Suitable for digital immigrants

079 | Data Analysis: From Atomic World to Digital World

I. Concept: data analysis

Data analysis is a quantitative analysis of user behavior.

Data is a trace of behavior that users can't hide.

Data analysis is to deduce behavior from traces and then tell you all the secrets of users.

Difference: Many times, users express what they want, not what they really need, through questionnaires and interviews, but the data left by users' behavior is very honest.

Second, three data analysis methods:

First, analyze the search data-determine the user's needs.

Users' needs will be clearly laid out in front of you through "search keywords".

Second, analyze statistical data-determine the number of users with certain needs.

Users do have needs, but how many users have such needs? At this time, you have to analyze the statistics.

Many industry analysis reports such as Analysys International and iResearch Media provide industry data of different dimensions.

Case: According to the inflow and outflow ratio of urban population, real estate developers judge that the demand in that city is still strong, and then decide which city to buy land on a large scale.

Third, analyze behavior data-determine the key direction of products and services.

Case: The wireless transaction ratio of Tmall Double 1 1 is changing year by year: this figure is 20 14 years 45%, 20 15 years 68%, 20 16 years 82%, 20 17.

This set of data reflects the purchase behavior of 90% users and has been moved to mobile phones. The focus of related industries and services should also be adjusted to mobile clients.

Get it now:

Ordinary mobile phone users basically have three behaviors when using communication functions: voice, short message and traffic.

At present, the short message service has basically shrunk into a big customer in the industry, and ordinary users are increasingly inclined to use chat tools such as WeChat and QQ as substitutes. Voice services have also remained basically stable. Among the three behaviors, traffic-related data contains the most information.

Traffic has two main parameters: the number of attached users (the number of users who use traffic services in real time) and the size of 2/3/4G traffic. If we distinguish between 2/3/4G and time for the peak number of users, combined with the age in the user registration information, we can basically analyze some behavior habits of users. Similarly, a more detailed classification of 2/3/4G traffic, combined with the corresponding data such as time, will find more interesting rules. For a simple example, through the analysis of traffic proportion, we can find the most frequently used network services and the most frequently used time periods, which can further help us see the development trend of social tools, video platforms, live broadcast platforms and so on in the Internet industry.