There are so many data that all kinds of data are expressed in different ways. What exactly should I do? Some people say: "In the initial stage of the product, activities are supplemented, and data processing lies in stability." Some people say: "in the middle of products, activities are the mainstay, and the processing of data lies in supervision." Some people say: "At the end of the product, activity is the core, and processing data lies in motivation." Others said: "There are four steps in processing product data!"
Step 1: Look at the overall data, mainly to see what anomalies exist in the overall data and what data trends are good (for example, the overall data, the number of players is stable and the monthly income is extreme).
Step 2: Look at the breakdown data (for example, the breakdown data shows that the number of new users in the game is directly proportional to the loss of active paying users, and the new users don't pay, and the loss of big R is serious)
Step 3: Combine data analysis (such as analyzing data, why are paying players losing? No paid competition? Or is it fun to get to the top of the game pyramid after paying? )
Step 4: Take action according to the data (for example, update the version, recall the players, change the quantity ...)
Such knowledge is estimated that students have memorized it. In this article, the author will discuss the value of data and business thinking on the Internet with children's shoes. The author thinks: data ≠ mathematics! If you look at the game with functional thinking, you can only say that you are good at math; In the Internet industry, it is necessary to combine user behavior data with business thinking to create Internet value.
1. Cultivate the commercial sensitivity of data. Recently, I read the news that the top management of a studio resigned frequently, the project team laid off employees one after another, and the major headhunters were busy grabbing news. I recently chatted with HR and learned that the resume of a certain website has now risen to 15 yuan. At first glance, it has nothing to do with this. If you savor it carefully, it is very appropriate. If you change your mind, it is another scene. ...
Take Zhilian as an example. The website mainly depends on the number of registered people and the routine of hard and wide/waiting for him. If HR buys a resume and goes to Zhilian, it may not be able to get the resume he likes. And headhunters can be said to be motivated by the wind, and often take the initiative to master most high-quality resumes, which not only saves money, but also obtains good resources; Grasping the market trend and cultivating business sensitivity, it is not difficult to find that if an MMO game loses a large number of users (because of trust? Guanfu? Others ...) As another product of MMO, can you smell this commercial atmosphere sensitively in advance? If not, the user will return to the channel (other games), and you will undoubtedly have to wait for him and beg the channel to give away traffic, which is undoubtedly a failure.
Of course, we can't get data from other people's backstage, so where do we usually look at the data trends of other games? Look at the competing forums, game updates, and the user activity of the forums, and you can see a little clue. Then contact with users, and everything will naturally come out. As for how to attract users, it naturally varies from person to person.
2. Derivative sensitivity of cultural data
If the sales of toothbrushes in the market increase, can you feel that the sales of toothpaste will also increase? If you put it in the internet market, it is not difficult to see a very pessimistic fact. Sales of toothbrushes have increased. Overnight, white jade toothbrush, ivory toothbrush, cartoon toothbrush, agate toothbrush and other products rose overnight, and finally inflation, everyone could do nothing.
In this case, it is the mentality of the developer, and I can't say anything; This paragraph mainly talks about the derivative sensitivity of data. For example, if a piece of rare equipment rises from 100 yuan to 200 yuan, then the entry props for making copies of rare equipment/special maps will also rise from 10 yuan to 20 yuan. Props rise, the player's recharge efforts have to increase; As the player's recharge increases, the arpu value will also increase. How to maximize ARPU value? From the product level, increase recharge activities, adjust the output probability of equipment, grasp the needs of users, do what they like, and maximize benefits; Adding all kinds of equipment, instead of increasing the value of equipment, will only be counterproductive.
3. Look at the data from the perspective of empathy
Some CPS choose channels and attach great importance to traffic. No matter what the product is, as long as the channel traffic is good, they will go to the channel, promote it and engage in marketing.
Flow is related to suitability. No matter how much channel traffic there is, it's not yours. Even if it is yours, it is not yours alone. Think about it from another angle. Look at products from the perspective of channels, channels look at products, look at transformation, look at payment, look at retention; Knowing the root and the bottom, data is a two-way street, but they look at each other from different angles. If it is really necessary, at least exchange channel data with product data.
On the other hand, if all kinds of data of products are high; It is best to find out which channel users come from and who are the main user groups. In this way, product design can be more inclined to users' preferences, which is a good way to improve the conversion rate. (Taking MMO mixed service as an example, it is easy to understand that users can use channel identification to mark packages. )
4. Look at behavioral data with business thinking
Behavior data, that is, user behavior share, such as activity, retention rate, payment rate …
Business thinking, that is, benefit analysis, such as user cycle value, the exploration of user exploitable value …
For example, the two guilds are in conflict, the game is hot, and the guild members are active, compared with the level; Guild presidents compete for equipment, compared with recharge, both sides are on fire, online and recharge to a considerable extent; What do you do? If you don't do anything, you will be there secretly happy and complacent; I read a book that said, "Sitting is waiting for death!" "If you don't believe me, two guilds will make up for it next month, or one guild will be kicked out of the game, and it will be too late to regret it.
"Do you want to do nothing?" If you don't want to, you have to learn to look at behavioral data with business thinking; For example, what is the competition platform of these two gangs? Forum? Post it? Who is taking the initiative and who is paying? Who is the person who influenced them? Are they still worth exploring?
How to balance this relationship? The ball effect is very important. The harder you push, the higher you bounce, regardless of anything. You will only bounce lower and lower and return to calm. At present, the more common game is to support this catalyst; Indeed, trust has played a certain role, but it is not a cure; If we think with business thinking and take Tianya Post Bar as an example, the topic already exists and the real users already exist, then word-of-mouth marketing can be easily realized, attracting potential users through the topic of the original users and bringing more business benefits; By controlling users' habits (such as joining in the fun) and human weaknesses (such as the higher the status, the better the face), a marketing can be created. This marketing effect is remarkable, and the most important thing is that it doesn't cost money!
A very simple user behavior, very common user behavior data, from another angle, may be a commercial marketing!
5. Look at the relationship between users and products through data.
Many people attach great importance to fixed data. Arpu and other core data form a set of advertisements. Countless people check one by one to measure the quality of their products, and countless operations use this to check and measure the success of operations. If you only do it for KPI, then you are successful. If you want to do higher, then this is not enough.
The relationship between users and products, most students also defined in user positioning, product positioning; Further, it is a set of core data evaluation and operation processes. ...
The author believes that data, users and products; There are too many aspects that can be discussed when the three forms a triangular relationship. For example, one user is online for 5 minutes and one user is online for 10 minutes. What's the difference between them? If 10 minutes is defined as an active user, where are the online goals of 5-minute users and 10 minutes users? What level of users stay for how long at what time? These are complicated and easily overlooked.
For another example, if a user buys two gems at a time, is it a one-time purchase? Or repeat purchase? Don't underestimate this kind of data. The user's single purchase and divided purchase directly determine the user's needs. Before the same amount, the interval value is very large!
Finally, think about another industry. The screenwriter industry has a definition of the script, and the script is only 5 minutes! This five-minute conversation is not about the movie cycle, but that you only have five minutes to impress your users. If five minutes is not enough, users will lose patience. The same is true of games. Back to the beginning, one user is online for 5 minutes and one user is online for 10 minutes. The difference between them lies not only in the time difference, but also in the time viscosity of the product. Take this as an example, if the opening animation is exquisite and the game screen is great, it is very easy for users to indulge in it for 10 minutes. If the initial beginner's guide is troublesome and protracted,
Finally, we can see the relationship between users and products through data, find problems through data, sort out problems through users, and solve problems through products, which involves not only operations, but also planning, art design and other departments. After all, the product is not delivered to the operation online, and unity and cooperation are the key points!
There is a lot of data, and it is also very miscellaneous. They form a network of relationships with each other and will touch the whole body; As for how to understand it, different people have different understandings and can only say: "Data is very important! What matters is not how accurate his algorithm is, but the grounding gas! He told us what to do next! "
The above is what Bian Xiao shared for you about how to analyze user behavior data with business thinking. For more information, you can pay attention to Global Ivy and share more dry goods.