Can AI supercharge your bank’s digital transformation strategy?
By John Spooner, head of Artificial Intelligence, EMEA at H2O.ai, an AI technology software company
Though some banking organisations remain confused about AI and its relevance, a rapidly changing new set of post-COVID realities means it has to be central to digital transformation initiatives, warns AI software expert John Spooner.
Even before the start of the (sadly) still-ongoing global health crisis, banks were moving ahead with what we may not have then called ‘digital transformation’, but for sure was: opening up multiple new channels like online and mobile, with their eyes always on the rise of the non-traditional bank challengers and payment apps.
And even before the pandemic, they also faced tough trading conditions that had eroded profitability. Then, one sort of animal somehow crossed its DNA with another, and everything changed.
To cope, banks scrambled their fighter jets and lots of amazing, and almost all of it successful, pivoting to digital went on (which I am happy to acknowledge as digital transformation). A new report from IT analyst firm Juniper Research says the majority, 53%, of the world’s population will be using digital banking by 2026, double the number now—and that by 2026, 4.2 billion people will have access to banking-as-a-service online.
But I’m afraid to say that now is no time to sit on your digital laurels. All over the world, COVID fundamentally changed the way people spend money and the way they interact with their banks (see MasterCard’s recent decision, sparked by the unstoppable rise of chip and PIN to make small payments instead of cash, to ditch the magnetic stripe on its plastic).
But our spending is now going into a post-COVID phase, and we’re going to see, again, a change in activity: we don’t know what that will be, but it will certainly be very different to pre-COVID and very different from during COVID. Any applications for predicting behaviour and activity need to be looked at again to ensure any and all predictions can be constantly updated to track the adjusting of customer behaviour over the next few weeks and months of post-COVID life.
The lesson has to be that consumer (and business) behaviour change is now a constant—and that is a statement that, as I will show, applies to everyone who has invested in data, and the best mechanism for understanding and predicting trends out of that data, AI/Machine Learning. Here’s how:
Addressing the Fraud Challenge
A lot of the fraud models you have in place are now probably not fit for purpose: either because they were still using older rule-based methodology, but also—and this is critical—even if you have machine learning at the heart of your fraud and AML (anti-money laundering) processes, those machine learning models are no longer based on the current reality. As a result, your fraud-fighting systems will soon start to generate a large number of false positives, which will lead to increased cost to do a manual check of those false positives. In addition, customer satisfaction will get dinged as well, because if their card or account has been stopped too often with no fraud having taken place, they’re going to feel frustrated.
How AI can help
If you haven’t taken the step to use AI, now is the time. Why: it’s your best way to increase the speed and accuracy of checking transactions.
Addressing Optimal resource deployment
What is the demand here on this resource? How is that going to shift? Do you want everyone back in the office? Do you put the people in the call centres or your branches? Will people coming off furlough still be needed?
How AI can help
As stated, to remain competitive you will have to be able to quickly adapt, based on how customers are going to be changing their behaviour. So, as the picture becomes clearer of what customers are going to start to do, you’ve got to have a mechanism that can build accurate internal sourcing and deployment models and quickly deploy them—and just as importantly, having a feedback loop to make sure that they are changing and adapting on the fly as customer changes and adapts after.
Making hybrid working work
Some banks say they want everyone back: some banks are tag-teaming and suggesting a flexible model; some say they want to be able to offer employee choice.
How AI can help
Looks like there’s going to be a lot of new, different ways of working here. To work out what’s best for both your customers but also your team and internal staff, you’re going to need to base your models on accurate data. So, in the same way you’re now used to analysing customer behaviour, you can also be analysing employee feedback, too, and this can be at the level of tens of thousands of people across many sites and geographies. How do you capture all of the feedback about the new changes quickly and then adapt?
A range of tools and techniques could help here, including sensors and desk usage data to map if people are behaving differently within the office, are they interacting differently, or to see if they are going around the building in different ways? What impact is that having on office capacity and safe usage, e.g., are they having conversations safe spaces away from each other, is everyone vaccinated and so on? An AI platform for great, real-time driven internal employee-based apps will be a huge aid.
A word to the cautious
These are just some of the ways that AI can help banks gear up for a changing present and immediate future, bolstering their pre-COVID and post-COVID digital transformation successes. But can I finish with a hopefully reassuring note to what we might call the AI laggards?
Those organisations that don’t have AI in any serious fashion yet should bear in mind that everyone’s going through a process of either building or rebuilding their models here. So, if you haven’t started on your journey, now’s a really good time to start, because no-one’s got a head start. Those organisations that have invested in AI to some extent need to start again, to be honest, as their pre-corona models can’t now predict the future well at all.
So, to sum up: accept that COVID-era change hasn’t ended, and that by embracing change and using data at scale is your best way of learning how to not just live in challenging times, but thrive in them.