(Wealth Briefing) -- This article explores how AI can help firms implement granular segmentation strategies streets ahead of the rather blunt categorisations now so often prevalent.
Thanks to the efforts of Google, Netflix, Amazon and the like, consumers have been trained to expect slick user experiences and the utmost levels of personalisation in all their interactions with service providers.
As these companies have proven, fully leveraging client data is the key to market dominance. In an increasingly digitalised – and competitive – industry, wealth managers will be seeking to follow their lead and ensure client segmentation delivers the tailored service and operational efficiencies it promises when done well.
An abundance of data may have long been available to help wealth managers forge a deep understanding of their client base, but relatively few have been using it in a nuanced and systematic way – or indeed at all. Even in the hotly-contested UK market, a quarter of institutions have not been formally segmenting their clients. Elsewhere, segmentation seems to have generally been based on only a handful of variables.
Industry luminaries have long advocated more granular segmentation techniques and their potential to improve operational efficiency, client-centricity and marketing success can hardly be disputed. Rather than a lack of enthusiasm, it seems likelier that technological challenges around data gathering and analysis have held firms back.
Solving segmentation challenges
In the view of Alessandro Tonchia, Co-Founder of Finantix, segmentation has always been “a necessary evil for economies of scale”, but never a particularly valuable exercise due to the sheer diversity of individuals’ wants and needs, and the difficulty - hitherto - of forming a meaningful picture of them.
“The cluster of characteristics that make you, you as a wealth management client are incredibly personalised and broad,” he said. “That’s why segmentation didn’t really work, because there were so many attributes to consider and not enough aggregation capability.”
Now, however, wealth managers are seeing that they can push these barriers aside through AI. Powerful groups of Machine Learning (ML) algorithms are able to take a non-linear approach to both structured and non-structured client data, performing fast, complex calculations on huge datasets and synthesising the information into actionable insights – all with as much automation as desired.
“With AI, you can track what the client reads, what they invest in, how they travel, what their interests are, how complex their family is and so on,” Tonchia explained. “Then, you can adjust everything - from how their portfolio is constructed, to the news that you send them, how frequently you meet and the products you recommend in future - to that unique set of attributes. We are truly moving towards ‘a segment of one’.”
Rather than being reductive and static, AI allows for a sophisticated, dynamic approach to segmentation that allows firms to act upon the factors most relevant for each specific type of interaction or element of service delivery.
What is more, they can then easily adapt their approach as those classifications change over time. A client’s wants and needs rarely remain fixed for long, so AI is not just a game-changer because of the sheer volume and variety of data-points that can be taken account of. Very impressive feats of “data-crunching” can be carried out automatically, and then the predictions made via ML will also improve over time as new data emerges.
Seizing the AI opportunity
At its best, sophisticated segmentation will simultaneously enhance the client experience and a firm’s efficiency in delivering it. But this calls for a clever combination of closely observing individuals’ behaviour and lessons that can be drawn from broad groups, along with a level of responsiveness that has only recently become achievable through AI.
The opportunity has already been seized, however. Several big-name banks are already using AI to dynamically segment clients based on factors such as subtle changes in their behaviour, whilst simultaneously predicting financial events in their lives.
“Real-world use cases of AI/ML techniques assisting wealth managers in extracting value from client data and management information are now beginning to emerge,” said Phil Tattersall, Director in EY’s UK Wealth & Asset Management Data and Analytics advisory practice. “They aim to better inform the role of the relationship manager with powerful analytics, with the goal of improving client engagement via better decision-making by the advisor and an improved capability to answer difficult questions in real-time.”
AI is improving the suitability and timeliness of the advice given to individuals, as well as driving the delivery of tailored content and alerts. But these tools are also being used by firms to look across their entire client bases to predict attrition and identify product and service opportunities. Wealth managers are increasingly keen to improve their use of client data and management information as a way to drive growth.
This last point is incredibly important when we consider how much larger and diverse the addressable market is becoming. Wealth demographics are going through seismic shifts and firms are having to rapidly get up to speed with serving countries and client segments that may be outside their comfort zones (millennials, female entrepreneurs and non-traditional families being but a few).
Wealth managers are going to need real focus to differentiate their offerings to all the segments they might seek to target while still maintaining efficiency, and Tonchia believes AI can play an invaluable role here.
“AI will give the wealth managers wings to implement their very specific strategies,” he said. “So, if a firm is concentrating on family succession and estate planning, they’ll have the tools to really execute on that through targeted client education and events; or, where they are really focused on investment returns they will have powerful tools for news automation and market alerts.”
According to one study by Cambridge University, intelligent algorithms can determine a consumer’s personality better than their friends, just by analysing the posts they have “liked” on Facebook. However, as elsewhere, there is a large element of fintech co-dependence in AI’s potential to make granular segmentation techniques a profitable and efficient reality. With challenges like legacy technology and data siloes still looming large, many firms will require significant upgrades to be able to gather, store, analyse and manage the huge amounts of data required to really “move the dial” (this is one area where new entrants may be at very great advantage).
But it should also not be forgotten that much depends on buy-in from front-line personnel. Despite the onward march of digitalisation, advisors remain the main conduit for wealth management relationships and institutions’ most potent means of deeply understanding their clients. The days when relationship managers held the bulk of client knowledge “in their heads” may be long gone. Yet asking them to input every single useful scrap of information they glean into “the system” may still be a stretch. The issue of who “owns” clients – institution or advisor – is not unproblematic.
For our experts, the answer is twofold. Firstly, the process of constantly improving the institution’s understanding of the client must be easy. Here AI can also help, such as by Natural Language Processing (NLP) technology sifting pertinent details from meeting transcripts rather than advisors being asked to laboriously type everything up. Secondly, advisors must be clear that “giving up” the knowledge of clients they have earned really is in their interest. As with all technology upgrades, success rests on securing staff support.
“There may be some psychological resistance, but if bankers can see that the technology will help them increase their revenues and be more efficient and professional – which it will – they can hardly complain,” said Tonchia.
Putting the case even more strongly, David Teten, Managing Partner of HOF Capital, warns that unwilling relationship managers will rapidly become anachronisms. “Imagine a sales person who doesn’t use email – they just couldn’t get a job,” he said. “But two decades ago there would be sales people saying ‘Oh I don’t use email; I don’t need it’; and it’s the same dynamic with these new technologies.”
In the race to improve client acquisition and retention, AI technologies will be able to make a huge positive impact. But it is also clear that both institutions and advisors will have to be willing to seriously rethink the way things have previously been done, particularly in client segmentation. In the past, wealth managers have struggled to capture the “essence” of complex private clients from a handful of bare metrics, but now they no longer have to - if all stakeholders commit to playing their part.