A good digital transformation strategy in 2024 should consider concept to execution across every aspect of the business. But before any bank gets started, they must ask themselves: what are we trying to solve? I distinctly remember a time – a couple of decades
ago now – when a few banks decided to automate anything and everything they could. A huge number of resources were used to chase these changes, but some projects did not have the desired return. In addition, while resources were directed to automating smaller
processes, significant changes could not be made. When establishing a digital transformation strategy, the goal of optimisation for growth must be included. In 2024, the goal, the measure of success and the list of priorities must be clearly defined, and technology
can support this.
What is the role of technology in the optimisation of banking processes?
Banks should not start with technology and decide how they should use it, the problem must be defined first and foremost. We are all excited about AI as an industry, but not everything that needs to be automated is going to require AI. Five years ago, everything
was about robotic process automation (RPA) and organisations rolled out RPA solutions for processes they should not have used that technology for. Next year, it could be better integrations, APIs, AI, God forbid, it could be blockchain. Any technology could
be a potential way to resolve a problem, but we must define the problem before defining the appropriate technology.
For instance, if growth in deposits is the biggest problem a bank is experiencing, and they rely on branches to drive customer retention and acquisition, technology can help optimise this process by expanding their efforts within digital. Ultimately, there
is no single technology out there that is specifically designed to optimise for growth.
What about the role of AI?
We need to remember that AI is not a single technology. There are various flavours of AI that we’ve already used for nearly a decade in underwriting, fraud, and identification. These applications rely on predictive AI which is very different from what is
exciting for the industry today.
Generative AI, using large language models and natural language processing, is a different discipline of AI which is dominating headlines today. However, advances in predictive analytics and AI can be used to improve how machine learning is applied, to make
decisions quicker, and to be more thorough about crunching through data faster.
Evidently, there is a journey that companies need to take on when to apply AI. While organisations are excited about AI, they need to understand what it is, how it works, and what it is built for. The most important factor to consider is where the data is
coming from. We may already have a good handle on data, we may be scrubbing it appropriately, and we may be looking at bias, but the models need to be trained with trusted data, as well.
If you’re going to deploy something like generative AI for customer service, the data that is used for learning has got to be appropriate for that use. The other piece is regulation, particularly for generative AI. The EU is further ahead of both the UK
and the US, but when the founder of Open AI speaks to Congress, we all fall at his feet to hear the cool things he has to say, and not to understand the issues. This is a common failure. Before deploying AI, we need to understand how regulation is likely to
impact AI, and to do that, we need to know what AI does, where the data is coming from, and the bias we need to test for. That’s where we may get in trouble in financial services.
Large language models have a lot of promise, and I can see it addressing the most basic queries from a consumer’s perspective. Both Bank of America and Chase have chatbots, but they are nowhere near the caliber of ChatGPT or other generative AI models. It
can only answer the absolute, most basic questions, that are worded in an exact way. In fact, combining the capabilities of a ChatGPT with a human is even more exciting to me. For instance, we can use AI to help analyse sentiment and emotion while serving
up possible solutions for a contact center agent to offer a customer. The augmented human is more exciting. Today, a chatbot answering questions and getting 80% of it right is not good enough. It must be 99.9% — I can’t see ChatGPT doing that right now. I
think, the best customer experience can only be provided if there is a human in the loop.
How have branch closures impacted consumer trust and driven them from regional to larger banks?
Despite the events of 2023, banks have done a good job of continuing to prove that they can be trusted to keep consumer funds safe. However, banks believe that trust means that people trust them for all kinds of financial services, and I think there’s a
misunderstanding there. For example, trust isn’t there for advice on financial matters, and there remains a gap or an opportunity for banks. In a banker’s view, a branch was needed because consumers would want to visit the physical location to get advice,
but today, very few people would go to a person sitting behind a desk to get advice. Funnily enough, the folks who sit behind a desk don’t have the tools to give the level of advice a customer may need. That’s not what their job is anymore, and the bank isn’t
equipped to deliver this information in a way that is digestible for today’s consumer, which is why they go elsewhere.
The fact of the matter is bank branches are closing across the globe. Yes, at times there are news reports that a particular organisation is opening new branches. I would say that this is a very expensive marketing play. Often times large banks are opening
10 branches in a specific area where they have not had a presence, but at the same time closing 50 branches elsewhere, so overall, there are fewer branches than there were before. This is because of changing consumer behaviour, particularly the increase in
use of digital services. As consumers, we are remote, we are automating activities ourselves, like paying bills, and that’s how we manage our finances. However, there is still a need for a branch, primarily for services that cannot be automated, or better
yet, for the products that the bank has not automated yet.
There are many banks that have great technology and digital processes, but they introduce a physical step. Having to physically go to a branch to give them your wet signature happens more often than you would think. If you didn’t have a branch, where would
you go to do that? Banks are not deploying digital technologies in a way that is looking at how consumers could use them to reduce the activity at a branch. There are also some types of customers for which a physical experience makes a lot of sense, for example,
small businesses that are managed by entrepreneurs that are not professional business managers. These individuals are practitioners, but need to be handheld through banking processes, and a bank branch is the best place to do this.
It's all about convenience, and this involves providing both physical and digital services. In some cases, we have seen digital-only banks open physical locations, just because of convenience. Banks must be prepared to personalise customer experiences and
think of the segment of one. Given the capabilities of AI, personalised customer experience is the future. The industry is not far from being able to offer that level of personalisation, even when the bank does not have a lot of data on someone. Taking a leaf
out of the likes of Amazon, Apple, and Google’s books, which are great at personalising experiences, banks must learn how to brand their services and do a better job of presenting themselves as being truly customer centric.