Hyper-personalised financial advice: Creating better consumer outcomes with AI

19 Jul 2024

First seen on Professional Adviser

Having worked as a financial adviser myself for more than a decade earlier in my career, it seems to me that good financial advice was always about hyper-personalised advice – the more personalised the better. Sound financial planning has always been built on trust and trust has relationships at its foundation.

Fact-finding was never just about facts, but facts flavoured with feelings, aspirations, worries, stresses, behaviours, and family values. Financial objectives risk being dry without colouring them with what makes for happiness, and contentedness, as well as an understanding of personal goals, bucket lists, idiosyncrasies, mad ideas, safety and changeability.

The quality of that fact find data and how well it is documented is vital, not only for the advice firm's head of compliance, but as the basis of highly customised hyper personised recommendations based on financial goals, and as many ‘soft facts' that the client is prepared to entrust.

Can AI help to reduce the advice gap?

The Retail Distribution Review made that sort of personalised, face-to-face-based advice a much rarer thing. The number of regulated financial advisers shrank dramatically. Before we knew where we were, a yawning advice gap had opened up. So that according to the Financial Conduct Authority (FCA), only 8% of UK adults have access to regulated financial advice today.

The big question of the moment is:
Can generative artificial intelligence (AI), along with automated processing and analytics technologies, be used to collect and interpret personal customer data to generate tailored, and eventually hyper-personalised, semi- or fully automated advice, thereby reducing the advice gap?

It's a question that many major banks and advisory groups have been wrestling with for some years. For example, in 2018 in the US bank Morgan Stanley rolled out a new capability to its 16,000 US-based financial advisers to aid them in their work with clients.

Called Next Best Action (NBA), the system is both a platform for personalised communication and engagement with clients, and an AI-based recommendation engine for investment and wealth management ideas that advisers can present to their clients.

The NBA system, if used correctly, has the potential to significantly change the ways Morgan Stanley's advisers work with their clients and undertake their roles. Usage of the system is voluntary, but some advisers are seeing positive impacts from extensive usage.

In the 2020 Covid-19 pandemic, the system became even more important for Morgan Stanley's Wealth Management business when face-to-face meetings between advisers and clients were impossible. In the first two months of the pandemic, the NBA system was used more than 11 million times!

For UK advisory firms exploring how to transition to hyper-personalisation, there are two key elements to consider:

• Access to high-quality, up-to-date personal data is key.
• Recognising that enabling hyper-personalisation is a journey: you cannot flip a switch and expect your systems to be producing gold-plated highly personalised advice recommendations for all your clients tomorrow.

So, if the quest to offer hyper-personalised advice to a larger number of customers than you can serve today is a journey, what are the steps along that path? The second half of this article is a personal view on what hyper-personalisation is and what it isn't.

What hyper-personalisation isn't

It isn't communicating with clients by using non-personalised generic data, like providing automated nudges to all customers based on ‘generic' non-personalised calendar events. One obvious event of this type is using the approaching tax year end as the trigger for a marketing campaign encouraging customers to optimise their annual ISA allowance. You see these sorts of generic nudges being put out by many direct-to-consumer platforms today.

It isn't utilising basic ‘low level' personal data such as your clients' age at their next birthday. So, you might send a letter or send a personal message to all clients' as they approach their 55th birthday, wishing them an early Happy Birthday and using that event to prompt a pension review session or retirement income planning conversation.

This type of nudging is relatively easy to automate. Many of us will have received letters (mostly mandated) from providers as we approach our 55th birthdays, reminding us of our pension options. However, by their nature these nudges are generic. They do not inspire much trust because it is evident that they are not speaking to the individual, or reflecting on their personal situation, specific goals, circumstances and concerns.

It isn't even automating outbound nudges based on more granular personal data. Children's ages often coincide with big shifts in personal and financial goals. Automating marketing messages around end of school, start of university, end of university can all prompt productive client engagement as priority adjustments and affordability shifts occur.

How might AI and automation create some hyper-personalisation?

Let's take the case of a client in retirement who has a regular monthly income set up from a self-inested personal pension (SIPP). Generative AI may well have already been used to elicit and capture hard and soft facts while setting the right level of income and selecting an appropriate investment strategy.

AI could have been used to keep the client up to date with the performance of their plan and aid the financial planner to prepare and present review information, hopefully reducing the overhead and improving the quality and personalisation of the information produced. But let's take this a stage further:

What if AI was used to monitor the performance of investments far more frequently than a human could? What if AI could monitor the cash in the client's current account at regular intervals? After all, we have got used to the idea of micro-savings providers automatically increasing a direct debit payment if it looks like a client can afford to save more.

What about an AI intervention that reduces a monthly income if the client doesn't need as much that month? Just as with micro-savings, the client can always say no to the nudge.

We all know sequence risk is one of the biggest factors on income longevity in retirement. What if AI worked with a client to agree suitable rules for adjusting monthly income based on investment performance?

These very personal, high-frequency adjustments could only be affordable with automated non-human actors, but the customer outcome benefits are potentially huge.

What role might digital assistants have in increasing engagement and capturing key changes in customer behaviour or changes in circumstance? Not wanting to make this sound too creepy, but it probably won't be too long before client activity on a portal is monitored by AI in real time.

The benefits? How about flagging activity normally associated with a concerned or worried client to their adviser? How about the digital assistant asking the client directly some questions designed to probe changes in personal circumstances or objectives that allow proactive contact and adjustment of recommendations?

One thing is for sure, the world of financial planning will be very different ten years from now. I will be, hopefully, comfortably retired. However, one other thing I hope for is better customer outcomes, especially for customers who cannot currently afford great hyper-personalised advice.

Previous Article Dunstan Thomas Products Contact Us