The AI revolution in finance seem to keep going without losing steam. Multiple start-ups are launching and those who have already entered the financial fray are getting more exposure.
A PwC report states “FinTech will drive the new business model” as the number one technology force needed to compete in the financial industry in 2020 and beyond. In addition, it also states that “Digital” will need to become mainstream.
Often, what many imply when they talk about digitalisation and FinTech is the power of AI. AI needs to be understood loosely here, and it can sometimes imply machine learning or data analysis, but it is always the power of data science and the information that we can capture from data.
This is not new; many financial institutions have already started their digitalisation journey, as an early 2019 Deloitte study, summarised in the chart below, reports:
Yet, the biggest challenges that organisations face in turning digital and, most of all, in embracing the AI revolution, is about the data.
What we keep hearing from data scientists and data science organisations all over is that data is hard to get, data is siloed, fragmented, dirty. This imply that most of the time and effort is spent in actually making the data usable. A Forbes article states that about 80% of the time is spent on cleaning data and the global online community CustomerThink states that 85% of AI projects fail and 96% run into problems with data quality, data labelling, and building model confidence.
In addition it is more difficult to define success metrics for AI projects than it is for regular engineering projects, and ‘agile’ does not work the same way.
So why do we insist on AI given all the problems?
An October report from West Monroe Partners, for example, found that digital maturity is linked with higher revenue growth, while a recent series of studies by McKinsey & Company revealed that increased digitalisation could boost Mexico’s GDP by as much as 15% by 2025, showing how even large countries, not just financial companies, can benefit. Therefore, despite the drawbacks, investing in digitalisation and AI is strongly associated with higher revenues and reduced costs.
In fact, another McKinsey report states that companies that have adopted AI have seen an uptick in revenue in the business areas where it is used, and for 44 percent of companies AI has reduced costs.
In addition, AI provides data justification for any business objective, can bestow those business objectives with streamlined processes and can create automated monitoring of any environment change that may disrupt those processes.
The above PwC report also states that “‘Customer intelligence’ will be the most important predictor of revenue growth and profitability” and “Advances in robotics and AI will start a wave of ‘re-shoring’ and localisation”. This will improve the labour market and produce more trust in the financial institutions utilising AI. As the same reports goes on to state: “When ATMs were first introduced, many customers refused to use them. Gradually though, after time and training, they came to see that ATMs could offer a better service experience. And trust followed”.
A VisionCritical article clearly defines what we mean by ‘Customer intelligence’ and how a deep understanding of our customers can drive revenues and sales. “In the age of the empowered customer, companies must move beyond traditional market research. Their challenge is to harness intelligence about their customers more quickly and comprehensively. By understanding the empowered customer, a company can make effective and intelligent business decisions”.
It is clear that a deep understanding of customers can only be achieved through AI and big data. As the above VisionCritical article says, “If they’re unhappy with a product or an experience, customers will tell you so, and they will expect a response. Companies that neglect to engage with angry customers risk a public relations disaster”. But customer engagement is also more and more being managed through AI, whether it is through a chatbot or a mobile app. Two years ago Gartner predicted that 25% of customer service operations will use virtual customer assistants by the end of this year and by 2022 two thirds of all customer experience projects will make use of IT.
What this trend implies is that while the customer becomes central, the customer experience, as well as the customer service associated with it, will become more and more AI based rather than human-based. The face of the company, and of the financial institution, will not be the physical bank and the tellers in it anymore, rather it will be the mobile app and the AI-based experience driving it.
Moving from the financial sector to the insurance industry, if we think about the AI revolution in car insurance, we should understand and expect that the real revolution is not the one currently underway, which is about calculating better premiums or rates, but rather it is the revolution that will arrive with autonomous driving. Autonomous driving will completely change the way the car insurance industry will need to operate, since it will not be insuring the person anymore, but it will have to insure the car manufacturer or the company that created the AI behind the autonomous driving. Cars, not humans, will be insured. AI and not people will be insured.
Financial companies, as well as other major customer-facing companies, will be presented with the same dilemma and face the same fate; their customers will be, in a sense, their insurer. As such, customers will not be evaluating a financial company based on the people who work there or its assets, but rather they will judge them based on the AI that will be driving their customer experience.