AI in Wealth: Digitally-Infused Human-Centric Business at the Front Office

Imperative for Change

The massive market growth in FY’20-21, resulted in financial celebrations for leading wealth managers, but also threw in a keen realization that the AUM growth (and therefore the fee income associated with it) was possibly a result of primarily market growth, and therefore masks underlying mis-performance. Key operational health metrics actually did not improve - Net new customer additions stagnated, advisor churn accelerated by 2-4X relative to previous years, Fund flows per new customer is declining, and revenue per advisor trailed that of assets growth.

Leaders in Wealth are questioning the resilience and sustainability of the current business model, and business risks in FY 22-23 as the market fluctuates. In parallel, advisor churn is affecting client retention and growth, in part driven by changing business fundamentals (e.g., Wells Fargo decided to shut down international wealth business) and in part due to new exploding digital possibilities (opening up fintech and advisors exploring more independent ventures). A TD Ameritrade survey in 2020 found that 40% of advisors reported “more likely” now than 6 months ago to break away and go independent.   

Fractal conducted exhaustive interviews with financial advisors of leading wealth managers in the US, to understand pain points and how AI could help.

Three main themes emerged:

Help to Grow high-value acquisitions (80% of FAs expressed)

1. Acquisition tends to be a manual, volumes exercise, with a 2% conversion. A typical FA has to work their way through 10,000+ leads, which may result in 200 clients. The process of working through is often using cold emails, phone calls or using social media like linkedin – without prior intelligence on who to prioritize.
2. Major need for intelligent lead generation with size of investible assets, pipeline scoring and prioritization.

Provide better advice and support to manage existing customer relationships (60%)

1. Avg 200 client base, $150-250M in AUM per FA, across guided and managed services. Basic tooling around client portfolio and profile is typically in place.
2. Advanced insights including portfolio metrics like beta, diversification scores, correlations, goal optimization and simulated impact from new buy/ sell actions.
3. Personalized next best product and service recommendations.
4. Easy to access, insightful information on hand-held devices.
5. Identify high off-us value and potential high wallet customers.

Provide operational conveniences (40%)

1. Wishlist for (a) “automated FA assistant” to handle most typical queries (b) semi-automated onboarding and document data extraction/ form fills.
2. Easy to search, catalogue, summarization of home office recommendations and research.
3. Remove low-value operational queries, increase self-service by customers.
4. Automated, cognitive workflows for onboarding, KYC, documentation steps.

Key roadblocks to driving Digital transformation

In a recent Fractal hosted Digital Transformation roundtable with 8 of the US’ leading banking, payments and wealth organizations – several roadblocks were pointed out as common themes:

• Challenge #1 (By far, the biggest issue) - How much to invest in Digital: While CEOs are conceptually excited about Digital, the decision to migrate funding from well-functioning legacy channels (i.e., hire more advisors etc.) is far from easy. Several questions remain unanswered to the C-Suite: What is the role of Digital in an FA-led business, How much can we attribute fund flows and growth to Digital as a channel?

• Challenge #2 – What metrics should the C-Suite care about: All the chief digital officers at legacy wealth orgs agree that digital will not be a standalone P&L unit, and therefore the specific metrics to measure digital success is not consistent. Traditional digital engagement metrics (visits, clicks etc.) are a-plenty, however – unclear what they stand for, and how they contribute to the bottom-line.

• Challenge #3 – Not clear alignment on hype vs. reality (“Even if digital needs to be invested into, just get a couple of tools. AI is mostly hype in the front-office”). This is common to most human-intermediated businesses, and any survey would reveal that “it’s a people business, clients would not respond to a robot”. The leaders however, have approached differently by defining this as a Digital-infused human business. Therefore, the problem is not to replace the FA, but to provide them the tools/ intelligence/ alerts needed to operate much better.

How can Wealth businesses catch up to the rising table-stakes

Despite being an AI company, Fractal believes that AI is only going to be partially successful here, given the deeply disjointed backend data, systems, challenges of adoption, and therefore hesitation by Digital/ AI groups to be accountable to a business outcome.

Therefore, unlike in other industries we do not believe a full-scale transformation is the key to impact here. Successful front-runners have taken a highly surgical approach to build out AI, ensuring optimal and accretive choices to drive transformation. We used the A-D-A-P-T approach to define meaningful agile sprints towards transformation, which became hugely successful in a few clients:

Action: What actions are likely to see highest impact (using FA Input). Describe the current and target state with defined benefits of change. Illustrate, and gain early adoption with the biggest “customers” for the new process

Data: Hate to bring this up, but usually the best laid plans get disrupted because of a late realization that data readiness is a couple of years away. Therefore Data is fundamentally, right at the top of the framework

Analytics: The part where AI comes in - less about the algorithms itself, but more about the ability to generate the right level of insight/ action, at the right latency, to the right consumer. These ofcourse, are defined by Action, constrained by Data, and enabled by …

Process: AI without a redesigned process will not work. As alternate actions get defined, are there different measures, new interaction models to be designed? How can existing processes be simplified, and how can use design and behavioral sciences to simplify, educate and drive adoption

Technology: The underlying backbone, which would include the consumption tools, automated pipelines, scaling up frameworks, CRM systems, testing and orchestration tools

ADAPT approach, helped to uniquely develop bite-sized outcomes, delivered collaboratively in pod-structures and, unexpectedly, helped to break down several silos in the organization.