The 24th April was a very busy day for me.  Firstly I was speaking at Henry Stewart Evans' Marketing Analytics event.  Unfortunately, I missed most the presentations but, I did see a very interesting one from GAME and the work that they are doing with 5one on segmenting their customers to 1. get a better understanding and 2. market to them as individuals.

One of the questions that I was asked during my presentation (and it was infront of my biggest audience since I was doing Shakespeare in school plays!) was how to merge online and offline data.  Now, in financial services I just cannot see a way of doing it.  However, with retail outlets like GAME there is a great opportunity to understand the behaviour of online and offline users because they have a loyalty card. 

I'll get in touch with GAME and see whether I can post a copy of their presentation on my site and perhaps get an interview with the speaker.  He was rather impressive.

People will now say that financial services organisations have branches which are classed as high street outlets.  However, what they don't realise is that a customer is identified by a number of different identifications.  They have their card number (perhaps several) and a customer id (which they may also have several of and it is unlikely that they are going to know this). 

However, the main thing is I would question is the benefit of spending time trying to understand this behaviour.  A customer goes into a branch to either pay in a cheque or pay a bill; and I'm sure that anyone in financial services would agree that the cost of this transaction is far higher (given retail rent, salaries, insurances etc) than if the transaction were to be completed online.  However, if they were to try and migrate the customers to the online channel for such transactions then surely they would be losing the opportunity of identify a cross sale opportunity - because lets face it - financial services products are hard to understand.

So, this raises the question of what segmentation would be useful to financial services analytics teams?  I'll write some suggestions down here after I hear your own thoughts.  Please leave comments