What big banks of today have that places them at the forefront of the digital revolution is an overwhelming amount of data. Many traditional financial institutions have already created big data operations that incorporate technology, data science and analytics departments in an effort to glean deeper insights into what customers need and want - because the huge volume of data that these banks have at their fingertips provides information of a depth and accuracy not seen to date.
For those institutions that want to respond effectively to digital disruption in the financial services sector and generate maximum value from the mountain of data collected, the next important piece of this puzzle is Artificial Intelligence (AI).
The reality is that customers are visiting branches less and less as they look for fast, simple and easy access to banking that adds value. With the FinTech industry and technology companies tapping into the banking market at an increasingly rapid rate, there is an opportunity and need for banks to evolve in a number of areas in order to provide significantly better customer experiences.
The advent of machine learning
In today’s world, we have seamless data input from devices that are recording and observing everything we do, with the capacity to store infinite levels of information in the digital cloud. This is the first key ingredient to developing AI.
The next key ingredient is machine learning. The ability of machines to learn from data, particularly with the recent advances in deep learning, has unlocked the ability of companies to build predictive models from large and complex datasets. Deep learning models, although computationally expensive, have demonstrated significant benefit when it comes to dealing with cognitive information such as image, text and voice data. Machines that can learn have the ability to reason on new or unseen data, thus creating AI.
Inevitably, AI will revolutionise nearly every aspect of human life. While neither AI nor big data are new terms, the value when used properly together will transcend the true power of both. For banks, this presents a massive opportunity to improve their customer insight exponentially.
The immediate value that AI holds for the banking industry is three-fold: it can significantly improve the scale, speed and quality of service to customers; it aids in the tracking and prevention of fraud and cyber-crime; and it unlocks massive scope for innovation.
1. Scale, speed and quality of service
Through automation, we increase scale and speed and can provide more personalised service to customers - for example, our wealth advisors can draw up-to-date tailored performance summaries for their clients, giving them the knowledge they need to make the best investment recommendations.
As machines aren’t subject to human working hours, we can also extend banking hours and enhance our responsiveness.
By reducing the wasted time and effort spent on generating the massive amount of reports required in the banking industry, we can free up staff to focus on creating more meaningful, richer face-to-face interactions with customers.
Finally, every call to the bank’s customer care centre can be analysed to gauge common questions and issues, with the aim of predicting and pre-empting these and ultimately getting to the point where customer service is intuitively handled by machines, faster and more effectively at any hour of the day. Currently, Absa uses an AI virtual assistant to handle basic queries, transferring to a human representative when it doesn’t have an appropriate response. The virtual assistant observes the way this is handled by its human counterpart, learning from this and integrating it into its database for future use.
2. Fraud detection and prevention
Through AI, we now also have the ability to identify fraudulent behaviour while it is happening, as well as identify what the next pattern of suspicious behaviour will be.
With every single customer call, transaction and interaction being recorded and analysed, a wealth of data is obtained and fed through the bank’s machine learning databases. Over time, and with enough data to sample, certain irregularities emerge.
The machines pick up these ‘exceptions’, which raise red flags. For example, at Absa, we recently noticed that multiple users were acting identically on different accounts and we realised it was fraud. We identified nine cases like this and reported it to the regulator.
3. Increased scope for innovation
AI makes it possible to build a world where virtual agents complete financial transactions and resolve complex user queries by applying deep learning, such as the ability to automate judgment on images through facial recognition or customer validation.
It also enables us to experiment with big data to find new discoveries. Using statistical techniques or machine learning to discover relationships between separate data points transforms the discoveries of those correlations into actual explanations of identified relationships. Imagine if we could have answers to business problems like, “What causes us to lose a customer?”
While there are multiple projects taking place behind the scenes, presently the face of Absa’s exploration into AI is our social media banking platform, Chat Banking via Facebook Messenger (a world first) and Chat Banking via Twitter. It’s cutting edge, it’s innovation in action and it has the potential to change the way millennials bank.
Some have questioned whether this particular investment is paying off when only the ‘early adopters’ are using it. However, tools like Chat Banking aren’t necessarily launched with the intention of getting everyone to switch over immediately - email and Internet Banking, for example, both took time to earn the trust of users. Rather, it is about adopting an agile approach in everything we do. We could wait two years to perfect it and launch something, but instead, we build it, launch it, assess who uses it and how, then learn from that and constantly improve on it, iteratively.
Coupled with Absa’s AI capability, this all forms part of a much larger picture and long-term vision. Giving customers something new to interact with provides a new input for data and a new avenue for learning, which will help to improve our future tech and innovation.
Achieving all this, however, depends on banks approaching AI as a business solution - rather than a technology solution - that converts big data into revenue and customer satisfaction.
When we look to the future and how AI is going to shape it, there are several changes we believe we’ll see in the next 10 years. The first is deep personalisation, where every customer will experience completely customised, holistic service based on their individual needs. AI will facilitate the scale and speed to make this viable.
Secondly, we will see prices driven down across the board, with volumes increasing exponentially as machines make it feasible to multi-task and perform business functions much faster than ever before.
And lastly, interconnected services will become the norm. We’ll see industries such as banking, telecommunications, health and wellness, travel and more start to merge in these integrated ecosystems. Some industries will die, giving rise to entirely new ones.
There’s no doubt that pressure will be felt in white-collar industries, and we’ll see a growing need to re-skill and redefine many occupations. When, for example, self-driving cars become standard, the motor insurance industry will either die, or evolve and change. It’s harsh, but it’s a case of adapt or die. As AI becomes ubiquitous, we’ll see many ethics and compliance struggles, but we’re optimistic about the future. We will see the dawn of shared growth and corporations collaborating towards a common goal. With enhanced productivity and true AI-aided multitasking capability, the economy will grow and jobs will be created.
We need to embrace this future or be left behind.