Using powerful powerbi Sangji diagram to express complex operation business flow

Want to make APP Same thing as WeChat , Can run small programs smoothly ？ | Experience will send you to Xinjiang 、 Huawei 、 Cherry keyboard ！>>>

On Sankey graph , Last time, we talked about the general construction method of Sangji graph at any level . The question left is ：

• Scenarios where traffic is not necessarily equal

• There are scenarios of cross layer flow

• There are backflow scenarios

In a real business scenario , There is such a complex structure , This article is to realize .

Business scenario

Take the retail scene as an example （ It can be extended to many fields ）, With AARRR For reference , There is probably such an effect ：

A lot of exposure on the line will spread the whole picture, causing other contents to be affected , Using logarithm to get ：

The business implications are very clear ：

• Online exposure gets browsed ;

• Browse to get online registration ;

• There will also be registration offline ;

• Sign up to become a member ;

• Members generate purchases ;

• Not registering as a member can also generate purchases ;

• It may be converted into a member after purchase ;

• You can also recommend after purchase ;

• Recommend a new registration .

In this business scenario , It's about the most complex use of the Sankey diagram , At the same time, we need to deal with ：

• Scenarios where traffic is not necessarily equal

• There are scenarios of cross layer flow

• There are backflow scenarios

Then let's see how this is done .

data

If you express this directly with data , You can roughly get ：

stay PowerBI It is obviously more intuitive to use Sangji diagram in China .

Prepare dimensions

use PowerBI DAX The dimensions of preparation are as follows ：

SanKey_Source_X_AARRRR = 
SELECTCOLUMNS(
{
    ( "0.online_show" , " Exposure "),
    ( "0.online_view" , " Browse "),
    ( "1.online" , " Register online " ),
    ( "1.offline" , " Offline registration " ),
    ( "2.reg" , " members " ),
    ( "3.buy" , " Buy " ),
    ( "4.share" , " recommend " )
} , "SourceCode" , [Value1] , "SourceName" , [Value2] )

as well as ：

SanKey_Dest_X_AARRR = 
SELECTCOLUMNS(
{
    ( "0.online_view" , " Browse " ),
    ( "1.online" , " Register online " ),
    ( "1.offline" , " Offline registration " ),
    ( "2.reg" , " members " ),
    ( "3.buy" , " Buy " ),
    ( "4.share" , " recommend " ),
    ( "5.new" , " New registration " )

} , "DestCode" , [Value1] , "DestName" , [Value2] )

You can find , Here we use  Code  and  Name  In a way that will KEY Separate from the display content .

The main idea here is ： Data as source and destination must be prepared in dimensions . Later, measures are used to control rationality and combination and calculation .

Implement the calculation

use PowerBI DAX The implementation measures are as follows ：

SankeyX.AARRR = 
VAR _source = SELECTEDVALUE( SanKey_Source_X_AARRRR[SourceCode] )
VAR _dest   = SELECTEDVALUE( SanKey_Dest_X_AARRR[DestCode] )
RETURN SWITCH( TRUE() ,
    _source = "0.online_show" && _dest = "0.online_view" , 8045 ,
    _source = "0.online_view" && _dest = "1.online" , 60 ,
    _source = "1.online" && _dest = "2.reg" , 60 ,
    _source = "1.online" && _dest = "3.buy" , 5 ,
    _source = "1.offline" && _dest = "2.reg" , 25 ,
    _source = "1.offline" && _dest = "3.buy" , 10 ,
    _source = "2.reg" && _dest = "3.buy" , 30 ,
    _source = "3.buy" && _dest = "2.reg" , 15 ,
    _source = "3.buy" && _dest = "4.share" , 15,
    _source = "4.share" && _dest = "5.new" , 55 ,
    BLANK()
)

In practice, , You can replace the constant values with more specific measures to achieve dynamic calculation .

Core business scenario

If you don't consider the exposure scene , The sanguingi map can be more refined , give the result as follows ：

Combined with the actual operation scenario , Look at the picture like this ：

about 100 Members , Yes 60 people From online registration , Yes 25 people From offline registration , also 15 people From the return conversion after purchase .

Does not include reflow conversion members , There will be 45 people （ Registered at 85 people ） Complete purchase .

among ,15 People will continue to recommend , Bring after recommendation 55 People are newly registered as members , The conversion ratio is approximately 3.

Suppose the model is stable , We can continue to periodically deduce the realization effect after any stage .

summary

This article uses free PowerBI Visual objects sangee map , Combined with the growth of hackers AARRR Model and actual business scenario simulation , More complex implementation and effect display are given .

thus , The implementation of a very complex sangee diagram and can be used for business expression , It includes ：

• It can be any level of flow ;

• Scenarios where traffic is not necessarily equal ;

• There are scenarios of cross layer flow ;

• There are backflow scenarios .

Using the ideas of this paper , You can immediately implement your own business presentation , Try it .

After subscribing to BI Zorro taught 《BI In progress 》 Course area , In addition to downloading this case , You can also watch videos to explain .

Learn systematically PowerBI Please be ready to , There will be many course updates this month , We must pay attention to the official account .

Let data really be your power

Create value through simple and easy with fun by PowerBI

Excel BI | DAX Pro | DAX  Authoritative guide  |  Offline VIP Study

Scan the code and PBI Elites learn together , Verification Code ：data2020

PowerBI MVP Take you Correctly and efficiently Study PowerBI
Click on “ Read the original ”, Start immediately

This article is from WeChat official account. - PowerBI The comrades in arms Alliance （powerbichina）.
If there is any infringement , Please contact the [email protected] Delete .
Participation of this paper “OSC Source creation plan ”, You are welcome to join us , share .