Multi Layer Sanky Chart

[Amar-Agnihotri]

Hi All,

I want to create a multi layer Sanky chart for Sales Pipeline Analysis of Salesforce Data.

I want a Visual to be look like -

Requirement Description- 

• Will likely need to utilize the Opportunity Snapshot table
• Pipeline - Start: should show the full pipeline value (incremental ACV) at the start of the quarter that is made up all opportunities with a close date in Q3
• Pipeline - 45: should show the value at the 45 day mark
• This would need to include opps that have already closed won and still open at 45 day mark
• Closed Won: should show value of opps that were closed won in Q3
• Closed Loss: should show value of opps that were closed loss
• Pushed: value of opps that were once set to close in Q3 but now have close date outside of Q3
• Pushed Next Quarter (orange): value of opps that were pushed in to the next quarter
• Pushed 2+ Quarters (yellow): value of opps that were pushed out farther than the next quarter

• Width of bars should be based on Incremental ACV
• Ideally we would have a way to show number of opps as well
• Filters/Slicers that should work
• Region
• Line of Business
• Opportunity Owner

My Data Model Looks Like --

The table That need to be used for the sanky would be Opportunity Snapshot because Opportunity Snapshot contains the following columns -

The Pipeline Start Date

Opportunity Stages - Closed Won, CLosed Loss ( Pushed is not Present in the Stage right now )

Incremental ACV

Last two pillars Push (Yellow Color) and Push(Orange color) are not present in any column of Snapshot table. 

Now can anyone suggest the structure of Sanky table . It will help me to do transformation.

Thanks 

[lbendlin · Accepted solution]

Your data rows need to contain "From" and "To" columns. You need to prepare these values outside of Power BI. Usually you would use the Opportunity Field History object data from Salesforce.com directly. It has these fields.

Once that is in place you can use the normal Sankey visual.

[Amar-Agnihotri]

Thanks @lbendlin