Unlike firms that have ramped up their spending on technology in the race to have the latest AI-driven advances, Ameriprise has kept its tech budget largely steady.
Even before the public release of Open AI's ChatGPT large language model in late 2022, Ameriprise was putting lots of money into projects like automating routine parts of client meetings. Of course, AI now takes a large share of the $1 billion the Minneapolis-based wealth manager devotes to technology every year.
But the goal remains the same, says Gerard Smyth, Ameriprise head of technology and service delivery.
"I think from the firm's perspective, we're always grounded in return on investment," he said. "You know, is there a return to the firm? Is there a return to advice? Is the return through clients? I would say we're grounded on those principles, and it's not about topping the headlines."
It's not just large wealth managers like Ameriprise that are seeing returns from AI spending. A Cerulli Associates survey of 68 RIA firms between May and July found that 64% of the respondents think AI has reduced their manual and administrative burdens, and 46% think it has improved their client communications.
READ MORE:
Those are exactly the sorts of results Smyth has set his mind to at Ameriprise. A native of England, Smyth came to Ameriprise after helping to build trading systems for the Australian Stock Exchange and later moving to Minneapolis to work for American Express Financial Advisors, which was spun off from its credit card parent company as Ameriprise in 2005.
"I came here for a one-year assignment," Smyth said. "And 25 years later, I'm still here."
Smyth stepped into his current position in 2020 and is now responsible for technology used by the firm's advisors, other employees and clients, as well as organizing services and operations. He recently sat down with Financial Planning to discuss how AI is helping advisors save time in client meetings, how AI can relieve pressure from advisor retirements and his firm's future plans for technology.
The following interview has been lightly edited for length and clarity.
Financial Planning: Of course, we hear a lot about AI improving various routine tasks that advisors have to do in their back offices. Do you have some concrete examples of how you guys are using AI to make advisors' lives a little easier?
Gerard Smyth: Sure. I think it's very important to understand at Ameriprise that we've been on this journey of delivering technology to support advisors' work processes for years and years and years. And certainly, AI pre-generative models, machine-learning models were a major part of quite a few of the workflows we've built out.
To give you some concrete examples,
Certainly a great example is we were able to add an AI capability of meeting summarization into that e-meeting flow. And because it's just a natural part of it, the adoption has been tremendous.
FP: Have you heard already from advisors who are able to spend more time with clients in face-to-face meetings or doing tasks that they're better suited for.
GS: Absolutely. We estimate that meeting automation, on its own, has freed up 10 to 20 hours per week, per advisor. And then certainly adding summarization and the client letter post that meeting is saving another five to 10 hours a week.
So, you know, very large savings that can then be used to create capacity for the advisor to engage in advice and for the client experience.
FP: When you have those summaries of the meetings, is the advisor going back over it to make sure there aren't mistakes and that it's indeed summarizing it in an accurate way?
GS: Oh, absolutely. I mean, certainly, a key part of the rollout of AI in our firm was within the enterprise risk management program. Right at the heart of that, responsible AI was a key tenet and a principle we live by. So a human in the loop and confirmation in that workflow that the advisor has reviewed the output.
READ MORE:
FP: Have you found that AI has improved? Was it more liable to make mistakes early and then, maybe as time goes on, it becomes more familiar with the subject matter and less prone to make mistakes?
GS: You know, I think it's that combination of the learning and also the improvement in the models. The models have just improved so dramatically over the last few years. So, yes, it gets more accurate as the technology gets better, and as the learning gets better.
FP: Are there other examples of how you're using it, either with back office tasks or just regular things that advisors used to do as part of their day-to-day work?
GS: Certainly. We have AI-powered insights, which is also a tool that we've had for many years. Insights is really about providing advisors opportunities to engage with their clients and to take action on behalf of their clients. It's about growth opportunities and client insights, and we've added that across the book of business, across the practice.
Again, our theme is integration, and insights are provided at the front end of preparing a meeting. These AI-generated insights can be used to customize a meeting to meet the client's needs.
FP: What have you had to do to add AI to your existing technology and systems?
GS: We've been on the journey of taking the data from all the disparate tools that advisors use and really bringing that together into a data lake that enables us to bring all those key pieces of data together for a client household, an advisor book of business or a practice.
So we're not going tool by tool and trying to slip it together. We actually have a really good data layer that we're leveraging for our AI solutions.
FP: What's next in your AI plans?
GS: We plan to roll out an advisor assistant. That's going to be a little later in the year. That's going to start marrying not just the client's information and the advisor information, but also our internal research and external research, as it makes sense. But again, very much an advisor in the loop, very responsible, and, of course, hitting that data layer, that right context, so you get meaningful value from the solution, and not just random data.
FP: That leads to the inevitable question: Do you see this as taking away anybody's job?
GS: We're certainly not seeing it yet. I think our goal is to take away the [business-as-usual] work that we can automate, so advisors and their staff can spend more time on meeting their clients' needs, being more responsive to their needs, better engagement, but also it gives them the opportunity with that capacity to grow their practice.
FP: One thing everybody talks about in this industry is
GS: I definitely think the pressure is there, and obviously recruiting more advisors into the industry is a huge focus area for us as a firm. But I agree with you completely. The automation through AI and through good investment solutions and advice automation all create capacity to make up for that shortfall of advisors.
FP: Some firms,
GS: It's certainly not part of our strategy. The AI capabilities we're enabling are really as a way to advance the advisor-client relationship across all segments. You know, we certainly haven't started to think about segmenting into different types of solutions. It's more about all of our segments.
