18 June 2018

How to get robo-advice to work? Try being human.

Thomas Ridley Siegert

By Thomas Ridley Siegert LinkedIn

A number of financial services providers are looking to robots to deliver advice to consumers. But are they at risk of repeating the same old mistakes?

There’s not a week goes by without another financial services provider announcing they are doing something ‘innovative’ with robo-advice. From LV= investing millions at the end of last year, to Moneyfarm getting £40m of funding to grow their robo-advice division – it seems like everyone is jumping on the bandwagon. Last year, the Big Four banks all announced they were looking at offering robo-advice.

The reasons why are pretty clear. Robo-advice can be a cost-effective way of delivering better outcomes for millions of customers. And from a customer perspective, it’s much cheaper than face to face advice – meaning many more people are likely to use it. Other sectors, such as retail, have integrated AI seamlessly for years. See Ravn ACE or KLM for interesting case studies.

However, there’s a huge difference – the consequences of bad advice are much greater in Financial Services than for most other sectors. This is why the FCA has been keeping a close eye on it.

Same old song

Last month the FCA criticised the online robo-advice market when they found serious failings in several app-based investment companies. They analysed advice given to consumers on a one-off basis when the customer first interacted with the firm. Some of their key findings were:

  • Fees and charges not being clearly disclosed
  • The customer’s knowledge, experience with investing and understanding not being properly evaluated
  • Information taken whilst receiving robo-advice was outdated or inadequate
  • Vulnerable customers not being identified and supported appropriately

The day the earth didn’t stand still

These findings aren’t exactly shocking. They apply to almost all areas of financial services. For example, 17% of car insurers don’t disclose their fees at all in their policy documents, according to our latest Customer Experience Ratings. Similarly, 97% of credit card providers had at least one issue with website accessibility, such as colour contrast errors or a lack of keyboard-only functionality.

I’m sure that robo-advice will fall victim to other common issues too; heavy use of jargon, high reading grade of language used, too many words per sentence, and so on. We see these issues time and time again when assessing documents as part of our Customer Experience Ratings.

Basics first, everything else second

Warren Buffett said never to invest in anything you don’t understand. If the advice a robot is giving to customers is incomprehensible, full of jargon, and aimed at people with PhD reading abilities then consumers won’t be able to take advantage of it.

Regardless of channel, product or brand, not giving customers the information they need in a way they can understand is one of the largest problems facing financial services today. Robo-advice is a great tool, but AI is only as good as what you put into it. If the bot is littered with the same problems plaguing financial services currently then what’s the point?

AI may be the future, but companies need to make sure they address the fundamental problems first. Invest in communicating clearly with your customers, and ensuring you are experts in explaining the benefits and pitfalls of your products. Only once you’ve nailed that challenge, should you be investing in AI technologies.