Print this article
Personalised Relationships Without Overburdening Personnel
Wendy Spires
23 January 2018
Wendy Spires, Head of Research at WealthBriefing, explains key ways AI could lighten the relationship management workload, enabling deep personalisation for clients but without overburdening personnel. Right products; right people; right point
The industry’s digitalisation continues apace, but wealth management in the true sense – with all the complexities that connotes – is still dependent on trusted relationships and high-touch, personal service. Indeed, this service ethos could be the sector’s main defence against the influx of low-cost new entrants into the space.
Yet not even client-facing advisors are immune from the efficiency imperative engendered by the industry’s unprecedented profitability squeeze. Many have been left struggling to maintain the highest service standards across burgeoning books of business: WealthBriefing research suggests that a quarter of UK RMs serve 51-100 clients each and 15 per cent even more, while in Asia’s premier banking segment ratios of 400:1 are common.
Wealth management will remain a “people business”, but the fintech revolution has led to a growing acceptance that personnel need not be the sole means of delivering highly-personalised service.
Companies like Amazon, Netflix and Google have “trained” consumers to expect service providers to “know” them intimately .
Forward-thinking wealth managers will now be keen to emulate how these masters of data have harnessed AI technology to deliver continually refined - and yet highly cost-effective - customisation. In fact, according to Phil Tattersall, Director in EY’s UK Wealth & Asset Management Data and Analytics advisory practice: “One of the primary AI/Machine Learning use cases for wealth managers is the ability to provide personalisation at massive scale very efficiently through algorithms”.
Moving away from the mundane
A key theme running throughout this report is how AI helps advisors move up the value chain in their use of time. Here wealth managers have a lot to learn from retail banks in their automation of the more mundane interactions relationships entail. Card cancellations and balance enquiries are just the start of the commoner enquiries that might be automated, freeing up advisors’ time for conversation which add greater value.
Importantly, technological solutions are becoming increasingly sophisticated and therefore applicable in a wealth management context. As EY’s Tattersall notes, “chatbot solutions utilising machine learning can answer the more simplistic types of questions from clients very effectively, while also learning from these queries and improving over time”. Still more intelligent are the “cognitive agents” already being used by several large banking groups to handle quite complex customer service interactions.
Customising content
To promote themselves as trusted advisors lightening investors’ mental loads, wealth managers need to ensure they are cutting through, rather than adding to, the investment “noise” HNWIs are bombarded with. Therefore, one area particularly ripe for AI amelioration is in tailoring the news and research clients receive to ensure that they are relevant, timely and delivered via their preferred channel.
According to Alessandro Tonchia, Co-Founder of Finantix, “with AI, firms can quite easily move from ‘information overload’ to content being precisely mapped to clients’ portfolios and interests with distribution happening instantaneously when market-moving events occur”.
Simultaneously, usage patterns like email open rates and website dwell times can be examined via machine learning to ensure investors are receiving only what will be of most interest, when and how they prefer – all of which would be a dramatic improvement on what many firms currently have in place.
Maximising the value of meetings
Yet the “reading” capabilities of Natural Language Processing mean personalisation and time-saving for both client and advisor can be pushed very much further. Client-advisor meetings are by necessity a fairly infrequent occurrence , and these stand to become very much more valuable and efficient for both sides.
Tonchia explained: “Imagine an advisor is meeting with a client invested in a particular fintech company and they have hundreds of 50-page research PDFs available to discuss, of which only a fraction cover the sector. Very quickly, AI-enabled technology can ‘read’ every sentence of every PDF, not only to single out the ones dealing with fintech but to extract content mentioning that specific company or their competitors. It can then assemble the most relevant sentences into a ‘real-time’ contextual report.”
At an even higher level of AI sophistication, an NLP meeting transcription might pick up that the client is concerned about China, leading to this being added to the next meeting’s agenda, additional information being gathered and perhaps a discussion with a specialist being scheduled – all automatically. The technology could even gather client feedback on agendas and literature ahead of meetings in an end-to-end process fostering continual improvement and added value.
And streamlining meeting preparations should be a real priority given their onerousness and high client-loading figures. Close to half of advisors say accessing three or more systems and significant manual work is required ahead of annual reviews, with a fifth taking over two hours to prepare for each one.
According to Tonchia, AI could lift much of this burden, as well as eradicating a great deal of the other more routine relationship management work advisors do, such as making appointments, collating emails and compiling reports.
As David Teten, Managing Partner of HOF Capital, argues: “A good rule of thumb is this: if a junior administrative professional can help an advisor do something via email today, then that function is likely low-hanging fruit for AI to pick.”
There are clearly innumerable opportunities for AI to automate the day-to-day work of delivering a personalised service that might otherwise fall to advisors to do manually and at great time cost. But the technology’s potential to deliver true and dynamic customisation should be seen in far broader terms.
Cost pressures and compliance obligations mean that wealth managers must be more focused than ever on proposing the right products and services, to the right people, at exactly the right point in time. So too does heightened competition: making appropriate recommendations is key part of personalisation, and making clients feel deeply known and valued by the institution. Again, this is something digital consumers have come to expect and technology commentators have even flagged the phenomenon of users today actually feeling quite offended if Netflix or similar suggest products misaligned to their preferences.
The broad point is that these algorithms are usually “scarily right”. For wealth managers, it is that AI can furnish an elegant and far more impactful alternative to the rather crude way that marketing by segment has tended to play out hitherto.
Wealth managers’ sales efforts have previously relied on “weak abstractions of broad client categorisations” like age or nationality, Tonchia explained, but with AI they can assign actions based on one or multiple specific attributes. What’s more, previous experiences with either individuals or sets of clients can inform where efforts would best be concentrated. “You are continually refining your picture of each client’s preferences, but you can also use AI analytics predictively,” he said. “If investors with a shared profile have approved of a proposed investment, you can assume others will too; whereas if it has never worked with a certain type of client, you know not to waste your time.”
Automated engagement systems
Not only can personalisation through AI boost client engagement, it could also help them to become better investors, by serving up education at precisely the right point in time or nudging their behaviour in the right direction.
As Greg Davies, Head of Behavioural Science at Oxford Risk, explained: “Firstly, wealth managers could look to build automated engagement systems where computers observe things like what clients are doing with their portfolios, and when and where they log on, before carrying out ‘brute force’ AI data pattern recognition to identify ideal opportunities to engage. This could be delivering little nuggets of just-in-time education. If a client’s profile suggested they had great understanding in some areas, but were fuzzier elsewhere, then the firm’s communication efforts could be tailored to bring them up to speed in the weaker ones.
“Secondly, you could implement an automated system of ‘nudges’ to help investors trim their own behaviour. They could sign up to be reminded when they are monitoring their portfolio too much and you wouldn’t need the advisor to be doing that, just a system running underneath.
“This is where AI meets behavioural design identification: you can start to design engagement systems that interact with clients in very different ways depending on their particular profile, and then gradually bring them along a path in a highly tailored, engaging way.”
AI technologies clearly have much potential to optimise sales through far more personalised marketing, but where they may have an even greater impact on the bottom line is at a deeper, psychological level.
Enhancing the client experience is a rather nebulous concept, but engaging with clients more productively and making them feel better about themselves as investors are two key ways to make it more concrete. As our expert panellists have highlighted, AI is at the leading edge of the industry’s quest to achieve ultimate personalisation at scale, and truly optimise the experience of both clients and time-pressed advisors.