Basic knowledge of data that operators must understand (Part 2)

Which indicators do website operations pay more attention to?

First, let’s talk about traditional website operations, which I also divide into four categories: traffic, visits, activity, and conversion.

flow:

Let’s make a vivid analogy. If we regard a large number of users as water and each person as a drop of water, then tens of thousands of users, like hundreds of thousands of drops of water, will converge into a stream and become traffic. Traffic in website operation refers to the number of visits.

The following terms should be noted in traffic: PV (page view), UV (user view), and VV (visit view).

Let’s take Didi Toilet as an example. If it is a website and a user goes to the toilet three times a day, then vv refers to the number of times I go to the toilet, that is, the number of visits. Because there is only one user, there is only one UV. If going to the toilet once requires visiting many pages of the website to complete this activity, then pv is formed, which refers to the number of pages visited. These three data are the three most common indicators used to count website traffic.

access:

In fact, this indicator is not used to measure the number of visits, but to measure the user's access behavior. It mainly includes the following indicators: bounce rate, second bounce rate, and conversion rate .

If I enter the homepage of Didi Toilet and find the page unsightly or the experience inconvenient and don’t want to use it anymore, I will not proceed to the next step and exit. This behavior is called a pop-up. If 10 people visit this page that day and only 1 person bounces out, then the bounce rate is 10%. This metric is used to measure the attractiveness of a website.

The second indicator is the second bounce rate. When you first come into contact with this indicator, it is easy to form the same understanding as the bounce rate. In fact, the meaning of this indicator is completely opposite to the call-out rate. If I think the Didi toilet page is well designed and I want to experience the next step, then I will enter the next page through a function or link on this page. This behavior is called a double jump. The method for calculating the second bounce rate is the same as the bounce rate.

The term conversion rate is not unfamiliar. I have explained it in detail in the previous article, and this indicator appears here again. If you want to understand the word conversion, it is recommended that you define it first and then use it, so that we can better apply this indicator for analysis.

So what does it mean on the website? There are several pages on the website that form a complete process or steps. In this process, the ratio between the previous page and the next page is called the conversion rate.

Let's use the same example. There are three pages for registration, login and payment. One hundred people registered, fifty people logged in, and twenty people paid. Then the conversion rate from the registration page to the login page is 50%, and the conversion rate from the login page to the payment page is 40%. The reason why this indicator appears is that it can be applied between any two pages to measure the conversion rate between them, so as to analyze whether the next page is attractive enough to users.

active:

The activity here still needs to be defined before use, but in this process you will encounter abbreviated indicators: DAU (daily active user) daily active users

MAU (monthly active users) Monthly active users.

Conversion:

This conversion on the website simply refers to transactions, that is, on a website, we focus on indicators such as order volume, payment rate, average order value and payment amount. Here, the average order value is equal to the total payment amount divided by the number of paying users. This indicator is used to measure the revenue that each user can bring to the website. If the total payment amount is divided by the number of orders, it is used to measure the revenue brought by each order. This data indicator is often encountered in the e-commerce industry.
Which indicators does APP operation pay more attention to?

In fact, what we have talked about above is almost enough. With the development of mobile terminals, app operation ( product operation ) is becoming more and more popular. The indicators encountered in this industry are still new additions, activity, retention, and conversion.

Added:

It refers to the number of newly added devices, because for an APP, it is very difficult for the program to identify a user. Sometimes you install a software but may not register it. However, for programs, it is easier to identify a machine. Each machine has a code that will not change since it leaves the factory, so the new additions on the mobile side refer to the number of new devices.

One thing that is particularly important to note is that the number of downloads is not equal to the number of new devices. Registration and activation are also required so that the data will remain in the database. We still use Didi Toilet as an example. Our user is set as Xia Yu. The first time Xia Yu downloads and uses Didi Toilet, it becomes a new device. If he changes his mobile phone and downloads another Didi Toilet and logs in to Xia Yu's account, this becomes the second new device. However, the newly added account Yu 2 has nothing to do with the device, so it is naturally not considered a new device. In other words, this indicator has nothing to do with the account.

active:

On the mobile side, users are more concerned. In most cases, activating applications is called activity, but you can also define activity based on the specific requirements of your industry.

Retention:

In the retention link, we often encounter the indicator of next-day retention rate . To better explain this indicator, let me give you an example. During a certain promotion, Didi added 1,000 new toilet equipment. The next day, only 200 of the newly added equipment were used for Didi toilets again. The next day, the next-day retention rate was 20%. In addition to the next-day retention rate, there are other n-day retention rates divided by time dimensions. Their meanings are similar, but they are defined for the purpose of analyzing specific situations.

The TAD indicator is an extension of the retention rate. By adding up the next-day retention rates of the previous n days, we get a number, which is the n-day TAD. When measuring the TAD indicator, the time needs to be long enough in order to eliminate randomness in the data, eliminate some fluctuations, and make the results more scientific.

Conversion:

For APP operations, the core indicators are similar, including order volume, payment rate, average order value, and payment amount. Generally speaking, a transaction is a conversion.

Operations in different industries focus on different core indicators?

Core indicators, as the name suggests, are the indicators that you care about most. These can determine your performance, your KPI, your promotion and salary increase, and your resignation. Below we divide them into four major categories and give examples.

Content type:

Content products live on content, such as Toutiao or NetEase News. The more people read them, the better. Therefore, content-based products pay most attention to reading volume, PV, UV, VV, PGC (the number of professional content), and UGC (the number of user-generated content)

Social :

Social type: the core function of social communication is used for communication, which is mainly divided into synchronous (QQ) or asynchronous (forum BBC). Social products pay more attention to the number of people who speak and post, and some customized active behaviors. The definition varies according to different social products.

E-commerce type:

This is relatively easier to understand. For e-commerce sales, performance is king, so this industry pays more attention to sales revenue, order volume, and average order value.

Game type:

The purpose of this industry is to attract more people to play and pay, so gaming products pay more attention to the number of active users, payment rate, and revenue.

However, one must especially point out the AR PU (average revenue per user) indicator, which is the total revenue for a period of time divided by the total number of active users for that period of time. For game operations , the more RMB warriors the better, or if 0.1% of all users are RMB warriors, but this person is a big V, that's also acceptable. This indicator is used to calculate these two forms of data.
To better illustrate the metrics that may be used in specific industries, I have listed the five most common operational position metrics.

Channel Operations:

The function of channel operation is to attract new users, that is, to place external advertisements and build good relationships with major application markets . So what does the channel pay most attention to? The two most important indicators are new additions (focusing on the final traffic results) and ROI (return on investment). The reason for paying attention to these two indicators is that it costs money to invest in advertising. If the revenue from a channel's external advertising is lower than the cost, then why should the channel place this advertisement?

Product Operations:

This refers to a position, not a department. It means helping to optimize products through operational means. What indicators does this position focus on most? The most important indicator for this position is very similar to that of the product manager . It focuses on how many people are active in my product every day and how is the retention rate of my product. If a hundred people come in, how many of them can I keep?

Marketing Operations :

When it comes to marketing , revenue is always king, so indicators such as revenue, number of orders, average order value, and ARPU are essential.

Content Operations:

The indicators that content operations and products focus on are very similar, so I will not elaborate on them here.

User operation :

User operations are the people closest to users. If we organize an event, how many people will participate in the end and what will the effect be? The most important indicators for user operations are user activity, user engagement, and activity effectiveness.
The essence of data analysis

Having talked about the meaning of so many indicators, what is the essence of data analysis? Let's make a comparison of active users. You can see that the number of active users of Didi Toilet has increased significantly. This behavior is actually the essence of data analysis - comparison. A single number is meaningless. Only through comparison can we find differences and formulate strategies. There are three common data analysis methods: composition, comparison, and change.

constitute:

Generally it refers to the whole and the part. For example, Didi Toilet has 200,000 users, of which only 10,000 are women, and the rest are men, so we can measure the relationship between the whole and the part.

Compare:

Comparison is the essence of data analysis, and only comparison can produce results.

change:

Changes are more often related to time. What kind of changes occur over time? For example: with our promotion, Didi Toilet had only 10,000 users last month. Later, we found a spokesperson, and the number of users this month successfully dropped to 200. This is change, and time will produce results.
The recent divorce incident of Wang Baoqiang is given by Baidu Index as follows: The essence of data analysis is comparison, so what should we compare?

The first is the peak value: by comparing the highest values ​​in two time periods, the peak value of the search volume for the keyword Wang Baoqiang can be easily found. If divided into two time periods, this one is higher than other places.

The opposite of the peak is the trough, which is to compare two time periods to see which is lower or worse.

The third type is cumulative, which is the total number over a period of time.

The fourth type is the average, which is the cumulative total divided by a length of time, such as a few days.

The fifth category is trend: whether the trend of data change is increasing or decreasing.

There is one thing you need to pay attention to. When comparing data, you must try your best to ensure objectivity, control variables, and ensure that the data are comparable. This is the condition for comparison, otherwise the data are not comparable.

Commonly used data analysis tools

After talking about the methods of data analysis, I will introduce you to several commonly used data analysis tools. You can choose the analysis tools according to the needs and complexity of your data analysis.

Excel:

A common and easy-to-use tool. Basically all the data involved in my data analysis can be processed using Excel. It is very powerful. Once you learn functions and pivot tables, you will basically be fearless when traveling around the world. This can double the speed at which you can process data.

Matlab:

It is more like a programming software because it was originally developed for mathematical processing and writing mathematical programs. Data analysis is just one of its functions. This software is also very powerful, but it requires programming skills, which is quite difficult.

Spss:

It is a software developed by IBM specifically for data analysis. It is very professional. Since this software is very large and has many functions, you need to spend time and energy on it, or even take classes to learn how to use it. It feels similar to PS.

Online Tools:

Most of the time we use online tools, which is very convenient. The most common one is Baidu Index. Umeng requires payment. ASP100 and GROWING IO are both very popular recently. They can help you with statistics without you having to do any processing in your product.