Generally speaking, we hope that every content of the new media we operate can get high opening rate and high sharing rate. But the facts are not always as expected. Therefore, in the process of new media operation, we need to constantly optimize our own content in order to get more attention and traffic. So let's discuss how to optimize our content with data blessing today. We simplify data analysis into the following steps: setting data indicators-obtaining data-data analysis-optimizing and adjusting 1. Setting data indicators Before setting data indicators, we must first know what data we need. In daily work, many people often start to retrieve data without knowing their own data requirements, and then think of one to look at. In fact, this is ineffective, because if you don't form a clear goal, you can only think that a point is a point, and these points can't be well connected in series for your system analysis, so you can't give suggestions on system optimization for the analysis results. Therefore, we need to understand our data requirements first. When you don't know, or the company doesn't give you specific needs, you can think about what we want to achieve in the account we operate and maintain, and then disassemble the required data in reverse. For example, if I want to have a high open rate and high sharing rate, then I must first know the number of articles opened and shared and the total number of users. In this way, we know that there must be at least three data in my data requirements: the number of open articles, the number of shared articles and the total number of users. Generally speaking, the daily data needed to analyze the content of articles are as follows: the number of articles read, the number of articles sent, the number of articles forwarded (shared), the number of newly collected fans and the number of clicks on the menu bar. On this basis, we can also analyze the entire operating platform through weekly data and monthly data, so the data we need to use can include: the number of articles read this week, the number of articles sent this week, the number of new fans this week, and the number of clicks on the menu bar this week (the above is just an example, and the specific situation can be adjusted according to the media we operate). After having the data, we need to set an analysis index, which is to judge whether the data meets the effect we want to achieve. We can draw historical data and calculate an average data index to judge the data quality of the article. Of course, it is also possible for a company to use the data of a good article as a qualified indicator, which mainly depends on the company's requirements. If it is a new media account, you can set the index standard according to the data of competing products. Second, how do we get the data? Generally, these data can be seen in the operation background. What operators need to do is to organize the data into excel tables and form daily and weekly reports, which is convenient for our later records and analysis. If you want to get the data of competing products, you can generally use some reptile tools, such as octopus and jujube. Of course, if you can't use tools, you can also pick them manually, but it's a bit time-consuming and laborious. Third, data analysis with data and indicators, the next step is to compare the existing data with standards. If it is below the standard, we need to find out the reason for the low, improve and optimize it; If it is higher than the standard, we should also find out the reasons for it, and increase efforts in this regard to expand the effect. When it comes to analysis, many people often feel that there is no way to start. Here are two ways to teach you. If you really don't know how to start data analysis, you can follow them step by step. Method 1. Ask yourself three questions: Who? -What did you do? -Can I do better? Ask yourself these three questions, but also to let yourself have a basic clear understanding of the media you maintain. Who is it? Is to find out the situation of your user audience. You can know your user information through some data in the background, and it is best if you can get a portrait of the user. Observe the characteristics of most of your users. What did you do? Know who cares about your media account, and why these users care about you, what they get from you, or what they want from you. What is better to do? When you understand the above two points, then think about how to do the second point better. That is, when you know who your user audience is, what they like and want, then cater to their preferences, or attract them and guide them. For example, through the data, I know that all the people who pay attention to me are college students and recent graduates aged between 18-25. Most of them started to pay attention to me because of a highlight of a variety show. Then, through horizontal comparison, we find some related articles, such as some entertainment news and emotional articles, with high opening rate and forwarding rate, and we can send more articles in this field in the future. Method 2: Analyze according to four quadrant dimensions. For example, we can analyze the opening rate and sharing rate that we are concerned about through four quadrants: As can be seen from the above figure, quadrant A represents articles with high sharing rate and high opening rate, which can be used as our excellent standard and also the direction that we need to enlarge and intensify; Quadrant B represents articles with low sharing rate and high opening rate. This kind of article may be of low quality, or it is difficult for users to share, or users may just open the article because of the title and then quit when they find that they don't want to get the content. The specific situation should be analyzed in detail; Quadrant c represents articles with high sharing rate and low opening rate, which may lose the number of times of opening because of poor title selection. However, the quality of the article should still be ok, so the analysis rate is high. For this kind of articles, we can make more efforts on the title to further improve the opening rate; D quadrant represents articles with low sharing rate and low opening rate, and it exists to tell you how to avoid pits. Prove that your users are not interested in this kind of article, or that the quality of the article itself is not good. We can put the obtained data into four quadrants for analysis. Quadrant A is excellent and needs to be strengthened and expanded. Quadrants b and c have their own advantages. We can make further efforts in quadrant A according to the concrete analysis. Quadrant d is basically a guide to avoid pits. You can analyze whether the quality of the article is poor or the direction of the topic is not good, and then make further adjustment and analysis. Note: In the operation of new media, data analysis is undoubtedly more objective, which excludes the subjective assumptions of our operators and puts the specific situation in front of us more intuitively. However, it should be noted that you should not memorize data by rote or rely too much on data. Users are people, emotional individuals and changeable. Articles that were popular in the past may not all be popular later. Users may pay attention to many accounts of the same type at the same time. He may have read too much about some things and felt nothing. If you rely too much on data, you will be limited by yourself, and you may still feel puzzled. Obviously it is the same type of article with good data before. Why don't users buy it? Is there something wrong with the data? The data is objectively presented and there will be no problem. The problem is our judgment. Therefore, in operation, we must learn to be flexible. Data is a tool for us to optimize our operations, and it is good to use it well. So what else can we do besides data? We will discuss them one by one later. Welcome everyone to exchange and discuss.