Categorize the customers

Feras95p

Dear all,

I have a data for 35,000 customers, the data include that

The date of the first order.

The date of the last order.

The average days between the orders.

The total number of the orders.

The total dishes of the orders.

The total cost of the order.

These are the data of the customers which are identified by their mobile phone.

I would like to do on some how an analysis that give me the percentage of each customer that a customer will be return if I did some marketing to him. I don't know how I can explain for you but I want to know for each customer in a percentage that the probability of him will be high or low to be returned.

I have attached the data without the phone number of the customers due to privacy to be more clear for you.

Thanks in advance for your kind support.

Sample.xlsx

Accepted answers

atcodedog05

Hi @Feras95p ,

Here is a case study which i felt would be very helpful which demonstrates the RFM model and tags them as Potential loyalist , Recent customer. You might want to calculate the RFM matrix to proceed with it.

https://community.alteryx.com/t5/Alteryx-Designer-Discussions/Customer-segmentation-in-Alteryx/td-p/157848

To help you with calculating each of R, F, M value. Here is a link of a resource.

https://www.coursera.org/lecture/foundations-marketing-analytics/computing-recency-frequency-and-monetary-value-with-r-recital-1-rEvkZ

And here is a blog on Principal Component Analysis cluster analysis to group customer of similarities.

https://community.alteryx.com/t5/Data-Science-Blog/Tidying-up-with-PCA-An-Introduction-to-Principal-Components/ba-p/382557

You might want to take a look at it too. And try to visualize it. It helps in grouping and targeting.

Thank you for sharing this scenario. It was a great exposure for me too.

Cheers and Happy Analyzing : )

All comments

atcodedog05

Hi @Feras95p ,

Your proposed scenario is a strategy based situation. There is no direct answers to it.

There are multiple ways how you approach them.

One would be to go by Recent Frequency Monetary model of customer analytics.

One more being you visualize the data on different breakdown which might expose patterns to you which might be helpful.

You can use clustering algorithms like PCA to group customers in group and strategize how to target them.

Building Time series model on customer groups would also help.

Its a lot of trial and error strategizing to arrive at the desired output which you can take action on.

I hope it was of some help. All the best in your venture.

Cheers and Happy Analyzing : )

Feras95p

Thanks for replying, I have just searched about RFM module which exactly what I'm looking for and I think with the data I have it will help to have a good module. Can you help on how I can do that in alteryx?