The principle is simple, but I don't know how to use it? Understand "contemporaneous group model" in one article

Reprint / Coder bear

Today, let's continue to talk about the eight model series of data analysis . Shared today , It's a simple principle , But widely used models ： Contemporaneous group model . In commodity analysis 、 User analysis 、 Channel analysis , It's all useful .

One 、 The principle of synchronous group
The principle of simultaneous group analysis is very simple ： Grow beans . How to distinguish which beans are good ？ The easiest way , Is to divide into groups （ So-called “ Group ”） Plant it together （ So-called “ At the same time ”） Then watch which grows faster .

In order to achieve this goal , We need to ：

Select the comparison object , Classify according to conditions
Select the appropriate comparison index .
Observe from a time node .
Compare the index differences , Find quality / Inferior groups

Take the example of beans , The practice of the same period group is as follows ：

The idea is very simple , Therefore, it is widely used in work , Many other names have also been derived .

Two 、 Commodity group ： goods LTV Model
Contemporaneous group model in commodity analysis , Also known as commodities LTV Model .

Doing this ：

Set product grade （A、B、C level ）
From the time the product goes on the market , Start to observe
Observe the sales volume after the product is on the market / Profit trend
Compare each grade of goods , Whether the average level of the product is reached
If the performance is better than the average , Focus on the shortage problem , Secure supply
If performance is below average , Focus on the backlog , reduce stock
（ Here's the picture ）

This model is very easy to use ！ Because many goods , Starting with the listing , Born with a life cycle trend . Through the analysis of synchronous group , It can not only simulate this trend , And you can set a reasonable price for each level of goods LTV Range , So as to find out whether the commodity sales meet the expectations as soon as possible . So as to further control the inventory , Maximize profits （ Here's the picture ）.

Not just commodity analysis , User analysis can also be used .

3、 ... and 、 User synchronization group ： User retention model
User synchronization group analysis , Also known as the user retention model .

Doing this ：

Set user grouping （ Generally according to the registration time or Registration channels ）.
From the time of registration , Observe each X Days later , Retention rate of this batch of users .
Fit the overall trend , Anticipation X Days later , The quantity retained by users in this batch .
Find the node with the most obvious decline in retention , Determine whether to further analyze .
（ Here's the picture ）

Because most of them APP Can only meet some needs of users , Therefore, the retention of users , Always slowly reduce . If some nodes are found , User retention has decreased significantly , This indicates that there is a problem with these nodes , Further analysis is needed . meanwhile , Based on the same period group data , The estimated number of retained users can be fitted , Can plan and serve user resources for operation , Provide data support （ Here's the picture ）.

Four 、 User synchronization group ： user LTV Model
When predicting user retention , Synchronously calculate the value expected to be generated by users , Then you can calculate the user LTV. The specific methods are as follows ：

Use the above method , First calculate the value retained by the user
Calculate each stage , User pay rate 、 The amount paid
Total user value = Retain users Rate The amount paid

Be careful , Calculated in this way LTV, Actually, it means ： Within a certain time , User generated value . Not strictly life cycle value . however , Considering that the enterprise makes business plans , Also in years / Quarterly for time units , Therefore, only users in 3 Months /6 Months /12 Value generated within months , Almost .

Be careful ！ In different business scenarios , User payment forms will be different , So right. LTV The calculation will have an impact .

Common situation , Such as ：

Ideal state ： Users pay a fixed amount every month , Pay in proportion （ Monthly rental business ）
The front is low and the back is high ： The more loyal users , The more you buy , The higher you pay （ Fan business ）
The front is high and the back is low ： Attract users to pay a lot in the early stage , Don't worry about the back （ Harvesting business ）

Corresponding data , Maybe as shown in the figure below . Computing users LTV When , Pay attention to the payment form . If you just use “ Average monthly consumption ” indicators , It is likely that the average value will cover up the real payment situation , So as to mislead business judgment .

5、 ... and 、 Channel synchronization group ： Channel quality analysis model
If in the user group analysis , Put the classification dimension , Change to ： from XX Users coming in through channels . You can make further ： Channel quality group analysis . The common forms are as follows ：

By channel + Advertising time , Categorize users
Focus on this channel + Launch time enters the user , follow-up X Day conversion rate / pay
Calculate user generated consumption , Compare channel launch costs
For transformed channels , Consider adding
For channels with poor transformation , Budget cuts / Rectification measures
（ Here's the picture ）

6、 ... and 、 The shortcomings of the synchronization group
The synchronization group looks good , Because it can fully expose the problem . But the contemporaneous group has no ability to explain the problem , Why did the quality of a certain channel suddenly collapse ？ The same period group alone can't explain , Other analytical models are needed to explain .

Focus on enterprise data analysis and application and digital transformation . Official account “ Business intelligence research ”, reply “ Information ”, Tidy up 6G The data warehouse of 、 Data center 、 Data governance 、 Enterprise data management case , For free ！