It’s been clear for some time that AI is set to transform just about every industry, especially banking and financial services. Indeed, AI is already widespread in many critical back-end processes and operations within the banking sector such as know-your-customer (KYC), credit profiling, anomaly and fraud detection, cyber security and risk assessment, as well as improving the efficiency of customer service functions.
However, with the exception of the offerings of a few innovative challengers like Starling or Revolut, many customers have yet to directly see AI improve their personal day-to-day banking experience in relevant or groundbreaking ways. Historically, AI has had a big appeal for amplifying back-office functions in banks: scaling those functions to deal with big and complex data sets.
But just like any good back-end process, they are intended to be things that a customer doesn’t see, and typically shouldn’t have to worry or think about. Therefore, many customers of banks aren’t consciously interacting with AI or seeing it provide a radically improved experience for them – at best, they’ve only noticed a slight increase in efficiency. This coming year, I think that’s going to change, as AI is leveraged more widely to help customers get real-life value from banks making better use of their data.
The top 10 Fintech companies disrupting financial services
Personalisation beyond selling
One of the other, and more subtle, back-end uses for AI in banking has been in “personalisation”; that is, taking data on a customer’s saving and spending decisions and offering them new insights, as well as more relevant services that they’d find interesting. While it’s definitely more convenient for customers to only be offered products or services that are of use to them, this moreso strays into the territory of effective advertising to sell additional services.
Some of the aforementioned challenger banks have begun to rock this trend, and go beyond “personalisation” as just a paradigm of trying to sell more appropriate products and services. Instead, they’re using AI to provide customers with a genuinely personalised experience of banking, taking into account changes in life stages and other knowledge about real-life events and activities of human beings.
But what does “genuinely personalised” banking look like? To answer that, we should compare these challenger banks with “business as usual” in the sector. Currently, most traditional banks still treat their online accounts as a digital version of a traditional balance statement. The odds are that your bank’s online account only provides a simple, itemised list of your ingoings and outgoings. If you want to calculate how much you spend, how you allocate that spending, set a realistic budget for next month, or estimate how much you might be able to save in an average month, it’s often the case that you simply will have to trawl through your statement yourself and do the hard calculations.
Want to easily see how much goes out on your subscription services or other automatic charges versus incidental spending, and perhaps manage some of those financial commitments? The data is all there, but has often yet to be transformed into easy-to-understand interfaces that can help consumers or small business owners get their finances under control.
This ends up being burdensome for people. And it’s also quite unnecessary. As your bank holds all your data regarding your transactions, it should be capable of doing any categorisation and analysis of your spending history itself. And that’s where AI really has space to shine in extending personalisation into a tangible reality for customers: providing analytics and insights, but making them accessible to what Steve Jobs used to often call ‘mere mortals’.
The challenger banks have shown it’s possible to furnish their accounts with AI to categorise and break down a customer’s spending. This doesn’t just save customers a lot of time they would have spent budgeting or balancing their checkbooks, but it can also draw on the bank’s own insights to help in delivering real-time analysis that goes above a traditional budgeting exercise.
For example, such personalised analytics can provide month-on-month comparisons between spending across different categories (rent, food, transport, etc.), helping customers see facts or trends about their spending allocation that they could try to change for the better. But there’s also the more humanistic element that banks need to start addressing – such personalised analytics also have room to make inferences from numbers to understand and anticipate people’s life-stages and real-world needs. To work, this must extend across the whole suite of financial products a person may own – whether they be insurance, ISAs, mortgages, or beyond – and be able to provide customers with comprehensive guidance as to what their financial ecosystem should look like.
Keeping it human
To work, though, these powerful ML capabilities have to be presented to customers in an intuitive way that allows them to quickly comprehend purpose, relevance, and how to take action. Clever analytics capabilities don’t really mean much if their results aren’t comprehensible to users – which means that nailing the interface and interaction design for personalised analytics may be almost as critical a task as developing and deploying the underlying models that will power them.
This focus on utility, ease of use, and end-user value also needs to likely extend to a final domain that I think ML will become increasingly important to banks in 2022 – protecting cyber security. In practice, people’s bank relationships have now become the “hubs” of their economic and personal safety and security well-being, for now and into the future. As banks increase their own cyber security for near term and future risks, perhaps they can extend this value and expertise to their customers as well, be they consumers or small businesses. Given all that, AI has huge scope to help people navigate risks associated with security.
At the heart of all the above innovations sits a simple premise: providing the insights and power of AI directly to customers in ways that are meaningful to their lives, their activities, and their current real-world priorities.
As opposed to streamlining back-end processes, banking in 2022 is going to take AI directly to customers and use it to radically overhaul and improve their experience. And, ultimately, this may lay the foundation for banks and the financial services sector being able to directly assist customers through AI truly acting on the user’s behalf, finally ridding them of having to deal with the more mundane and tiresome parts of their financial lives.
