Subscribe to DSC Newsletter

How Do i Develop a CLTV (customer lifetime value) Model using SPSS Modeler? I have all the transactions data of retail cusotmer base.(Demographic and Financial data). As per my understanding reading through some articles..........I got to know that a tenure model and Average revenue model needs to be built which will give a CLTV for each customer? But How do I proceed now to built these models?? Please help :)

Views: 2140

Reply to This

Replies to This Discussion


let me show one basic approach based on churn rate.

Lets have data at least for 2 years.

Average yearly churn rate  = percentage of active customers from year 1 who were active in year 2.

If you know that you loose 10% customers yearly - how many years will average customers stay active? Think about the formula.

You may go deeper to customers level - predict individual churn rates. Basic variables are such as '# of purchases so far', '# of month since last purchase', '# of categories purchased' and many other. Target variable is boolean - will be customer active in year 2? Then apply a formula to predict individual tenures.

Average revenue spent can be predicted in the similar way. Start with simple statistics then build predictive model for individual revenues. 

And finally: CLTV = actual revenue + # of periods customers will stay * # average revenue

Hi Thanks a lot very much ..........this really helps!!.....if any doubts again........will get back to you!!

I am retired prof in Statistics from a business school I have read on CLBtV. Under CRM.Very familiar with SPSS.
I am planning to gets a Research project from a bank where i like to use CLTV. I am yet to have discussions to find what sort of Data of retail customer base I should ask for
give me reference of articles which wd enable me to use SPSS

Hi Hema,

it's a bit trickier in banking. There are more products, all with different value and life cycle. 

This approach could work:

1. Define products in scope

2. Ask about data for all products preferably on customer level.

3. Data should have attributes as for example:

- date of activation

- date of deactivation

- average account balances

- interest rates

- revenue and costs on product

- delinquency status

- and much more

4. The success depends on your ability to prepare data into right structure. SPSS eats flat files or tables.

I wish you good luck! And if you would need some support, just let me know.


We are working on churn predictive tools with prorpietary algoritms. if your company is interested in developping some specific tools we can speak.

we have a team of researchers, statitician, churn manager and consultants on analyttics and churn; mastering this aspect at very advanced level.

If any interest contact me

Best regards

Remi Mollicone

Innovation, Alliances/Partnerships, Business Development
[email protected]
Skype: remimollicone7
Tél: + 33 630 729 013


On Data Science Central

© 2019 is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service