How to do data analysis? 2 major analysis models and 6 data display charts!

What is Data Analytics ?

Data analysis refers to the process of analyzing large amounts of collected data using appropriate statistical analysis, extracting useful information and forming conclusions by conducting detailed research and summarizing the data. In practical terms, data analysis can help people make judgments so that they can take appropriate actions.

What is the purpose of data analysis?

The purpose of data analysis is to concentrate, extract and refine the information hidden in a large amount of seemingly disorganized data in order to find out the inherent laws of the research object.

In practical terms, data analysis can help people make judgments so that they can take appropriate actions. Data analysis is the process by which an organization purposefully collects and analyzes data to turn it into information. Data analysis processes need to be appropriately applied throughout the product's entire life cycle, from market research to after-sales service and final disposal, to improve effectiveness.

In an enterprise, data analysis can help us understand the enterprise's operating conditions, product sales, user characteristics, product stickiness, and so on.

What are the steps in data analysis?

Steps in Data Analysis

1. First, clarify the ideas and purpose of the analysis:

Data analysis must have some business purpose. It may be to track user usage of a new product after it is launched; it may also be to observe user retention over a period of time, or it may be to operate a certain coupon to see if it is effective. With a certain purpose, determine which angles to analyze from. Then find the metric that speaks to that purpose.

For example, you want to verify whether the latest batch of coupons is valid. We can analyze from two aspects: the receipt of coupons and the usage of coupons. The indicators of coupon receipt can be broken down into the receipt rate; the usage can be broken down into: usage rate, average order value, etc.

2. Data collection:

After determining the core indicators for this data analysis, data collection should be carried out based on the data indicators.

Some companies have very thorough data preparation, and data warehouses, data marts, etc. have been built early on. Some companies are lagging behind in data analysis, so we need to do a lot of data collection work ourselves in the early stages. For example, use some of your own company's or third-party data analysis tools to track data and obtain logs. Or use existing data in the database, such as order data, basic user information, etc.

3. Data processing:

After the data is extracted, dirty data must be removed (cleaned) and then the data must be transformed . After performing the most basic data summary and aggregation, we can get a wide table of data with relatively simple fields and relatively rich data.

4. Data Analysis:

Data analysis is the process of using appropriate analytical methods and tools to analyze processed data, extract valuable information, and form effective conclusions.

The data that a general company needs to observe can be roughly divided into the following categories:

• Business data: payment amount, number of paying users , payment rate, average order value
• Operational data: number of new users, daily active users, weekly active users, monthly active users ( AARRR model )
• Product data: PV and UV of key pages ( funnel model )
• User data: user life cycle , user retention , user average order value, user type ( RFM model ...)
• Product data: product sales, gross profit analysis….

As the importance of data becomes more prominent, more and more companies have realized that data is crucial to their operations. Therefore, most companies have a dedicated BI department to carry out preliminary data processing and analysis, and summarize them in the form of weekly reports for management to use for daily data needs and corporate decision-making.

Here we mainly introduce two simple data analysis models:

AARRR Model:

Acquisition, Activation, Retention, Revenue, Refer

A AR RR model

1. Acquisition

How to acquire users? Online, it can be obtained through SEO , SEM on the website, and through market launch , ASO and other methods on the app. There are also H5 pages for operating activities, self-media and other methods. Acquire users offline through door-to-door marketing and flyers.

2. Increase activity

After users come, we increase their activity by offering price discounts, editing content, etc. Increase the content, increase the products, and provide favorable prices, but keep costs under control so that there is room for growth. Such users are the most valuable to be active.

In terms of product strategy, in addition to providing operation modules and content deepening. Implement product membership incentive mechanism and growth system to cultivate active users. Not only are there icons for product discounts and VIP, but for long business processes, there is an incentive system for the process, and the product strategy is more diversified.

3. Improve retention

By increasing the activity level and having loyal users, the number of users will gradually settle down. In terms of operation, the community users build UCG together by using content and mutual messages, breaking away from the early PCG model. E-commerce improves retention through product quality, and O2O improves retention through high-quality services. These are all business-level retention improvements.

In terms of product model, retention is improved through the sign-in and reward mechanism of the membership mechanism. Both app push and SMS activation methods are product ways to activate users and improve retention.

Monitor the user churn of the app through indicators such as daily retention rate, weekly retention rate, and monthly retention rate, and take appropriate measures to encourage these users to continue using the app before they churn.

4. Earn Revenue

Generating revenue is actually the core of application operation. Even a free app should have its own profit model.

There are three main sources of revenue: paid apps, in-app purchases, and advertising. The acceptance of paid apps in China is very low, and the Google Play Store only promotes free apps in China. In China, advertising is the source of income for most developers, while in-app payments are currently more widely used in the gaming industry.

The aforementioned increases in activity and retention rates are necessary foundations for generating revenue. Only when the user base is large can revenue increase.

5. Self-propagation (Refer)

The previous operating model ended at the fourth level, but the rise of social networks has added an aspect to operations, which is viral spread based on social networks, which has become a new way to acquire users. This method is very low cost and has the potential to be very effective; the only prerequisite is that the product itself must be good enough and have a good reputation.

From self-propagation to acquiring new users, application operations form a spiral upward trajectory. And those excellent applications make good use of this track and continuously expand their user base.

Funnel Model:

User access path

Funnel Model

The funnel model is widely used in daily data operations such as traffic monitoring and product target conversion. It is called a funnel because users (or traffic) enter from a certain functional point (which can be set according to business needs) and may complete the operation through the process set by the product itself.

Monitor users who follow the process at each conversion level to find optimization points at each level; map the conversion paths of users who do not follow the process to find room to improve user experience and shorten the path.

A typical example of using the funnel model is the conversion of e-commerce websites. When users are purchasing products, they will inevitably place orders according to the pre-designed purchase process and finally complete the payment.

It should be noted that a single funnel model is meaningless for analysis. We cannot evaluate the conversion rate of each step in a key process of a website from a single funnel model. Therefore, we must analyze the conversion rate of each step in the process through trend, comparison and segmentation methods:

Trend: Analyze the changes in the timeline, suitable for monitoring the effect of improving or optimizing a process or a step in it;

Compare: By comparing the conversion rates of purchase or use processes between similar products or services, problems in certain products or applications can be discovered;

Segment: Segment the conversion rates of different sources or customer types to find some high-quality sources or customers. It is usually used to analyze the effectiveness and ROI of website advertising or promotion .

5. Data presentation:

Data Visualization - Basic Charts

Data visualization is the scientific and technological study of the visual representation of data. Among them, the visual representation of this data is defined as information extracted in a certain summary form, including various attributes and variables of the corresponding information units.

Charts are a common means of "data visualization", among which basic charts - bar charts, line charts, pie charts, etc. - are the most commonly used.

Data Visualization - Charts

Some people think that basic charts are too simple, too primitive, not high-end, and not grand, so they pursue more complex charts. However, the simpler the chart, the easier it is to understand. Isn’t understanding data quickly and easily the most important purpose and highest pursuit of “data visualization”?

So, don't underestimate these basic charts. Because users are most familiar with them, they should be considered for priority whenever applicable.

1. Bar Chart

The bar chart is the most common chart and the easiest to interpret.

Bar Chart

It is useful when you have two-dimensional data sets (each data point consists of two values, x and y), but only one dimension needs to be compared. Annual sales is two-dimensional data, "year" and "sales" are its two dimensions, but only the "sales" dimension needs to be compared.

The bar chart uses the height of the bars to reflect the differences in data. The naked eye is very sensitive to height differences and can identify them very well. The limitation of histograms is that they are only suitable for small to medium-sized data sets.

Generally speaking, the X-axis of a bar chart is the time dimension, and users tend to believe that there is a time trend. If the X-axis is not a time dimension, it is recommended to use color to distinguish each column to change the user's focus on time trends.

Bar Chart

The above picture shows the number of wins of each team in the English Football League in a certain year. The X-axis represents different teams and the Y-axis represents the number of wins.

2. Line Chart Data

Line charts are suitable for large, two-dimensional data sets, especially when the trend is more important than a single data point.

Line chart

It is also suitable for comparison of multiple two-dimensional data sets.

Line chart

The figure above is a line graph of two two-dimensional data sets (carbon dioxide concentration in the atmosphere and average surface temperature).

3. Pie Chart

Pie charts are a type of chart that should be avoided because the human eye is not sensitive to area size.

Pie Chart

Bar Chart

In the picture above, the area order of the five color blocks in the pie chart on the left is not easy to see. It would be much easier if we change it to a bar chart.

In general, you should always use a bar chart instead of a pie chart. But there is an exception, which is to reflect the proportion of a certain part to the whole, such as the percentage of poor people in the total population.

Pie Chart

4. Scatter Chart

Scatter plots are useful for three-dimensional data sets, but only two dimensions need to be compared.

Scatter plot

The above picture shows the medical expenditure and life expectancy of various countries. The three dimensions are country, medical expenditure, and life expectancy. Only the last two dimensions need to be compared.

To identify the third dimension, you can add text labels or different colors to each point.

Scatter plot

5. Bubble Chart

A bubble chart is a variation of a scatter chart that reflects the third dimension through the area size of each point.

Bubble chart

The picture above shows the path of Hurricane Katrina. The three dimensions are longitude, latitude, and intensity. The larger the area of ​​the dot, the greater the intensity. Because users are not good at judging area size, bubble charts are only suitable for situations that do not require accurate identification of the third dimension.

If different colors (or text labels) are added to the bubbles, bubble charts can be used to express four-dimensional data. For example, the picture below uses color to indicate the wind force level at each point.

Bubble chart

6. Radar Chart

Radar charts are suitable for multi-dimensional data (more than four dimensions), and each dimension must be sortable (nationality cannot be sorted). However, it has a limitation that the number of data points is at most 6, otherwise it cannot be distinguished, so its application occasions are limited.

Here are the stats for the Miami Heat's five starting basketball players. In addition to the name, each data point has five dimensions: points, rebounds, assists, steals, and blocks.

Radar chart

Draw it into a radar chart, it looks like this.

Radar chart

The larger the area of ​​a data point, the more important it is. Obviously, LeBron James (red area) is the most important player for the Heat.

It is important to note that users are not familiar with radar charts and have difficulty interpreting them. Try to add instructions when using it to reduce the burden of interpretation.

VII. Conclusion

The author of this article @hooly compiled and published by (Qinggua Media). Please indicate the author information and source when reprinting!

Product promotion services: APP promotion services, advertising platform, Longyou Games