AI is on!
Implementation of first uses cases around real-time contextual added-value banking notifications and offers will make their appearance in 2018, predicts Sophie Guibaud, vice-president for European expansion at Fidor Bank.
Artificial intelligence (AI) is beginning to make some real inroads into the consumer banking experience. While some might have dismissed it as a gimmick a year ago, there’s no doubt that AI is going to used more and more by banks into 2018. What’s more, it won’t just be used in the customer-facing services at the front end, but at the back end too, helping banks to make better use of the data they have.
Much of what we see in the headlines about AI’s use in banking tends to focus on the services customers are directly interacting with. Many banks, for example, are using AI-powered chatbots to provide round-the-clock support for customers. These chatbots are able to answer a range of basic queries by instantly accessing relevant information from knowledge banks, CRM databases and so on. While in most cases, they still require humans to fill in the gaps and answer more complex queries, the theory is that the more questions they answer, the more they will learn and the better they will perform.
While few banks are prepared to talk about them, we are seeing a growing number of these chatbots. While RBS was one of the pioneers back in 2016 with its Luvo chatbot, Bank of Scotland has begun a trial of an iPhone-based system, while Standard Chartered plans to introduce one early next year. In Brazil, Banco Original is already thinking about several areas of its business where AI can be used effectively.
The advantages of using AI-based systems in customer service roles are clear. With chatbots on the front line for incoming customer queries, fewer human customer service staff are required. Those that do remain can use AI as a performance-boosting tool, helping to provide quicker and more consistent customer services. Banks can then, either reduce headcount, or reallocate staff to roles requiring more advanced skills, those that cannot yet be replicated by machines.
But as well as cutting costs by streamlining customer services, there are also opportunities to use AI to boost revenues within banks while delivering a very personalised concierge-like experience to theirs clients. With the European retail banking market shaping up for the implementation of PSD2, the age of true Open Banking is just beginning. It is an era of increased choice and innovation where customers have more power than ever before, able as they are to easily switch between banks and choose the services and products that suit them best. The banks that manage to construct the most flexible and personalised offerings for customers will be the ones that win through.
While these banks will need to have a strong strategy based on their use and integration of open APIs, they will also have to find a way of presenting the manifold and varied tools and services to customers in a way that is not overwhelming. Third-party partnerships with financial service providers that offer something useful to customers are all very well, but if they cannot be linked up to the right customer at the right time then the opportunity may be wasted.
This is where AI can step in – by applying AI and machine learning techniques to customer data, banks can create detailed profiles of each individual. From this, they can offer the products and services that are relevant to that customer, providing the kind of personalised service that they will find useful. The bank can earn commission from the third-party provider, while the third party gets a qualified new business lead. In this scenario, everyone wins.
To be able to achieve this kind of model, banks need to be in a position to not only gather customer data that they had in the past, but also add up new data coming from other financial services providers thanks to PSD2 enforcement, and analyse it effectively. While many banks have vast quantities of data at their disposal, much of it exists in siloes, with data about one customer’s online banking activity held in a different place to data pertaining to their telephone banking activity, for example.
The first challenge for banks is to bring all this data together in order to create a unified view of each customer, using all of the data they have. An additional step would then be to bring in data from third-party services where it is available – social networks, for example – to fill in any gaps in the profile. This then gives them an effective starting point for applying AI techniques, with an end goal of being able to offer relevant, tailored solutions to each customer in real time. For example, if a customer makes a big purchase of a car, then that is the perfect time to offer them a suitable insurance policy. This kind of automated, intelligent action will help to boost customer loyalty by providing them real time contextual added-value notifications and offers.
A word of caution, though. One of the main limits is not to appear too intrusive by managing customers expectations, asking for their consent and introducing the use cases that matter to them gradually. Research from UBS suggests that some markets – including the UK – show higher levels of resistance to AI assisted processes such as chatbots and virtual assistants among customers. This could explain why many banks are reluctant to talk much about implementing AI within their operations.
However, in my view, customers can be won over by AI services as long as they are well executed. Chatbots need to be genuinely functional and helpful, and available at all times. Targeted offers automatically prompted by a customer’s activity should be timely and relevant, but above all they mustn’t be intrusive. There is a fine line between being helpful to a customer and appearing like you’re monitoring their every move.
A recent report from GFT shows that nearly a quarter of retail banks in the UK see AI as strategically important. What’s more, IT systems used by banks in the UK are in a better state of readiness for implementing AI than in other regions. However, those banks that are able to implement AI solutions in their back ends to help create truly personalised services will have a competitive edge in the open banking world, no matter which market they are operating in. Expect to see more and more examples of this kind of banking service in 2018.