The predictions are high. UBS research amongst 86 banks says AI could boost banks' revenues by 3.4% and cut costs by 3.9% over the next three years. According to Tabb, of 200 global tier one and tier two banks more than 83% have evaluated AI and machine learning and 67% have actively deployed them (FinExtra).
Currently, front office is king in AI applications, according to the findings of a recent survey of the 30 of the world’s biggest banks by the Financial Times. Some 17 of the 18 banks that provided detailed answers are already using AI in front office, ranging from Citi’s Facebook messenger chatbot to UBS’s use of Amazon’s virtual assistant Alexa for customer service. Front office is also where banks see the biggest potential for AI-related savings. But the survey results also show that broad is best: eight banks are using AI in front office, middle office, back office and data analytics. The other 10 are using it in three of the four areas.
The growing competitive threat from big techs and digital-only banks should be a wake-up call for all incumbent institutions that they need to dip their toe into AI. There is also another reason for incumbents to focus more on AI: it can unlock a potential value of about $200 billion in banking sales and marketing alone, according to McKinsey. This is a big prize. Financial services firms have already shown considerable success in applying AI-driven tools to boost user experience and customer engagement. But sales activities seem to be lagging behind.
Banks are sitting on a wealth of customer data, which is a great advantage compared to some of today’s digital disruptors. As the financial services sector is turning to insight-driven sales strategies, machine learning (ML), a subset of AI, is a big help to improve effectiveness in sales. Predictive and prescriptive analytics could help identify purchasing patterns or forecast customer habits and behavior.
In an example of using AI in analytics, a bank has commissioned an analytics solutions firm to end the migration of high-value mortgage customers to rivals. The company compared the attributes of loyal customers with those that had churned, and identified over 100 factors related to customer, product and transactional data. The data was fed into a predictive modelling tool that uses neural networking techniques to predict churn behavior. The number of factors was reduced to around ten and the model was applied to all mortgage customers, ranking them in order of their likelihood to leave. Based on this ranking, the bank launched a targeted marketing campaign, which cut the churn percentage by nearly half.
There are also low-hanging fruits to reach for in AI development – think low-complexity AI use cases that offer significant benefits with a relatively small implementation effort. These are often missed, as institutions jump straight to the most challenging use cases of AI. Only about 20% of companies are implementing these ‘must-do’ AI use cases with a nice payoff, Capgemini found in a survey.
Banks can also rely on third-party solutions to tap sales growth potential offered by ML technologies. Third-party insight-driven digital sales and engagement tools use AI and predictive analytics to show, for example, if a customer is planning to buy a house or need a consumer credit. These tools collect, aggregate and analyze non-traditional data and other information, like geolocation, behavior patterns, social media interactions or even the weather.
Banks seeking to boost digital sales should consider building AI into their strategies. There are many examples in the financial services industry and in other sectors showing the immense benefits and endless opportunities AI solutions offer.