A summary of the data analysis of an Internet

Whether it is product or operation, data analysis is a section of the daily business, through data analysis, we can get problem feedback more clearly, find the root cause of the problem, and resolve user demand can not satisfy problems. Specifically, how the data analysis is to promote the business, help thinking? The author of this article made a summary, let’s take a look.

First, say

The number is originally abstract text expression, but the number and an event are associated with an event. After a series of data is generated, this changing data is a string that quantizes a specific event, and an abstract event icon Process.

This process that will affect abstract events, brief summary, can be divided into three phases:

The first phase is the collection of data; the second phase is statistical output data; the third phase is the analysis of the data.

Then, we will focus on the third phase of data analysis. Data analysis is to find “relationships” between quantitative events and time, find a connection or variation between two variables.

Second, the importance of data analysis

As the Internet industry is constantly developing, for many practitioners, the requirements of individual capacity are constantly expanding. Whether it is a full-link designer’s concept proposed by major Internet companies in the past few years, it is still a product designer post and user experience designer positions, and the user experience designer. It is necessary to continue to learn and continue to expand your personal capacity boundary.

Previously, the data analysis data was the product and operation of the product, but now every classmate of product development, there is a significant advantage that the data analysis capability is a distinction between other products research and development.

The design is on the line, only the head of the head is said to be a film, only the objective data will be more convincing, so the product developer’s data analysis ability is more and more attention to everyone. .

Third, data analysis process

In daily work, many product developers are more embarrassing after getting a dot data, especially for some product or designers who are not very sensitive to data, get PV, UV, access, and jump out. After the conversion rate, the experience of experience, the average access, the average data, and a pile of Arabic numbers have been taken out alone, but the combination comparison is as if you can’t understand.

In fact, good companies will have a data analysis system, which will help product R & D staff to make data cleaning and visualization of data, so that more convenient use of data is analyzed and thinking.

However, some product developers who offer product developers may need product developers to handle and analyze their product.

So, what is the process of getting data analysis? Here is a key steps to feel effective here:

For a product R & D, the data analysis step can be divided into four steps, and the purpose and goal of data analysis is the question of the primary consideration, with the result and is guided, with a clear purpose to do data analysis. For example, if you want to ask the new sign-in function of the app to keep the user’s substantive effect:

1) What is the expected goal: When the sign-in function of the app is used, the target of the user is reached.

2) This analysis focus: analyze the MUV after the new function is online, and the entire app MUV.

3) This analysis task:

Compare the APP date before and after the new feature; analyzes the change trend and trend of MUV of the function MUV and APP after the new function is online.

The second step is data refinement processing, removes some unnecessary data, or calculates the average of data, or the time of recording peaks.

The refinement of the data can be divided into three steps:

1) Data statistics

According to the analytical objectives, useful data; the designer should strip design-related statistics in design performance / user behavior.

2) Data refinement

Data cleaning, delete useless information data / data is not fully, selecting useful historical data, such as data in the same type in the same scene and different paths; calculate data results, clicks / mean / conversion.

3) Data visualization

The refined data is used to use visual tool processing, into PBI / Excel / Keynot, etc.

The above six basic charts can basically apply most of the data visualization.

The third step data analysis is to observe and analyze the data charts after the finishing.

The ability of data analysis is for designers, and it is not as fine as data analysts, as long as you can have a product or designer’s requirements for data analysis capabilities as long as you find the same and problem from the data. Can find and ask questions, then solve the problem by designing methods.

I also summarized 4 steps for data analysis:

1) Description Phenomenon

Objectively describing the performance of data; an abnormality or regularity is described by visualization chart trend.

2) Perceived problems

There are abnormalities or differences, there must be problems; there is a problem, it is necessary to find the reason; the data reflects the design of the design, trying to make questions. 3) Thinking / proposing

Suppose is a preset plan for “Question” to find a solution;

4) Analyze verification

Through anomalous / associated relationship analysis, the interference factor is ruled out to determine whether it is a problem of design itself; classification and layers of user feature / behavior / design performance are analyzed to verify whether it is a problem.

The final output summary is to analyze the output and summary of the results, this is not described in detail.

In a specific work, in fact, each team will have different requirements for the content of the analysis results, so there will be a thousand different contents and forms, but the focus can reflect why this data analysis is to be used.

Fourth, say in the end

Many people mistakenly understand the data analysis, and the key to improve data analysis capabilities, in fact, the core of data analysis capabilities is not in methods and tools, but thinking.

For business people, the core idea of ??data analysis is to obtain “quantization relationship” between two variables, to explain the phenomenon; the steps of data analysis, perceived problems, propose hypothesis, selection characterization, collect data, analysis and verification; Propose hypothesis and selection characterization is why many business people data analyzing; data analysis is the most important thinking is to determine the relationship between two group variables in the business to explain the business.

The data is used to describe the phenomenon, which is a means of simplifying an abstract event or a phenomenon. There are many ways to analyze the method, but the method of use requires us to choose what kind of method according to the nature of its team or R & D product, need to constantly verify the applicability of the method, to summarize a set of data suitable for yourself Analysis SOP is your own data analysis capabilities.