resulting in abnormal flow change reason is very simple, the holiday is different in a year. Usually in the two period and the end of the beginning of each holiday, website traffic will have a significant decline and rising trend. This change is very easy to verify, date and duration of flow observation change the historical data can also be compared to the same period last year. And the flow change holiday is more regular, and at the same time all traffic sources on the website of the impact. But usually not on the quality of the flow has a great impact.
recommendation: communicate with the familiar website or business personnel, understand the time the sales season, and the time period and prior to the creation of the holiday table.
. Start and end
usually in the face of this situation, we will carry out a quantitative analysis, find out that part of the flow through the breakdown of specific changes. Then qualitative analysis, find out the specific reasons for this part of the flow changes. The first step of the work is relatively simple, the noble baby analytics subdivision function layer analysis, it you can find the reasons. And the second step of the job is more difficult. Because the data itself can explain what happened, but cannot tell why, so we, the real reason we need to find their own
2: industry in the off-season and the season
in different sectors of the website by the holiday effect and the starting time is different. For example, during the Spring Festival change of most site traffic is from the national unified holiday this day, but for travel and ticket sales, the train ticket website, the time may be ahead of time. Because of the Spring Festival for this kind of website is the sales season. So a year sales season will also affect the changes of the traffic flow changes. Industry season relatively regular, usually caused by.
frankly, not every time analysis can find out the real cause of the abnormal flow changes (Qualitative). And in many cases, even when we find the reason and find the reason is so simple, so obviously. And we have ignored it in this article. A summary of the reasons for the 7 most common site traffic anomaly changes. Before the complex segmentation and hypothesis first look at these 7 reasons, you may reduce a lot of unnecessary work. If these 7 reasons have been ruled out, and then start your project..
: create a large table, record all holidays next year in time, then the flow rate changes in different holidays were recorded. This did not help too much may be the first time created, but when you complete record of the year after discharge, will become very valuable.
on site for all traffic sources The first