Forum Discussion
RFM classification help
Hi,
I am trying to do a RFM segmentation example for the marketing team.
I am able to do the segmentation using measures. Please see the sample file here.
My requirements:
1. Need to be able to summarise the customer counts by RFM Classification.
Champions | 1 |
New | 4 |
Loyal | 4 |
Hibernating | 6 |
Lost | 1 |
2. Need to be able to calculate which product/category drives the largest volume of sales for a given RFM Classification category.
Questions:
- Would it be better to move the logic of classification to database table? or Power Query/M?
- Would it be better to move the logic of classification to Power Query/M?
- Can the requirements be satisfied if I create Summary table (Summaize) in Power BI?
The actual data will have millions of Customers and millions of rows of transactions for any given month. Therefore I am looking for what will be the efficient approach.
Any suggestion would be greatly appreciated.
If the requirements can be done in PBI, I would greatly appreciate if anyone could show this in the attached file.
Thanks in advance.
2 Replies
- az38Community Champion
I think the best technique would be to "Split by delimiter" in Power QUery mode and get RFM Classification from cstRFMClassification field
and next there are 2 option:
1. simple scenario - just add Classification in visual and the second field "Classification" aggregated by Count (visual settings in visualization pane)
2. SUMMARIZE('Table';[RFM Classification];"Count";Count([RFM Classification]))
do not hesitate to give a kudo to useful posts and mark solutions as solution
- adudaniMemorable Member
Hi anandav ,
I am trying to achieve something similar.
I am curious about the output of the performance of your measures and what approach you took.
Link to the post for what I am trying to achieve with this segmentation.
https://community.powerbi.com/t5/Desktop/RFM-Analysis-F-amp-M-metric-grouping-along-with-Visual/m-p/3047732Any insight would greatly be appreciated.
Please let me know if you'd like additional information
Thanks in advance.