What is it about?

Digitalization is changing the shape of many industries and the way companies and clients interact. This digital revolution has been particularly relevant in the banking industry where the use of digital banking has become one of the most strategic channels used by bank customers. By employing a set of machine learning algorithms, this paper reveals that the digitalization process is originated from customers’ need to gain information about basic aspects of their banking accounts (e.g., checking their account balances), and this facilitates a transition to transactional services (e.g., transferring money). Once the initial adoption has taken place, the diversification of online and mobile services adopted by the customers becomes larger when they are conscious of the range of possibilities provided by the bank and when they perceive those options as safe. The adoption of non-bank payment instruments (e.g., PayPal and Amazon) happens when consumers are already diversified digital bank customers. Users of non-bank payment instruments seem to have previously reached a substantial degree of banking digitalization.

Featured Image

Why is it important?

Modern societies are undergoing a rapid digital transformation. A sizeable part of this change is related to the demand for financial services. Understanding the process of financial digitalization is valuable for the banking industry to design strategies that bring on board and retain digital users. It would help banks to obtain information on how they can face competition from new providers of financial services (BigTech and FinTech). Financial providers could benefit from the digitalization phenomenon by offering services that better match customers’ needs. In this sense, segmenting customers using similar techniques and data, would make possible to offer them more personalized digital services. Moreover, linking payments experiences to social media interactions could also be used to foster the adoption of digital payments. Additionally, policymakers may use this knowledge to implement more efficient policies to promote financial digitalization and enhance financial inclusion and literacy. To reach this end, this paper employs a machine learning approach to reveal the patterns driving the digitalization process and to offer a multi-dimensional comprehensive picture of the process by which bank customers become digitalized.


This publication helps explain how to make the transition from brick-based to digital banking services and how the population makes this transition. This is important for banks, for consumers and for regulators as financial services are a key ingredient of social inclusion

Santiago Carbo-Valverde
Universidad de Granada

Read the Original

This page is a summary of: A machine learning approach to the digitalization of bank customers: Evidence from random and causal forests, PLoS ONE, October 2020, PLOS,
DOI: 10.1371/journal.pone.0240362.
You can read the full text:

Open access logo


The following have contributed to this page