By FintechOS · September 06, 2021
9 minute read

Advice gap: automated personalized guidance is key to solving the issue


In our previous article on this topic, we considered how well-intentioned market regulations have exacerbated a financial “advice gap”. To solve this, insurance, following in the steps of the investment industry, needs to exploit technology to create scalable, automated customer guidance

says David Punter, Product Owner, Digital Insurance at FintechOS.

In my previous article in this series, I explained how the UK Financial Conduct Authority’s (FCA) 2016 Financial Advice Market Review made 28 recommendations to financial services firms. Many of these were intended to narrow the ‘advice gap’ between affluent consumers – who employ dedicated financial advisors – and lower-income individuals – who go it alone.

Financial providers only wanted to sell their products to informed consumers, but were finding themselves responsible for the financial decisions of their customers. To remedy the situation, firms had to employ financial advisors to support their customers in making good choices within their means.

Ultimately, another solution was found: so-called ‘robo-advisory’. Incumbent firms would make a one-time investment in automated systems that would collect and analyze a customer’s data and provide guidance on financial product choices according to pre-defined criteria customer’s financial position. The customer could then make the final decision on the purchase, removing the need for the financial services provider to employ human advisors, and also absolving them of liability where their customers’ financial decisions did not turn out well.

This would also be good for consumers, many of whom didn’t want to face lengthy consultations as part of every financial decision. Yet, there would still be some safeguard in place to help them make sound choices.

This all sounds like the ideal solution. So, what went wrong?

The two kinds of robo-advisors

Most insurers and financial institutions are built on legacy systems that can’t easily be upgraded. They aren’t Google or Apple. Creating intuitive, artificially intelligent (AI) platforms on top of outdated systems is a gargantuan task, particularly when there’s no proven return on investment.

Without the kind of sophistication that AI tools bring to automation, the first-generation robo-advisers created in response to the FCA’s request were primitive, at best. Generally, the systems fell into two, equally flawed categories:

1 User manual bots

Some robo-advisory systems focused on providing generic information about the product and leave it to the customer to read and process the guidance, then determine whether the product was for them.

This method relies entirely on the customer’s own expertise. Yet, customers can’t be expected to understand terms and concepts like ‘moratorium underwriting’. This could lead to the customer being refused treatment for a medical condition they had believed they were covered for.

This approach also relies on the customer actually reading the advice that’s provided to them. Unsurprisingly, most people will likely skip the fine print.

2 Choose your own adventure robo-advice

On the other hand, some robo-advisory systems simply took an elaborate paper form that a financial adviser would normally go through with a customer and place it online. The customer fills in the form through a web portal and the resulting information can be used with a decision tree to lead them to specific advice based on their own needs.

This ‘choose your own adventure’ concept can struggle to find the right balance between simplicity and over-simplification.

In contrast to lengthy paper forms and human consultation meetings, a robo-advisory questionnaire system leads the customer to a much-shorter and more-specific version of the full-length guidance. The customer will, therefore, be more likely to read the information as it’s briefer and more relevant. But it’s easy to see that if the customer skips information or doesn’t understand it, then the quality of their decisions could be far too unreliable.

However, if the robo-advice tries to mitigate this, it would require spending much longer – perhaps hours – filling out detailed forms: hardly any better than the situation the firm tried to fix.

Efforts to implement logic and validation can help, but the fact remains, the more automated a solution is, the more likely it is that inaccurate input leads to inaccurate output, in other words the resulting guidance might be correct from a software point of view but wrong for the human customer. Lacking the necessary flexibility and sophistication, the robo-advice system might not be capable of suggesting an alternative product that would be more suitable, or even rejecting customers who are supposed to be eligible.

This could leave the financial services firm at risk of penalty from the FCA for providing inaccurate recommendations, not to mention reputation risks among customers who made poor financial choices.

Robo-advisory is not a failure

This is not, however, a complete disaster for the concept of robo-advisory. Any software developer will tell you that a minimum viable product (MVP) with a series of flaws to overcome is the starting point of any innovation. The issue, however, is how to move forward with it.

First-generation robo-advisory may not have solved the advice gap immediately, but the benefits of great robo-advisory are still there for the taking to anyone who can keep iterating and solve the many small details required for the overall customer experience to be success. It may be a challenge, but solving the advice gap with automation is an achievable goal.

In fact, challenger insurers and neobanks are already focused on this goal, partly because their business models tend to be built around automation and scalability in the first place, and partly thanks to having no legacy that holds them back from implementing a bold new vision. As new consumer robo-type solutions often target low-hanging fruit, full-featured and larger financial institutions may not take their solutions seriously, but proof is in the growth curve. In every industry, digitally-native challengers are winning over ever bigger numbers of consumers and establishing themselves as a more relatable and customer-focused option than the more experienced players. On the one hand this is competitive disruption at work, and on the other hand this is exactly how the advice gap is supposed to be closed: via innovation.

Robo-advisory and audit

In addition to the above, robo-advisory offers a distinct advantage for regulated firms: automated advice is much easier to audit.

Private conversations between customers and human advisors come down to simple he-said/she-said. An angry customer can claim more was said than actually was, while advisors acting inappropriately can claim innocence.

For robo-advisory, when the advice is good, there’s a clear record of what was communicated and why that information was provided to the customer, proving the advice was credible. On the other hand, when the advice is bad, it’s easy to identify where it came from and why, then correct it.

To compete with challenger insurers already offering this service, established financial institutions need to offer comparable automated services that succeed in appealing to individual customers. The challenge is in overcoming legacy systems so a successful robo-advisor ‘knows’ who customers are and what’s best for them. There are some fundamental steps that can be taken to achieve this:

1 Brownfield automation

There’s no getting around legacy technology. Replacing often fragmented core systems in any established financial institution is too expensive and disruptive to make it a short-term option. The only practical way forward, since speed of innovation is essential, is to integrate brand-new technology into your legacy systems. Prioritizing modernization of the front-end means your customers can enjoy a state-of-the-art experience while outdated tech under the surface is not perceptible as a limiting factor.

2 Data Lakes

AI and RPA will be essential to robo-advisory achieving its ambitions, but these tools are useless without the raw data to feed them. This is why every major financial institution is currently working to collate all their legacy data and store it in ‘Data Lakes’.

Often the raw, unformatted data is brought into a Data Lake and then organized more formally in a ‘Data Warehouse’. Essentially, these are just cloud servers storing information to serve as fuel for sophisticated AI processing. These data stores are becoming essential for financial institutions to remain competitive as technology progresses.

3 Product-based guidance

Tailoring advice to each customer is a complex task. Perhaps easier is to provide guidance tailored to each product. Products will only appeal to certain consumers, so you can easily rule out a large portion of your customer base from each service.

This doesn’t contradict the importance of customer personalization, but it does offer a more pragmatic approach than trying to take the full range of scenarios that a human financial advisor handles, and then somehow digitalize it in a single guidance solution.

4 Hybrid process

Just as I recommend blending your legacy technology with the state-of-the art, I also suggest you look at integrating your existing and future processes. The right tech can make this seamless.

For example, should your customer become stuck on a difficult step in your onboarding survey, they can contact your call center and have an advisor remotely access their device and complete the step for them. Having a ‘safety net’ for customers means being more agile about implementing digital journeys which are new to the business, and then ironing out common issues in a more automated and self-service way later based on feedback from customers and advisors.

5 Do it yourself

Hiring a team to do a complete digital transformation can seem like a daunting prospect. Thankfully, new generation platforms such as FintechOS allow you to focus on digital product building without reinventing the wheel.

Adopting a no-code / low-code platform means you can build your new workflows and automated guidance rules – and keep tweaking and improving them – inside systems you control, without having to reinvent common components and infrastructure.

Let’s close the advice gap

A recent KPMG study showed that 85% of insurance CEOs say COVID-19 has accelerated the digitalization of their operations. Now is the perfect time for robo-advisory to move to the next level and close the advice gap once and for all.

To discuss your robo-advisory strategy and how to accelerate implementation of digital advice journeys, visit the FintechOS site for information on both insurance and banking solutions, and to book a demo.

Share this article with your connections