Consumer Understanding: Using AI Tools to support Consumer Duty

08 Nov 2023

Dunstan Thomas, pensions specialist fintech, has been focusing hard on enabling clear customer communications for policyholders through the statutory money purchase illustrations (SMPIs) and its simpler annual benefit statements (SABSs) for many years. It has been awarded the Crystal Mark by the Plain English Campaign for the layout and readability of these documents which it delivers for many providers.

Improving Financial Understanding

However, it remains clear that while readability is one thing, understanding is quite another. The UK scores poorly for financial literacy relative to other countries across Europe. According to both the FT and the Financial Conduct Authority which campaign on this issue, half of the UK population has low confidence in making decisions to do with money. Numerous studies link poverty with low levels of financial understanding.

Groups with the least financial literacy include those with lower incomes, young people starting their careers, women and ethnic minorities. It is these, often more financially challenged, groups which are most exposed if they make poor financial decisions. Therefore, if we are going to improve customer outcomes for all pension policyholders, we need to try to improve financial literacy, and at the very least check the understanding of those who are receiving their pensions illustrations.

The advent of a new wave of generative AI tools since ChatGPT launched last November is already producing some interesting tools which are worthy of investigation. For example, one we've been evaluating for our SMPIs and SABSs delivery services for clients was created by a tech firm called Amplifi.

Amplifi originally developed its AI-driven tool to assess legal contracts' intelligibility for typical consumers. Its core tool could also be adapted to determine what level of financial literacy is needed to fully comprehend the information being imparted in a typical pensions illustration. If too much of your illustration shows that your policyholders need an advanced understanding of pensions and at retirement options, to comprehend key information laid out there; then it might be time to simplify still further and retest understanding in the interests of improving consumer outcomes, thereby reducing calls into your helpdesk following annual statement distributions.

Personalised Virtual Assistants in Financial Services

Another area that is seeing increased adoption of AI is in the development of AI-powered digital human avatars. These tools enable text to be brought to life by using AI to provide human-like walk-throughs of complex documents including pensions and investment illustrations, even offering a light quiz at the end to check understanding.

These new AI tools offer the potential for personalised virtual assistants, capable of explaining what you are looking at and what specific phrases like Balanced Portfolio actually mean in language that can be understood by a typical end investor. It offers the option of providing the details behind the raw data with as much explanation as the customer needs or values. Tools which combine visual elements such as graphs and charts with the raw numbers and explanation from a human-like representative has been found to increase engagement and comprehensiveness markedly.

It should also be possible for customers to self-configure their projections to highlight what's most important to them - essentially moving the most interesting piece of information for them - to the top of the page view or turning some data into more meaningful charts at the touch of a button.

The customer could answer a chatbot question: ‘Do you like what you see?' with ‘No'. Which leads to a chatbot response: ‘What would you like to change?' The customer may then enter: ‘I want to retire earlier than 67.' That pushes them through to a new calculator session where they put in several new parameters and draw others from existing records.

There might be a prompt to review the risk level you have versus the current portfolio. That might trigger a Financial Health Check with your adviser. Or if D2C, then you go through a risk profile journey and can opt to rebalance your investments in line with the revised risk profile or change of circumstances.

Using AI to increase consumer understanding

What if you could take this level of multimedia interactivity a step further through AI to give the same virtual assistant the wherewithal to answer typical supplementary questions which customers might ask having been through their annual statement?

As long as the tool can interpret the question accurately and tap into its dynamic and continuously increasing database of answers to customer questions (linked to a specific type of statutory document), the opportunities for increasing consumer engagement and consumer understanding can only improve iteratively, all without the concern about hiring more new contact centre agents in time for statement delivery peaks. It also creates the potential for customers to tap in and check their pension status whenever they choose, rather than waiting for a notional annual review.

Increasingly being integrated into illustration delivery services is a comprehensiveness test at the end of reading a document. Similar to the sorts of test questions on a CPD module, policyholders might be invited to answer a short questionnaire to check levels of understanding on the key points of knowledge at the end of the reading of their annual statement.

It might be possible to go further still and enquire whether they have any comments or queries on the information in the document they have just read. We are becoming increasingly used to being asked for online feedback and reviews. Why should your pension documentation be any different? The stream of customer comments coming back could provide a feast of ideas for further improving customer experience and the readability of these documents in the future.

Delivering better consumer outcomes

AI-driven tools offer the promise of a new golden age in terms of the use of technology to deliver deeper pensions engagement and ultimately better consumer outcomes. You can use the resulting digital interactions to gather more data about the customer and ascertain whether their profile fits with the typical profile of customers benefiting most from this type of product or portfolio. Where some misalignment is evident from crunching the data, it should be possible to pass this feedback back to a support agent or financial adviser.

In summary, new generative AI tools offer the potential to address many of the challenges presented by Consumer Duty, especially its consumer understanding outcome which requires firms to deliver communications that meet customers' needs; are likely to be understood by them; and support them to make effective, timely and informed decisions; in addition to being fair, clear and not misleading.

Delivering this prescriptive outcome all needs to be evidenced, while ideally containing customer support costs. No mean feat! But new generative AI-based tools offer the potential to help in many of these areas. Specifically, tools now emerging can enable:

• A tighter fit between a product's core benefits and customer's core retirement income needs
• More opportunities to segway into financial advice around conundrums that are uncovered
• More tailored and personalisable communication, all within the parameters of statutory regulatory required communications
• Potential to iteratively improve the quality of your documentation
• Potential to improve your customers' financial literacy - engaging and educating them so that they are better prepared for retirement and less likely to experience financial shortfalls in retirement.

Previous Article Dunstan Thomas Products Contact Us