Unlock Your Access for the Data Webinar: Winning Data Teams: Getting Started and Scaling Your Team aired November 15, 2022

Artificial Intelligence products and platforms deployed at the enterprise level is still a relatively new experience. Even though wealth management firms are slowly embracing machine learning (ML), natural language processing (NLP) and other models to dig deeper into their customer data and make business decisions, we are only scratching the surface of the benefits of AI.

Dozens of vendors have launched AI products for the wealth management space and, of course, they all claim that their software is the best for whatever problem a client is looking to address.  But how can firms cut through the noise and get an unbiased opinion?  

That was the goal of a recent discussion I moderated at the In|Vest NY 2018 conference.  Two consultants and one  industry analyst comprised the panel and provided a vendor agnostic point of view that was as refreshing as it was informative. (See Digital Digest from the InVest 2018 Conference)

Artificial Intelligence for Wealth Management  

We’ve all seen or heard about artificial intelligence and we’ve experienced a very broad adoption of it across financial services. Nearly 1/3 of financial firm executives said their business had adopted some early artificial intelligence (AI) technologies, such as voice recognition, predictive analytics, or recommendation engines. We’ve also seen articles about how AI is going to destroy millions of jobs, but a small percentage of firms said increasing worker productivity was one of the reasons deploying the technology.  In this post, we will review some of the leading vendors in the AI space and discuss their strengths and weaknesses.

It is helpful to think about AI in terms of the intelligent automation (IA) of non-value added activities that free up capacity and improve productivity, explained Gavin Spitzner, president of Wealth Consulting Partners. Another form of AI is augmented intelligence, which helps advisors and clients to find opportunities through big data by identifying advice opportunities and scaling hyper-personalized advice and communications, leading to better client experiences and outcomes, he noted.

Numerous studies have shown that despite advances in fintech over the last decade, the average advisor still spends approximately 2/3 of their day on what Spitzner referred to as “administrivia”. This includes behind the scenes stuff like compliance and other tasks that do not directly help drive top-line growth. Advisors are spending only 1/3 of their time on sales and service and actual advising, the bit that should be their value-added, he pointed out.

The reality is even worse, because the percentage of clients who feel as though they are receiving good, personalized advice is much lower than it should be, Spitzner reported. Ultimately, AI in wealth management is about the combination of automating the dull stuff and scaling personalized advice, reducing cost, and spending more time advising, he stated.

There are three core attributes that we have seen as indicators of success in AI enterprise deployments, Spitzner observed.

AI can undoubtedly help asset managers improve performance, lower cost, etc., but what is most interesting is AI that helps wealth managers better help clients with their financial journeys over time, Spitzner declared. So the way we organize this, it’s loosely around ecosystems on the left and more points solutions and some of the newer entrants on the right side.  (See 6 Ways AI is Helping Build Consumers’ Confidence in Banking)

If there’s an 800-pound gorilla in AI for wealth management at this point, it’s hard to argue that it’s not Salesforce, Spitzner asserted. They’re the closest thing to the notion of a single pane of glass that everybody is striving for. That’s where the output of AI is going to be displayed, in predictive analytics and next best action recommendations.

Salesforce has developed an impressive set of enterprise applications including Einstein for AI, that when combined with their Lightning app development framework can serve up real-time, personalized activity recommendations for advisors, Spitzner noted. They also have a strong cloud computing platform, a unified view of the client, extensive API connections, and flexible workflows, which is an area that doesn’t get the attention it deserves, he added.

The challenge for Salesforce in wealth is that they are often brought in at the consumer and commercial bank level, which only represents 10% to 12% of the firm’s revenue, Spitzner explained. Relegated to a small corner of the bank, it tends to be neglected when it comes to internal training for configuration and optimization. Hence, the booming market for Salesforce consultants such as Scion, Appirio, and AppCrown.

Also, Salesforce pushes out three or four releases each year, which generates a lot of information to be absorbed. Firms need to have dedicated people in-house who can keep up in order to take advantage of the never-ending stream of new capabilities. Data scientists and practice management consultants to understand it, adapt, and educate. being able to educate advisors on where they have risks in their client base.

Other firms identified by Spitzner as AI leaders: IBM and BlackRock.

While Salesforce has been in the AI space for a while, who do you see as an emerging AI vendor?

NextJ Systems is a lesser-known CRM vendor that has built some strong AI functionality into their platform, Spitzner observed.  They showed some impressive AI components in their demo earlier in the conference that they built using a set of IBM’s Watson tools called Client Insight for Wealth Management.  (See 10 Disruptive Demos from InVest 2018)

Bryan Sachdeva, VP of Wealth Management for NexJ Systems, published a blog post that listed some actionable insights that AI can generate by leveraging the data available at a typical advisory firm:artificial intelligence wealth management

  1. Book of Business Optimization: recommending which customers to move to self-service or robo channels and which to spend more time on to maximize AUM.
  2. Customer Profitability Optimization: identify suitable products to recommend or activities to perform for clients that will maximize likelihood of growing AUM or Commissions.
  3. Service and Engagement Model Optimization: suggested service models or engagement models based on client interaction history, sentiment, and previous touch deferrals/tardiness.
  4. Intelligently Recommend Next Actions: for Opportunities, Leads, Complaints, and Service Requests, based on previous cases that closed quickly/successfully.

A surprising choice by Spitzner for an emerging AI vendor was Microsoft, since they have not developed as much wealth management domain expertise as Salesforce. He felt that Microsoft was a company people should be keeping an eye on because they have all the necessary components to develop enterprise solutions including Dynamics CRM, Cortana, Azure, Office, SharePoint, etc. So for firms that are Microsoft shops, it’s hard to argue that they do not present a compelling AI offering, he advised.

At Microsoft’s recent financial services summit, there were a lot of partners providing AI solutions around fraud detection, Spitzner related. One statistic he mentioned was that standard AML compliance sees almost 95% false positives.  This generates tremendous operational overhead and may also result in users missing real problems that are buried under a mountain of alerts.

We have seen proof of concepts from some of the AI providers on this list (see slide) that claim around 85% reduction in false positives, which saves 93% of the time spent evaluating them, Sptizner stressed. Microsoft is also deeply integrated with Adobe Marketing Cloud, so there are some solid use cases around marketing campaigns, workflow productivity and other things.

Data is the Key

One important aspect of a successful artificial intelligence implementation is the underlying data to support it, explained Doug Fritz, President of consulting firm F2 Strategy. Many wealth management firms spent the last 20-30 years just building plumbing between different tools but never took the time to consolidate their data.  This makes it quite difficult for them to take the next step, he observed.  (See Will AI be an Advisor’s Best Friend or Worst Nightmare?)

artificial intelligence wealth management

Fritz has seen many large wealth management firms that get started on AI machine learning projects, but then realize their data is not ready to support some of these solutions. Those companies that have spent time consolidated their data into a single, organized data lake or operational datastore are much better positioned to leverage AI tools like the ones Gavin mentioned, run analytics and gather insights from them, Fritz noted. 

We’re seeing data projects and data organization architecture projects come out of the AI space because it’s a step you have to take to be fast and nimble with these solutions, Fritz stated. These larger vendors like to claim that their solutions can do “anything” while the smaller vendors are very specific in what they do, he pointed out.

If your organization is “wanting AI,” as Fritz put it, but you don’t know what you want, then you’re not going to find a clear direction with one of the large firms since they’re going to overwhelm you with options. Fritz challenged the wealth management folks in the room to spend some time figuring out the best first steps they should take and then investigate some of the smaller AI firms and decide whether what they have meets your immediate needs.

Fritz recommended some good ideas for initial projects such as improving advisor-client communications and analyzing client interactions to pinpoint attrition problems. Insite Engines is one of the smaller AI firms that have developed strong solutions for these type of projects.

Agreeing with Doug about the importance of data was Will Trout, Head of Wealth Management Research at Celent.  Data is entering organizations through multiple inputs and the Internet of Things is an even bigger inflection point and reverberates throughout the organization. However, old style data warehouses are increasingly insufficient to manage the flow of information going in and out of the organization. So to Fritz’s point, getting your house in order is the first step towards leveraging the power of data for AI for machine learning.

Challengers and Emerging Vendors

Whether or not you are working with one of the big firms as your ecosystem anchor or one the challenger firms Spitzner listed (see slide above), you should have access to solutions around natural language processing (NLP), natural language generation (NLG) and conversational AI.

The following is a list of vendors identified as challengers, but many of them already have some of the top 10 or 20 banks as clients, so they are inching their way into the top tier of vendors. They deliver some interesting use cases around advisor productivity, converting data to text or text to data, and chat bots. 

Two of these four challenger vendors are focused on chatbots, which have developed beyond simple Q-and-A. These tools now enable banks to provide retail banking customers their credit score, enable them to set and manage their budgets, and notify them about specific transactions, all from the convenience of the bank’s mobile application or website. (See 9 Questions on Artificial Intelligence for Wealth Management)

Spitzner included Pefin, a direct-to-consumer robo-advisor, in the list of What’s Next/Emerging Vendors. Pefin’s use of neural networks is a differentiator for them and can improve their personalized advice.  They could have similar growth rate to Betterment or Wealthfront in their early years. 

Other emerging AI vendors include: