Over the past decade, robo-advice has emerged as a new way to provide financial advice to customers. It is a technology-driven approach that relies on algorithms to provide investment advice. While robo-advice was initially seen as a game-changer due to its low fees, ease of use and accessibility, the business results have been mixed. Typically, a one-size-fits-all questionnaire led to a short list of ready-made funds to choose from.
Various surveys have shown that the majority of users still reject robo-advice. The main reason for this rejection is the lack of human touch and empathy in the experience. In the UK, the FCA limits indeed the level of personalised advice a financial institution can provide without human involvement. However, with recent advances in AI, it is now possible to humanise the experience and make it more personalised, empathetic, and engaging.
Leveraging Natural Language Processing
One of the key advances is the use of natural language processing (NLP), which enables a better understanding of the context and the profile of the user. NLP can analyse the user’s input rate and understand their intent, tone, and preferences. It can then map it to the standard taxonomy used in wealth management and interpret different variations. This can then be used to engage in a more personalised and empathetic way, using language variations that the customer is most comfortable with, whilst staying within the regulatory boundaries.
One of the greatest benefits of implementing NLP for customer service is for back-end data analysis as it provides a way to contextualise and sort the full flow of customer data. One example of this is sentiment analysis whereby companies can align the emotional content of customer communications with the actions that they take. Companies can learn more about what their clients are talking about, where there are gaps in their offerings or resources and how they can address and fix these concerns.
Implementing conversational interfaces
In order to flex standard, static questionnaires, a conversational interface can further help; customers can ask the questions that they have in mind and Generative AI can provide dynamically generated responses on the fly, in plain language. A mix of conversational and generative AI can weave the required questions into a natural thread, alternating between questions that the customer might have and the questions that the financial institution must ask for wealth management purposes. The AI tool can also profile the user in the background on their Knowledge & Experience regarding investing, which in turn can be used for the taxonomy used by the NLP engine. Overall, this is more reflective of how a conversation would go face to face with an advisor.
Developing behavioural finance
Advances in behavioural finance have been made possible via AI, taking psychological factors into account. Analysing how people make financial decisions based on different risks and outcomes, an AI-powered profiling tool is able to more accurately assess someone’s true attitude to risk.
When it comes to presenting investment options, a conversational interface can again be much more “human” than presenting standard “KiiD”(Key Investor Information Documents in pdf format published by asset managers). For example, Generative AI can explain fund categories, asset allocation and investment strategies in plain words, drawing upon thousands of examples. Many customers struggle to evaluate the different choices offered to them. These are typically presented in the form of a table comparing a handful of funds, listing the attributes provided by the fund manager (per the KiiD) and some independent analyst ratings. This is where Generative AI can be of big help, letting users formulate questions in their own language, mapping these back to standard taxonomies used by asset managers and then providing answers back in plain language English.
Providing support and guidance
It is also essential to provide ongoing support and guidance to users, further bolstering the human dimension. This can be done through regular check-ins, updates on the investor’s portfolio, on-going investor profiling to ensure the service provider stays in tune with the evolving needs and preferences of the customers. AI can learn over time how customers want to be nudged or informed: do they prefer emails or pdf attachments, what time of the day are they most likely willing to read about their investments etc.?
In conclusion, robo-advice has the potential to revolutionise how guidance around wealth management is provided via digital interfaces, whilst still adhering to the strict rules set by the FCA. The focus is on taxonomies, plain language, education and conversations as opposed to clicking options. Advances in AI are making it possible to humanise this experience in a way that was not possible before. By using natural language processing, behavioural finance techniques, conversational engagement, and AI-generated responses and explanations (conversational techniques), robo-advisors can provide a personalised and empathetic experience.
