Data analysis methodology and previous experience summary 2 [notes dry goods]

One 、 Classification of data analysis

1、 Descriptive analysis ： What happened? ?
Describe the characteristics of something , It needs to be accurate 、 Perfect and even real-time data

2、 Diagnostic analysis ： Why does it happen ?
In the acute evaluation of descriptive data , Diagnostic analysis tools will enable analysts to drill down into segmented data , So as to isolate the root cause of the problem

3、 Predictive analysis ： What's going to happen ?
Describe the relationship between data characteristics and variables , Predict the future based on past data

4、 Normative analysis ： What needs to be done ?
The canonical model leverages an understanding of what happened , Why did this happen and all kinds of “ What may happen ” analysis , To help users determine the best action plan to take

Two 、 Key indicators that need to be memorized

1、 Some key indicators
1） Company average daily / Once a week / Monthly turnover 、 Active quantity 、 Traffic size
2） Weekend and mid week are generally different
3） morning , At noon, , Key differences of users at night 、 Active number 、 Traffic
4） Beijing and Shanghai , Market share of each major region 、 Consumption ability 、 Average turnover per user
5） The expected growth rate of the company in the next quarter , Expected turnover this year
6） Active ratio of desktop and mobile 、 Income ratio
7）web and app The proportion of
8） Ratio of free users to paid users 、 The main difference 、80% The turnover is provided by the top few% of users
9） Main funnel , If registered 、 Sign in 、 pay 、 Submission, etc , Conversion rate of each step 、 Loss rate
10） Customer retention of major products 、 Get the cost of users 、 Life cycle value of users

2、 Remember the need for key indicators
1） Memorize key indicators , When you see abnormal fluctuations , Will be sensitive to detect something wrong , That's what we usually say “ Data sensitivity ”, This is particularly effective in human interaction in companies , Because interaction is real-time , An immediate response is needed .

2） Another great advantage of memorizing indicators is that they can bring great credibility to analysts . In planning strategy 、 When positioning products , Many ideas are based on logical reasoning , Industry experience , Class hypothesis , Analysts have the opportunity to provide more quantitative indicators , Provide strong support for reasonable views . Facts speak louder than words ,“ We randomly sample , 90% Users support the current pricing ” It's better than “ Generally, everyone is at this price ” Much more persuasive . Long term interaction based on quantitative facts , Analysts can win a lot of trust .

3、 ... and 、 User Marketing （RFM）

1、 What is user marketing
A series of operational activities for users in the product , Including pulling new 、 Activate or wake up, etc , Through these activities , Enable users to buy repeatedly . Such operation or marketing is accurate , Finally, we can achieve the goal of improving the overall purchase conversion rate .

2、 How to conduct user marketing
1） Identify business scenarios , By drawing user portraits , To find our target users .

2） On the basis of Article 1 , After finding the target user , Precision marketing , Is the so-called activation or awakening , Make this part of the target users , Buy on our website , So as to improve the final conversion rate .

A usage of industry experience , It is the grading of user value . When we talk about user value , This word is actually vague , How can we accurately grade this value ?

Commonly used in the industry RFM Model ：R yes Recency, It is the time since the last purchase ;F yes Frequency, It refers to the purchase frequency of users ; M yes Monetary, It refers to the user's purchase amount .

3、RFM Of 8 Class user
We can from these three dimensions , Rank users , We can divide it into 8 class , this 8 Class can cover the value of the user we want to illustrate .

Four 、 Improvement of conversion rate （AARRR）

1、 Conversion rate
Very popular in Silicon Valley Growth Hacker The concept is the pirate law , AARRR Model , This model is centered on product design and operation , Get users 、 Rapid transformation 、 Improve retention , Then realize and spread , These are all centered on product design and operation .

Whether the previous traffic is king , It is still the king of product design and operation , The core of it is transformation . For the marketing department, we need to pay attention to the transformation and registration of different channel traffic ; For the Engineering Department ] Come on , Optimizing innovative products is also for better user experience , So as to help users complete the transformation of the core ; As for sales , The retention of users itself serves the turnover .

2、 The basic steps to improve the conversion rate
For different industries 、 Mobile terminals and web terminals of different products 、 Under different scenes , We all have different analytical frameworks to think about this problem .
1） Locate the right user
For example, with the help of [ User segmentation ] function , Filter separately from IOS And Android source users , Study the characteristics of these groups in a differentiated way , In order to make targeted product improvement and content optimization .

2） Improve the experience and fluency of the product
After defining the user population , We need to pay close attention to the user experience of the transformation process . There are many ways to improve the user experience , Such as the position of the deployment button 、 Interface color , Or make an overall revision 、 Product interaction, etc . Then analyze user behavior , Internet companies need to iterate quickly 、 Improve the product , Constantly improve the user experience of products , It's a process of trial and error .

3） Optimize the conversion path
We use the idea of funnel transformation analysis , Analyze the conversion rate of users at each step , To help better optimize the registration process , Improve conversion rate .
After understanding the transformation process , We can also further split these steps , Understand the differences between different user groups . In this process, we will find many transformation points 、 Analysis point , These analysis points will help us improve our transformation .

4） Pay attention to re purchase and additional purchase
Repurchase is a very important indicator , Its essence is to retain , We need to focus on and optimize it , It is conducive to improving the overall conversion rate . But re purchase is not so important for all platforms , such as SaaS product , One registration is enough , Let's just communicate with customers offline .

5、 ... and 、 Retain

1、 Important time nodes left
1） The next day ： How much is left the next day . We can see that , Pull new on the same day , For example, we pulled 100 personal , To
The next day, there may be only a dozen , This is for the next day .
2） Zhou Weisheng ： Generally, customers use a Products , If he carries out a complete use, his experience cycle is about a week
Or a few days , The more common one is Zhou Shuai . Of course , If your product experience cycle is 14 Days or 3 God , Let's decide 3 Day or 14 Daily retention . Customer experience , The product meets his needs , Will stay steadily , If you are not satisfied, you leave .

2、 Retention methodology
If we do a lot of innovation 、 Channel optimization and so on , It may improve the retention rate , But can this retention stay , The core problem is whether the product function design can meet the core needs of customers . If it can be satisfied , Can we go further , Can the design of our product be better 、 Faster 、 It is more convenient to meet the core needs of customers , This is the second point .
1） Analyze according to the time of obtaining customers
2） Analyze according to user behavior
3） Analysis of the use of different modules of products by different groups

3、 Retention is the core of growth
1） Improve the possibility of users recommending products . A very simple truth , The larger the retained base 、 The more core users , The more people and possibilities are recommended .
2） Extend the life cycle of users using products , It means that users will stay on our website longer and longer , The greater the possibility of bringing us wealth .
3） Improve the probability of paying . Upgrading refers to the user's initial use of free primary functions to later complex functions , Finally, pay to buy the full function .
4） Reduce back to this cycle . When more and more people stay , Buy... For a fee , Or recommend it to others , Can make money return to our company as soon as possible .
5） It is conducive to reducing the cost of getting customers 、 Increase channels . If the user's life cycle value ( LTV ) And customer acquisition costs ( CAC ) The proportion of 3:1 It's very good .LTV The higher the words , We can increase the investment in market channels , Because the more channels, the better customers can be found .

reference ：《 The first data analysis manual of Internet growth 》