Last week we explained the high-level clustering process that underpins FUSION, our SaaS, which segments data into clusters of customers exhibiting similar (financial) behaviour, by using an Unsupervised Learning algorithm.
Whilst FUSION can be applied across numerous different industries and for different purposes, this blog covers its application in the Financial Sector and, in particular, the segmentation of customer transaction data.
Financial Institutions, like any business, aim to maximise the value of their customers throughout the customer’s journey. This starts with marketing to and recruiting the right ‘type’ of customers to ensuring they remain loyal and high-value adding throughout their lifetime.
As mentioned last week, FUSION assigns each customer a similarity score within its cluster and enables organisations to identify customers with similar behaviours. This can be good or bad behaviours (e.g. customers making regular high-value purchases over time, defaulting on loans, or engaged in criminal activities).
By being able to identify, at an individual level, which customers behave like their peers and also display similar demographic features, this offers new insights across business units on which customer type to attract, nurture and service to maximise the customer value over their lifetime.
On the flip side, FUSION also singles out outliers or customers with spending patterns deemed to be ‘not normal’ in comparison with their peers’ behaviours.
For Financial Institutions this provides simple and effective means of clearing low and medium risk customers, whose behaviour is consistent with their peers, when performing KYC or transaction monitoring activities. To the same extent, it will also flag any outliers or customers that need further investigation as their financial behaviour appears abnormal.
In next blogs, we will cover how FUSION is key to detecting Financial Crime in Financial Services in more detail.
Get in touch ASAP if you want to book your demo of FUSION.