Personalized product recommendations

Banco Santander, S.A. is Spain’s largest bank and the main component of the global financial group, with some 190,000 employees, 13,000 branches and €1.5 trillion in total assets (2016).

The challenge

The bank was looking for a solution that would provide recommendations for personalized marketing. The ultimate goal was to spot those clients who are most likely to buy a particular service, thus setting the bank up to close new deals.

The solution

deepsense.ai designed a machine learning model that creates personalized recommendations, showing what should be proposed to a given client based on what type of customer he or she is. The model also recognizes peoples’ different needs. For example, a sole proprietor is definitely a different type of client than a young university graduate. The personalization created with machine learning techniques also takes into account the client's history.

The algorithm automatically searches the database to find clients fitting a particular profile based on a given customer and considers their behaviour when making predictions about their propensity to buy.

The effect

The model had 93% recall, which is 60% more effective than traditional product recommendation techniques. With the tool, the bank has access to the information about customers interested in particular products, so the sales team works more effectively, closing more deals and selling more products with less effort.