Machine Learning is most commonly used in mid and back-office operations (AML, fraud detection, credit scoring etc) where it is used to reduce operating costs. However, it has not yet conquered the front office, the realm of user interaction, where it can unleash a substancial increase in revenue.
"How?" you ask. By delivering on the Holy Grail of personalisation. Data relating to spending (transactional) behaviour is the most valuable asset that banks can leverage as a competitive advantage with respect to new entrants. But this data needs to be put to work.
It needs to help the bank:
- to know their customers in a much richer way than traditional high-level segmentation,
- to help those customers improve their financial life, regaining trust, and
- to monetise (let’s not be naive) the interaction at the right time in the relationship.
Personalisation delivers customer interaction and satisfaction, generates brand loyalty and an increase in share of wallet but it is not simple to achieve. It requires to the bank to know itself well before organisational change can happen. It requires a complex Machine Learning event-driven infrastructure, models that look at the customer from a human angle, and not through a finance lens, and data quality. It requires a good look at the products. And it requires a new mindset.
This article presents the steps to achieving real personalisation in banking, and provides the recipe for its successful implementation. And remember, almost paraphrasing ABBA, knowing you, knowing me, there is nothing banks cannot do to improve the financial life of their customers and be more profitable.
If you like what you've read, vote for the full article to be included in the upcoming AI Book !