New media operation: data mining method!

Unlike traditional marketing and promotion methods, online marketing is a more proactive marketing method. It needs to look at users from a developmental perspective, actively analyze user data and content data, find hot content that users are concerned about from the data, and use some data to analyze new users and lost users.

First; Analyze user data and content data

1. Collect data

How to collect data is a question that operators need to think about. The "platform" is the main source of data. The data collected on the platform is of great reference value. Next, I will explain several mainstream platform data.

WeChat platform:

WeChat is a must-have social software for each of us. Similarly, we as data analysts also need to pay close attention to the WeChat background. There are six analysis projects under the WeChat acquired statistical function, including user analysis, graphic analysis, menu analysis, message analysis, interface analysis, and web page analysis. In these projects, there will be designated trend charts. The advantage of trend charts is that they allow operators to see the data more intuitively.

2. Organize the data

When it comes to organization, you need to export the background data first, and then do some organization. There are many ways to organize. Choose one that is easier for you to use. In this way, the data you get will be accurate and you will not waste too much time.

3. Arrangement form

Data can be presented in many forms. In order to allow operators to refer to the data more intuitively, I have compiled several forms here so that you can easily choose the organization method that is suitable for you and your company.

• Line chart: Analyze the trend of data changing continuously over time.
• Pie chart: Analyze the relationship between the proportion of data and the total amount of data.
• Bar chart: By comparing the data, the difference in the quantity of each data can be intuitively reflected.
• Area chart: Analyze the increase or decrease in the amount of data and the total value over time.
• XY scatter plot: shows the relationship between several data points.

4. Analyze the data

After you have sorted out the data, what you need to do is to analyze the data you have sorted out, compare the data, and analyze its trend changes. For example, if you find that the number of views in a certain period of time suddenly increases or decreases sharply, then as an operator, you need to analyze and understand what articles are pushed in this period of time, and find some reasons from it. If the number of new users of your platform continues to increase sharply in a certain period of time, then it is very likely that the WeChat public platform has launched some activities at this time, and the number of users will continue to increase. At this time, what you also need to do is to find some reasons that lead to the increase in the platform, and record them, lay the foundation for future platform operations, and accumulate relevant experience.

Second, find hot spots from the data

Finding hot spots is what every operator needs to do. The way to open up the marketing path is that only when your platform itself gathers topics and hot spots can you gain the attention of users. If you want to discover hot spots, then what you need to do is to understand the source of some hot topics.

Baidu

As an operator, you can use Baidu Index to analyze recent trends. As we all know, Baidu Index is also one of the most important data sharing platforms in the Internet era. Through Baidu Index, we can understand the popularity of a certain hot spot. It can present competitive products, audiences, communication effect data and information in a scientific mapping method. In summary, the information you can get from Baidu Index is the distribution of netizens who pay attention to these words, what related words the netizens who pay attention to these words have searched for, as well as the search scale of this keyword in Baidu and its rise and fall trend over a period of time.

Weibo

Weibo is a must-have on people's mobile phones nowadays. Some popular topics can be spread on Weibo in an instant. Micro-topics on Weibo mainly show popular events with relatively high attention within 24 hours. Operators can find the areas of their concern according to their own promotion direction and then embed this topic into their own push information, which can increase user attention and reading volume.

Taobao

For e-commerce or platforms that mainly sell products, it is inevitable to pay attention to market conditions, so the Taobao rankings are your best reference. You can choose the product categories in the subdivided industry according to your own actual situation, and then choose the products you want to sell.

iQIYI

The iQiyi Index is a video data analysis platform. For video platforms, it is necessary to frequently use such a video index platform to analyze the playback trends of popular videos, user viewing behaviors, characteristics of viewing users, etc. Therefore, this platform is relatively easy to use for video platforms.

At the same time, analyze the new users and unfollow users

Yesterday's key indicators are mainly based on days, weeks, and months as time units, analyzing the changes in the number of users at different time points. We all know that as a new media platform, no matter how hard we try, it will be in vain without a sufficient number of fans. Therefore, operators should pay special attention to the dynamics of new users. Understanding the user's yesterday's key indicators is a good entry point. Normally, WeChat operators may not be able to see the changes in these data, but when your platform launches a new plan, these key indicators can play a very good role and can intuitively reflect the effect of the new plan, allowing operators to summarize experience and fill in the gaps based on these data and indicators. In addition, we must always find problems and analyze the data of the number of people who have canceled their followers. The number of people who have canceled their followers is also data that operators should focus on, because the cost of maintaining an old customer is much lower than that of adding a new customer. Therefore, if there is a cancellation of followers on the platform, it must be taken seriously. If there is a maintenance loss of fans, then the company must analyze the reasons more and try its best to solve and prevent this situation from happening.

Finally, the eight methods of data analysis

The first method: direct evaluation method

From the literal meaning, it means to directly define analytical data based on years of experience, so as to judge its quality. It is usually analyzed by evaluating the recent number of readings, sales, and whether the number of articles pushed on the day is normal. However, if you want to use the direct judgment method to perform data analysis, there are two rigid requirements. The first is that the operator must have rich new media operation experience and can correctly evaluate information such as reading volume. The second point is that the processed data must be intuitive enough to directly represent the pros and cons of a certain value. If these two conditions are not met, then you cannot use the direct judgment method. Therefore, the direct judgment method is a data analysis method that can only be used when both the operator and the data are of high quality.

The second method: comparative analysis

Comparative analysis, as its name suggests, is a method that can only be used to compare two or more groups of data. The differences between the two groups of data can be analyzed to gain an in-depth understanding of the rules represented by these data. Comparative analysis is divided into vertical analysis and horizontal analysis. The horizontal method refers to the comparison of different indicators in the same time period, while the vertical method refers to the comparison of the same indicator in different time periods. Through comparative analysis, we can directly understand the quality of operations and the current level of operations. By understanding the current level of operations, we can summarize and analyze the advantages and disadvantages. Of course, the advantages should be maintained or even further improved, and the disadvantages should be adjusted and focused on for breakthroughs. Therefore, comparative analysis is more suitable for an assessment of operational quality.

The third method: group analysis

The grouping analysis method refers to grouping and analyzing the analysis objects through certain indicators. The purpose of this analysis method is to have a deeper understanding of the different characteristics, properties and relationships of the objects to be analyzed. The grouping method needs to follow independent and completely different enumeration analysis methods, which means that there can be no overlap between groups, and there must be obvious differences between groups. Each data can only be classified into one category, and no data can be omitted in the grouping to ensure the final integrity of the data. Each group can accommodate all the data.

The fourth method: structural analysis

Structural analysis refers to comparing and analyzing existing data domains and common data in some groups with statistical data. Structural analysis is to determine the proportion of various parts in the whole.

The fifth method: average analysis

The average analysis method is relatively difficult. It uses the average to measure the general level of a certain type of data at a certain time and place under certain conditions. The average data is more objective than some data and can help operators analyze subsequent trends and patterns. The average data includes the arithmetic mean, geometric mean, logarithmic mean, etc. Among them, the most commonly used is the arithmetic mean. The formula for the arithmetic mean is, arithmetic mean = sum of all data in the population/number of data. This formula is very important, please write it down, it will be useful in future data analysis!

The sixth method: Matrix analysis

Matrix analysis is a method of quantitatively analyzing problems. It refers to using two important data indicators as the basis for analysis, and defining these two indicators as horizontal and vertical coordinates to form four quadrants, so as to find solutions more intuitively and provide data reference for operators.

The seventh method: Funnel plot analysis

The funnel chart is actually an inverted pyramid. I believe that everyone must be very clear about its shape. So what is the method of funnel chart analysis? In fact, it is very simple. Just fill in the relevant data you have summarized and you can intuitively see the status of each step of the data conversion. The prerequisite is to first count the relevant data. This analysis method is best used in conjunction with other analysis methods.

The eighth method: regression analysis

This method is more complicated. It refers to predicting future development trends by studying the cause and effect relationship of the development and changes of things. In other words, this method is used to study the relationship between variables and is often called the causal method. In fact, this method is not very commonly used, so I won’t explain it in detail here.

Well, this class ends here. New media data analysis does not rely on a single data collation and analysis, but still requires operators to carefully refine and summarize. The core should be presented through the idea of ​​raising questions, analyzing problems, and solving problems. So I think everyone should know how to do it. See you next time!

Author: Spider

Source public account: Spider