to identify false traffic," the second part of the site to continue to analyze the false traffic. In the first article, we introduced 7 ways to discern false traffic. The traffic is analyzed from 24-hour traffic distribution, visitor's geographical distribution, network attribute and so on, and the false traffic is identified by means of contrast and subdivision. In this article, we will expand the method of discriminating false traffic, by comparing the difference between false flow and real traffic, we introduce 4 kinds of methods to distinguish false traffic from the angle of visitor behavior.

　　

　　The characteristics of false flow and real flow

In the analysis of false flow, the first simple description of the false flow and real traffic characteristics, understand the characteristics of these two traffic can help us quickly find the site of false traffic shadow, and further separate it. Here's a look at the characteristics and differences between these two types of traffic.

　Characteristics of FALSE traffic:

Purpose: The production of false traffic must be related to a particular purpose.

Regularity: Specific purpose leads to false traffic must have special rules.

Real Flow characteristics:

Nature: True flow in the various dimensions of the performance must be natural.

Diversity: Netizens ' preferences vary, and behavior must be diverse.

After understanding the characteristics of the two types of traffic, we can start to analyze the site traffic, with nature and diversity of access behavior as a principle, to find those who have "regular" false traffic.

　1 single page Refresh analysis

Single page refresh refers to the behavior of refreshing the traffic on the Landingpage page of the website in order to reduce the bounce rate. This type of flow from the bounce rate indicator on the performance is good, but did not complete the conversion and purchase. At this point we are still very difficult to determine whether this part of traffic is cheating traffic. You need to do depth analysis by accessing the path or by clicking the Hot Zone diagram. However, even the path or thermal map analysis has become a very large project in the face of multiple landingpage. Because we might want to look at how traffic is accessed in hundreds of landingpage. Now we have a good way to solve this problem is to use custom metrics Pageviews/unique pageviews.

　　

Pageviews represents page views, and the unique pageviews represents the unique page views per page, equivalent to the number of visits per page. In one visit, the user browsing a page multiple times will only cause pageviews increase, while the unique pageviews will not increase. So we use different pages as dimensions, dividing by using pageviews and unique pageviews two metrics to see the number of times a visitor browses the same page in one visit. In general, visitors do not browse the same page multiple times on a single visit. So if the value of Pageviews/unique pageviews is high, then this part of the flow is worth noting. Of course, this is not an absolute standard. To be sure, the best way to do this is to compare the Pageviews/unique pageviews values of this portion of traffic with the values of the pages in the entire station.

　2 Visitor Loyalty Analysis

Visitor loyalty is an analysis of the frequency of visitor return visits over a period of time. Generally speaking, when a certain number of visitors come to your site, there will always be a part of visitors to visit again. Even this part of the visitor is very few. Even if there are only one or two. It is as if in a page, even if some of the links in a very hidden position, there will always be someone to click, even if the proportion is very small. Remember a real lesson, when we analyze a WAP site for our customers, the number of clicks on a link in the page is 0. I took it for granted that the link was very high because of the availability of online movies, so it was normal for no one to click. But the reality is quite different from what we imagined.

　　

Therefore, in the analysis of the flow of a channel, the appropriate pull time dimension to analyze the visitor return is also a way to identify false traffic. The real visitors will have a return visit, and the false traffic will not be done after the cooperation is over. So those who do not return after the end of the cooperation period of uniform flow is mostly abnormal.

　3 Visitor Coincidence Degree analysis

Visitor overlap is the ratio of the number of visitors to the top of the queue after a period of heavy weight. For example, let's say that I'm looking for 10 people to click on your ad every day for 10 consecutive days. At this point, there are 10 absolute unique identity visitors recorded in Google Analytics every day. 10 days added together is 100. But when we look at the time dimension up to 10 days, there are only 10 absolute unique identity visitors. This is because Google Analytics the visitor, so each visitor in the 10-day data is unique. According to this logic, we can calculate the coincidence degree of the visitors in different channels. The specific formula is: 1-row heavy visitor/not row heavy visitor *100%. For the case in the example above, the guest coincidence degree equals 1-10/100*100%=90%

　　

For different flow channels, we can also use the visitor coincidence index to identify false traffic. When the flow of a channel in a short period of time there is a high degree of overlap between visitors, we need to further check the flow of this channel quality.

　　4 page Access Long Tail analysis

Page access Long tail analysis refers to the extent to which the visitor's page browsing is extensive. According to the characteristics of real traffic, each visitor's characteristics, interests and habits are unique. They will browse the content of the site in a variety of ways, according to their goals. These natural and diverse features of visitors can be seen through popular content and exit pages in the site. These are false traffic that cannot be simulated.

　　

Popular content is the most popular page during the entire access process. The above image is the browse Volume trend chart of the popular content in the website. Because each visitor's purpose is not the same, there will be many pages to be browsed in addition to the most popular pages, and most pages will have a small amount of browsing, only 1-2 times. These are the long tail of page access, and they fully represent the nature and diversity of real visitors browsing the site. Similarly, there must be a long tail for exiting a page, because visitors will end up on different pages.

The content of this article ends here, using Google Analytics 4 ways to identify false traffic. Do you have any questions or better ways? Please leave a message for me to discuss.

--"All articles and pictures are copyrighted by the Blue Whale (Wang Yanping)." Welcome to reprint, but please specify the transfer from the "Blue Whale website Analysis blog." 】——

Original: http://bluewhale.cc/2011-04-18/google-analytics-identify-false-traffic-2.html