for operational data analysis, I’m sure a lot of small partners will have the following problems:

often faces abnormal data, "what seems to have been done?" What seems to be happening? So what might have caused the subjective conjecture?

face data reports, do not know how to analyze? Don’t know what to analyze?

data analysis is one of the most basic skills of operation. Do you really value it? The qualified operation must be data driven operation, not operation driven data!

1 thinking from a single dimension to a system is a change that must be made in data analysis! For data analysis, you need a structured data framework!

when we consider the problem, we will follow a train of thought, that is, from macro to micro, from global to local, data analysis is no exception. Data analysis in product operation in position here is also needless to say, the data analysis must be based on a detailed understanding of the product data system, doing data analysis to establish data system in mind, product data from large to small dimension system can be divided into macro, meso data, data micro data of three dimensions:

must be able to grasp the data analysis thinking, you dare say that will not do data analysis?

above data analysis latitude does not include all the data latitude of our products. In doing data analysis, we need to do useful data screening in line with our product situation. Of course, in the operation provided by background demand must be based on data needs, such as common user profile data, PV number, UV number, UID number, the number of starts, the retention rate, bounce rate, page access path, many operators in the background to improve data demand, put forward a lot of data, and a lot of data involved the definition and calculation of complex, this will only increase the pressure on the operation of the background data, and by analysis of the actual operation is not big, but the impact of the efficiency of data view. Operation data analysis according to the basic background data table with Excel export functions, and with the help of third party data platform for computer-aided analysis, it not only can reduce the cost of development background data, also can greatly improve the efficiency of data analysis.

2 to do data analysis, need to target oriented, learn to do data dimension split, structured thinking to do operational data comprehensive, systematic analysis.

in the product operation data analysis, we can follow the following ideas to carry on:

( );         Determination of data analysis target

          key dimensions of dismantling target clear data