Man and machine: A glimpse into the future of financial advice

abstract character shattered into pieces

Technology is waiting in the wings that could overhaul the advice process and dramatically change the way firms talk to clients.

These developments are not decades away, nor mere years away. Some of these radical systems are already here, making their way to the UK from US and European markets.

They include systems that map clients’ emotions through movements in facial muscles, and technology that mines social media to predict life events such as retirement or buying a first home.

But are advisers ready for such a step change in the way they give advice? Do these systems offer a way to enhance the advice process and make it more robust, or simply serve to create a bigger compliance problem further down the line?

The issue of man versus machine is fast becoming something else – in advice, it seems more likely to be a case of man and machine. So are computers beginning to deliver advice in the way humans can? And more importantly, are they doing it better?

Machines that read emotions

One of the latest innovations comes from Swiss-based technology firm nViso, which in July launched its Emotion Advisor tool.

An alternative to the risk profiling questionnaire, clients are asked to rank their financial priorities before sitting in front of a webcam. They are then played a short financial video, with the system measuring how their face moves in response to the material. These facial movements are mapped to emotions, such as happiness, fear and anger.

Following the video a report is generated, which allows advisers and clients to compare stated financial priorities with their emotional financial priorities.

nViso co-founder and chief executive Tim Llewellynn says: “So much of how we communicate is non-verbal, through tone and gestures, not just what you say but how you say it. It is one of the dominant criteria in how people communicate, and yet computers are missing it. That led us to thinking there was an opportunity to develop systems to measure innate human behaviour.

“The system tracks a taxonomy of 170 facial muscle movements, then relates that back to how to decode that and translate that back to primary emotions. We have taught the computer using visual intelligence to track these movements automatically.”

Llewellynn says Emotion Advisor can help advisers in terms of the suitability of their recommendations by providing a record of clients’ emotional response when it comes to financial decision-making.

He adds: “This information can be used to give back to that person to help them become more self-aware, but it also helps the adviser better understand the nuances of how to communicate with clients. We are not selling tools and dashboards here, this is about partnering with firms.”

IBM is behind another tech development in the advice space through its Watson project. The system uses channels such as social media to build a picture of a client’s wider behaviour. IBM analyses this data then claims to be able to predict life events or advice “triggers” such as the birth of a child, retirement or the receipt of an inheritance.

Alongside this, the system uses the aggregated consumer data to give advisers lists of segmented clients, including who are the priority clients and which individuals are at risk of leaving the advice firm.

Finance & Technology Research Centre director Ian McKenna says: “It is designed to create an additional level of support information when an adviser is talking to a customer, showing screens of information on the subjects most relevant to that client at that time. In a way it sounds a bit 1984, but on the other hand we are talking about mapping information so that we can better support consumers.”

Emotion Advisor is being offered as either an online tool which acts as a standalone website, an internal tool for advisers to use with clients either on iPads or through the company network, or integrated with existing financial planning software. Costs are broken down into licence fees and bespoke charges though nViso declined to give figures. IBM declined to comment.

Entering the mainstream

The application of this kind of technology could be widespread, and wider take-up of tools such as Emotion Advisor could see standard processes such as risk-profiling turned on their head.

McKenna says these kinds of systems are a natural evolution of advice as a service, and can help improve advice recommendations.

He says: “Emotion Advisor is really a metaphor for all the ways technology will transform the way in which the advice process will work over the next five to 10 years.

“In the next few years this sort of thing is going to start to become mainstream. It is about enabling advisers and other financial firms to embrace these developments and to understand far more about their customers and what they really think. To what extent are the answers given as part of the advice process what the client thinks they should say, rather than the answers they really feel? The technology enables us to cut through any bias and get to the heart of what the customer is really thinking.”

McKenna argues by being able to aggregate consumer data as IBM is doing, patterns can emerge which can inform advisers’ conversations with clients.

He says: “This is how big data becomes real in financial services. It is not just about being able to predict life’s milestones, as we go forward we’ll be able to better predict changes in people’s health, which may trigger a mid-life career change.

“It’s about getting into the detail of these subtle changes that are going on in our lives all the time and being able to pick up on them, enabling us an industry to become more effective at targeting consumers with the services they’re looking for at a time that’s appropriate to them. That’s the big difference, instead of starting with a product set and seeing which products should be recommended.”

Llewellynn says Emotion Advisor may appeal to younger clients, as well as younger advisers.

He says: “There will be a lot of pushback among some camps, but we are already seeing positive endorsements from millenials, and the so-called ‘new wealth’ generation. There will be a huge amount of wealth being transferred through the generations, so how are advisers going to attract that new wealth and engage with those clients?”

Llewellynn says he has spoken to advisers who have initially been sceptical about the system.

But he adds: “One adviser thought about it some more, and made the point that where he could see this being useful is with younger advisers who don’t have the same level of experience as older advisers, and who may not necessarily have such a strong rapport and relationship with clients. There will be pockets where this fits in very nicely.

“The more conservative will need more data and more time on this. But this is robust science, this is not a gimmick.”

Leap of logic’

But not everyone is convinced these systems are offering genuine benefits to advisers.

Threesixty managing director Phil Young says: “The problem for me with neuropsychology and these kinds of technological developments is a lot of these things are still in their infancy. It involves a big leap of logic from measuring facial patterns to pin it down to the extent that this emotion has been triggered by that response.”

Young admits there may be a role for technology such as Emotion Advisor to be used ahead of the client conversation. He gives psychometric testing as an example of something that “at its purest” should be done without an adviser present, as he believes the adviser can “contaminate the experiment”.

But he adds: “There may be robust science to say the software works, but the evidence still does not prove the use of the software in the context of advice works better than simply asking clients straightforward questions. This whole area is unproven, and seems to be trying to find a solution to a problem that doesn’t exist.”

Yet the technology also has its backers, including risk-profiling and financial technology firm eValue.

Strategy director Bruce Moss says he welcomes anything that progresses the science of risk-profiling, saying the standard risk questionnaire is a “fairly blunt instrument” in understanding the client’s motivations beyond a simple risk category.

He says: “Something we are interested in and are working on is using artificial intelligence to mine how somebody has interacted with a website. You can tell a lot about how the personality of an individual depending on how quickly they answer the questions, do they go straight through without pausing, do they take a long time over each question, do they go back and change their answers.

“The other thing we’re interested in doing is a “people like you” feature, so from the information that we have about a person, their wealth, their age, their family situation, we can say for people with a particular risk profile we can say this is the kind of solution they have been looking at.”

Distribution Technology chief executive Ben Goss says the emerging technology is a good starting point to have the risk conversation with clients.

He says: “The world of technology continually moves on – the power of computers in our pockets and our smartphones doubles every 18 months. We should be open to how best to deploy technology to reducing the cost and risk of delivering high quality advice.

“But the role the adviser plays is really critical in empathising with the client and helping them understand their needs. Putting them in front of a camera and seeing how they respond? I don’t know. Advisers are well qualified, skilled people, and core to their role is understanding, consulting and empathising with the client, and doing that over a period of time. The psychometric analysis is a great starting point but it needs to be evolved with the insight that a human being can provide.”

Man and machine?

Emerging systems such as Emotion Advisor and IBM’s Watson project shows just how far technology has come, and pinpoints where advice could end up.

But the fit between man and machine inevitably leads to a debate about whether computers are starting to give advice better than actual advisers.

Moss says: “Machines can’t do what advisers can do, but they do it differently. When an adviser is picking up body language, there is no record, an adviser is forming an impression. Of course, when it comes to review, you’ve got the risk questionnaire but not all the intuitive stuff the adviser overlaid on it. When you have a computer doing it, you will absolutely have something that has been audited at every single step of the process.”

Yet the relatively early stage of development means Moss is also wary.

He says: “These systems have to be thoroughly tested before they are applied because you are industrialising the advice process. If you get it wrong, you are going to have problems because a lot of people will have received bad advice.

“Understandably, advisers may be pretty cautious because there is a regulatory risk associated with these things. Mapping people’s facial expressions, for example, may be proved to be an extremely reliable approach, but if I’m banking my business on that, that is quite a bold thing for me to do.”

But McKenna goes further. He says: “Frankly it’s simply not true anymore that machines can’t do what advisers can do. The science is proving that machines can measure these things more accurately.

“This does take you to that uncomfortable space where the machines are beginning to do things better than humans.”

Expert view

Simon Collins

There is an analogy to be drawn with these kind of technology developments and another big tech development – blockchain. There is a lot of talk about blockchain at the moment, and a lot that is trying to happen, but it is quite Wild West and pretty scary.

No matter what the technology and how it develops, it does not disguise the fact that people still have to be provided with suitable advice.

There is a real possibility these developments will assist advice firms in being able to better comply with financial services regulation. Technology takes away some of the grind of an adviser’s life, potentially eliminating mistakes and giving the adviser more time to think about the solution.

US regulator the Financial Industry Regulatory Authority has previously stated: “We reinforce that a registered representative that uses a digital advice tool to help develop a recommendation must comply with requirements of the suitability rule, and cannot rely on the tool as a substitute for the requisite knowledge about the securities or customer necessary to make a suitable recommendation.” I expect the FCA to come out with something similar in due course.

What the Finra statement says to me is advisers really need to understand whatever technology they are using – not necessarily the depths of the algorithms, but what the technology actually does and how it can look different.

With Project Innovate and the Financial Advice Market Review in the background, the development of technology is positive and should serve to assist advisers. But advisers cannot be absolved of responsibility.

As with most things in our industry, technology is not a total panacea and one needs to tread a little carefully and remember who is responsible. Technology and the way these tools are being developed is a case in point – it all comes back to due diligence.

Simon Collins is regulatory managing director at Eversheds Consulting

Adviser views

Philip Milton

Managing director

Philip J Milton & Company

“Tools that measure facial expressions are somewhat irrelevant. When a client comes to see me, they want me to use my skills and do what they are paying me to do. They do not need a machine for that. Clients are paying me to help make the decisions for them – they want to get on with their lives without getting into the minutiae.

Hall-Dennis-Yellowtail-2013 700 x 450.jpg

Dennis Hall

Managing director

Yellowtail Financial Planning

Predictive technology needs volume in order to work, and facial recognition can work in an online advice context. As advisers we tend to rely on our own ability to recognise what is going on with a client, as that is what they are paying us for. These tools may be worth a look but I can’t see what they add, particularly when so much of our relationships are built on human involvement.