Weaving AI and ML into customer experience
Paul Mullins, former Managing Director, HSBC
Below is an insight into what can be expected from Paul’s participation at CeFPro’s Customer Experience Summit.
The views and opinions expressed in this article are those of the thought leader as an individual, and are not attributed to CeFPro or any particular organization.
How can institutions integrate artificial intelligence and machine learning into their strategy to enhance customer experience?
In virtually all industries, this thinking is already underway, and it’s an exciting space. It all begins with data, ensuring you have the right tools in place to access both structured and unstructured data to unlock opportunities.
Most people immediately think of chatbots or better-guided advice, but there are numerous ways AI, ML, and Generative AI can make an impact. This includes providing better access to information for agents to deliver quicker and more accurate responses. It also enhances fraud detection, which, while not always apparent to the customer, significantly improves the overall experience by reducing false positives and detecting threats before they even become apparent to the customer. In the area of credit, there is a great example from 2020, where Barclays partnered with Amazon in Germany, leveraging AI analysis of consumers’ online behaviors to offer real-time credit to shoppers at checkout.
With any new technology or innovation, it’s best to start small and seize low-hanging fruit. There are a number of consultancy and tech companies that can assist with the discovery phase if that experience is not in-house. It doesn’t have to be prohibitively expensive either to get the ball rolling.
It’s equally important to remember to bring along and engage the employees on the journey as much as it is the customer for any deployment to be a success. Without their buy-in, it’s likely to be a much slower transition – for the most part, AI will help employees focus more on the things that count, and that is the customer.
Are there potential risks and reputation damage associated with the deployment of AI and ML? If so, why?
AI and ML are well-established practices in the financial services industry, mainly used to streamline processes and reduce errors. However, their effectiveness is only as good as the user input, so any model developed needs appropriate governance in place and, importantly, traceability so it’s clear where content emanates from.
Generative AI carries even higher stakes, with a risk of bias and misinformation (so-called hallucinations). Despite the hype, the application of this technology is still in its infancy, and whilst it can happily provide you recipe suggestions from the contents of your fridge, it may not be ready to be 100% relied upon for complex tasks like retirement planning…yet.
There are also ethical considerations and trust issues. How people feel about interacting with a computer and the extent to which it can be trusted are important factors. While we accept that humans can make mistakes, we have different expectations for computers.
Additionally, there’s the reputational risk of not embracing AI and ML, as customers, shareholders, and employees may question why the company is not adapting to one of the potentially biggest changes in how we work and live since the Industrial Revolution.
What impact will ChatGPT have on AI product launches? What does this mean for financial institutions?
Already, ChatGPT can analyze customer data and provide personalized financial advice. It can suggest investment strategies, savings plans, and other financial recommendations tailored to individual customer goals and risk profiles. It can tailor insurance policies in the same way, providing personalized underwriting across any number of types of policies, potentially lowering premiums as a result of the enhanced risk management capabilities.
There are numerous use cases; it can also help predict future trends to inform product design and features better, assist with the creation, and also the quality control of documentation.
There are ethical and moral considerations to be mindful of. However, there will need to be appropriate governance and mitigation, especially in areas where regulation struggles to keep pace with innovation. It will eventually catch up, and non-compliance could prove costly. As a result, financial institutions may face constraints on deployment in a way that other industries do not.
The near-term solutions are likely to be more focused on freeing up employees to spend more time with the customer. A great example is Morgan Stanley, who, in March, introduced an internal-facing service that leverages OpenAI technology and their own intellectual capital to deliver relevant content and insights into the hands of Financial Advisors in seconds as opposed to the weeks it may have taken beforehand. This helps to drive efficiency, scale, and improve the overall customer and employee experience
In addition, companies that operate globally will likely face different rules and regulations across different regions and countries, and this will mean how they make use of the tools available may well be different market by market, dependent on privacy and regulatory laws.
One thing is very clear, though: work on preparing a business for the cultural change that Generative AI brings, from the board down, will be a prerequisite for any product or service changes to be successful and sustainable – work that all businesses should be doing now.
How can institutions leverage KYC to tailor products and improve the customer experience?
KYC provides insights into a customer’s background and financial situation. This information can be aggregated to customize risk decisions based on individual circumstances.
The same is true in Insurance; as an example, Sara Assicurazioni and Automobile Club Italia are already encouraging drivers to install ADAS systems in exchange for a 20% discount on their insurance premiums. Indeed, it has been demonstrated that such systems can slash the rate of liability claims for personal injury by 4-25% and by 7-22% for property damage.
It can also help inform product design and allow for a far more granular understanding of individual personas’ behaviors, supporting tailor-made propositions for customers based upon specific customer needs rather than the average of a million different customer needs today.
AI can also assist the process by automatically detecting and identifying customer documents and checking for fraud.
However, it is important to remember that, in most countries, data captured can only be used for its intended purpose, emphasizing the need for adequate controls to govern AI use in the financial services industry.