Opening Remarks & Fireside Chat: Michael Kitces, Head of Planning Strategy, Focus Partners Wealth

Transcription:
Transcripts are generated using a combination of speech recognition software and human transcribers, and may contain errors. Please check the corresponding audio for the authoritative record.

Brian Wallheimer (00:08):
Well, good morning everybody. We decided to get all the snafus out of the way now, so everything is going to go perfectly from here on out. I guarantee it. Welcome. Thank you all for being here today. A couple of quick housekeeping notes. As you can tell, this is a packed room. There are some seats—if you want to find one, please do fill in. They're scattered around, especially in the front. No one likes to be in the front, right? But there is overflow seating in the Flurry room. There is a video and audio setup in there. So if you want to find a seat or if you're ever in a situation where you can't comfortably sit down, find your way down to the Flurry room, and that works as well. During several of our sessions, you will see on some of the screens some notes being taken.

(00:50):
This is an AI conference, so we decided to have a little fun with AI and we're going to be using a program, I believe, called Rosie. Throughout the conference, it will be taking notes and making high-level bullet points about the things that are happening in sessions. It is accessible through your app if you've downloaded it. If you haven't, there are QR codes around. Please download the app. You can access that throughout the conference and at the end there will be a synopsis of the entire conference there. And it will be perfect, I'm sure. AI doesn't mess up. My daughter told me the other day, she was all upset because AI told her some celebrity was dead. And I was like, "There we go." But speaking of, I told my daughters that I was coming to this conference.

(01:36):
I've got three children and I told them I was coming to an AI conference. My daughter got very serious and she goes, "Are you for AI or against AI?" That's how she is about everything. And I told her, "Look, this is a conference about learning about AI. It's about education. It's about understanding the tools that are at our disposal." And so I did what I always do to her—and luckily I didn't make this about baseball—but I said, "Maybe you have a really great garage full of tools and you have the coolest cordless drill there ever was. It's got the most torque, it's cordless, it's got a long battery life, everything's amazing. That doesn't mean anything if you're trying to drive nails into a board, right? And if you add to that that you haven't gotten your building permits, you have a huge headache."

(02:21):
And that's what we're going to talk about in so many places here. What are the right tools for you? What are the right tools for the jobs? How do they work? Are you compliant? All of those things matter when we're talking about AI. So I told her all of that, and she just gave me a look and then a lecture about how much water AI videos use. So also, please use AI responsibly. It matters. It turns out I learned a lot from her yesterday. That said, I also want to talk real quickly about financial planning. If I can get our QR code up here: Financial Planning has done quite a bit of research on AI usage with financial advisors—how you're using it, what you're using, all sorts of things. If you scan that QR code, we made our overall story about the AI readiness survey free for you today.

(03:08):
Go in there and take a look. Tons of charts and graphs, a lot of data about what you all are doing with AI in the wealth management sphere today. I want to just make a few quick thank yous before we get on with the show. This was made possible in large part due to a wonderful advisory committee. There are plenty of you out there; if you go to the Advise AI agenda and look at the advisory committee, it is just a wonderful who's who of big names in the AI, tech, and wealth management space. I want to thank them so much for all of their time and expertise. To all of our speakers: you're all taking time out of your busy schedules. I know you're going to learn something, but you're sharing as well, and we can't thank you enough for that. Our Horizon team, who's put on such a wonderful show, and the tech crew in the back—we want to thank them as well.

(03:51):
I want to give a special shout-out to Suzanne Siracuse, founder and CEO of Siracuse Consulting. She is one of the driving forces behind the program today. If you like what you see, it is in large part due to Suzanne and her expertise. Last but not least, thank you all for being here. You're packed in here, and it just goes to show that there's a lot to learn. I am not going to keep talking. I'm going to hand this over to Suzanne, who's going to sit down for a fireside chat with a man who needs very little introduction in these spaces: Chief Financial Planning Nerd at kitces.com, Michael Kitces.

Suzanne Siracuse (04:44):
Good morning, everyone. And thank you, Brian. You guys have done a terrific job of carving out a niche for a conference completely devoted to AI and its uses in wealth management. I was here last year and it was crowded then; I think there are almost double the amount of people here this year, and it just continues to evolve. One of my favorite things I do is moderating, and the thing I love the most is sitting down with Michael because he's never short of an opinion—and on AI, there are lots of opinions. So we're going to get right into this. Michael, you've tracked the evolution of advisor technology for years through your research, writing, and consulting. How do you describe where we are today when it comes to adoption of AI across wealth management?

Michael Kitces (05:44):
From the adoption end, I would still say as an industry, we are in the "collective dabbler" stage. If you look at broad-based industry studies that ask if you have ever logged into ChatGPT and tried it out, 90-something percent of people say yes. We see headlines saying "90% of financial advisors have adopted AI," but "adopted" might be a generous word. It's nothing negative toward AI; I think our industry is just doing the normal adoption curve. There's a subset of early adopters who pick up the toys, dive in, and figure out what to do with them. Our industry has its share of early adopters—maybe a slightly smaller percentage than other industries because it's financial services—but they are out there.

(06:45):
But in terms of proactive use, it is still mostly the early adopters who are really in. The mainstream advisor is dabbling with AI for very focused use cases. They might use it as a brainstorming buddy for marketing ideas for 30 minutes, or they're buying an off-the-shelf solution for a particular use case, of which meeting notes is the most common. But it's not a "sidekick buddy" for the mainstream advisor yet. Part of that is trust issues; they're still getting comfortable. Some of it just comes down to not knowing what to ask. They're stumped about the use case.

(07:49):
It reminds me of the folks who say Excel is amazing—you can do anything, write macros, program in Visual Basic. Then the other 99% of people are like, "What?" The power users do amazing things, while the mainstream user occasionally opens a spreadsheet to keep track of a few rows. I feel we're seeing a similar thing in the AI realm. One is a calculator for numbers, one is a calculator for words, but the mainstream advisor is not really there yet. They're buying focused point solutions, but it's not part of their daily work life in a meaningful way.

Suzanne Siracuse (08:52):
So Michael, when you say the early adopters... there are people using it in multiple use cases, which you would say would be more power users. But do you feel that the reason the wealth management industry is lagging a bit is because we're dealing with people's money and it is a heavily regulated environment?

Michael Kitces (09:25):
Probably less the regulation per se and more that we're managing people's money and giving advice in very high-stakes scenarios. In an adoption context, we get two constraining factors. Number one: "Oh my Lord, what happens to my data?" still freaks out the mainstream advisor. We have conversations with advisors where we have to say, "You realize if you don't ask it anything specific to your client, it has no client data? You can ask it to brainstorm your marketing plan." There is a gut-level response that AI equals bad privacy. That is a major fear point, and while it's educatable, it remains a hurdle.

(10:21):
The other piece is that this is a high-stakes business. We're giving advice about life savings. If you get it wrong, you can get sued for large amounts. Even short of that, most of us have organic growth rates low enough that losing a client because the AI did something wrong is a catastrophe. We are in a world where we're fighting for a 97–98% retention rate.

(11:18):
We see that flowing into how AI is getting used. It's why most dominant use cases are "human-in-the-loop" functions, not full automation. If I'm still involved, I can make sure nothing bad happens. We see very low interest in anything client-facing. Advisors actually don't want to put AI in front of their clients. The only use case in our research where a majority of advisors said they would never use AI was direct client service. There's no amount of efficiency that makes up for costing me a client who would have stayed for 20 years. That level of risk is just not tolerable. Most of the hunger we're seeing is for human-in-the-loop cases. It turns out most advisors really don't want to be automated out of the process.

Suzanne Siracuse (13:16):
Yeah, they want that human interaction combined with it. When I look back at this time last year, what's been the most significant change?

Michael Kitces (13:34):
Honestly, from the advisor end, I don't give a crap about AI at all; I just have things that need to be done in the business. If AI is the useful thing to solve it, then yay. But if it's not, don't waste my time pitching it. I think we are much more use-case oriented now. A year ago, we were in the hype phase where it was going to change the world. Now, we're a bit more grounded. It's just going to be another tool like Excel.

(15:22):
Advisors live in the problems that tech doesn't solve, and we always have. That's why Robo-advisors didn't disrupt us, nor did smartphones, the internet, or Excel. Forty years ago, we were using HP 12C calculators; when the spreadsheet was invented, we were supposed to be gone by 1987. Now we're supposedly gone again. It never happens because we live in the parts of the client interaction that are between the tech. The more the tech does, the fewer administrative things we have to do, and the more we get to do the client things we enjoy. Now we're finally figuring out which AI use cases actually save us time.

Suzanne Siracuse (16:27):
I couldn't agree more. People are just more knowledgeable about what AI can and can't do.

Michael Kitces (16:49):
I love the demo stages here. For those of us on the advisor end, there's a lot of, "Can you just show me what it does?" Stop the AI-speak and show me the software. If it's cool, I want it.

Suzanne Siracuse (17:09):
It's the one conference where people want to go to the exhibit booths to see the demos. Quick show of hands: who was here last year? (Pause) A lot of newcomers! Last year, you were bullish on meeting notes as a leading use case and skeptical about AI in investing. Has your perspective shifted?

Michael Kitces (17:51):
No, perspective hasn't shifted. AI is like a calculator for words. I hunt for places where there are lots of words it can help me expedite. Meeting notes are a wonderful example because many words are said and most of us don't like typing CRM notes. Now we're finding the follow-on use cases: getting information into the CRM, kicking off workflows, and prepping for the next meeting.

(18:48):
Other places words show up include crafting communication to clients. We have tools doing this for newsletters, and I think we'll see more email assistants that bring in the context of the client—email history and CRM notes. Some advisors are two-finger typists; even for fast typists, a thoughtful email takes mental energy. It's much easier to edit than to create. I also see interesting compliance use cases. Historically, we tried to catch wrongdoing with keywords and random spot checks. Now, an AI can read every piece of communication in real-time to find complaints, negative sentiment, or activity that sounds like fraud.

(21:20):
The one that intrigues me most is the CRM. Our data shows a slow but steady decline in satisfaction with CRM systems. Firms are getting bigger and workflows are more complex. More fundamentally, we are in a long-term relationship business. A client record might have 273 entries. An associate is told to "get up to speed," but are they really supposed to read every entry back to 2006? CRMs started as sales functions—a Rolodex with a few line items like "called prospect" or "sold product."

(23:33):
We went from a transactional world to a deep relationship business where we have hundreds of entries. That flat database structure is a terrible way to engage with high volumes of relationship information. AI is very good at consuming high volumes of information. I'm intrigued by the possibility of "AI-native" CRM systems built around relationship data. I'll give a shout-out to the Salentica folks (Slant); they are building in this direction. Financial advisors are unique because we have multi-decade relationships with dozens of touchpoints per year that need deep, client-specific context. AI is uniquely capable of solving that.

Suzanne Siracuse (26:23):
That is such a key thing—the unique situation of advisors and the length of time they spend with clients. Your recent tech survey found that most advisors don't want AI to completely automate a process. What does the ideal use case look like to them?

Michael Kitces (27:50):
The big headline was that only about a quarter of advisors wanted the AI to automate things. Viewed another way, 75% said, "Please don't automate." What they actually wanted was for the AI to expedite. I don't want you to take my 30-minute task and make it go away; I want you to take it and get it down to five or ten minutes. The fundamental difference is that in expediting, I'm still involved to make sure it's right. The closer the task is to the client, the less desire there is for automation.

(30:26):
Advisors don't want vendors to do the "last mile." This changes everything from how you market to how you show up as a vendor. If you really want to scare advisors away, say your AI "automates" things. We found that two-thirds of advisors are less likely to engage with a vendor if "AI" is in the name or on the homepage. Mainstreamers want to expedite.

(32:52):
From a product perspective, you build different interfaces and user flows if you assume a human is in the loop from day one. Some vendors have struggled because they built automation-oriented things and now have to retrofit their design. You are valuable to me even if you don't automate the email entirely, as long as you get it from 30 minutes down to five.

Suzanne Siracuse (34:30):
Great points. What about AI and marketing? Where are you seeing real traction?

Michael Kitces (34:54):
Candidly, we're not seeing huge traction yet. Prospecting tools are getting visibility—using data to generate a list of prospects. As someone who started cold calling from a phone book, that sounds amazing. However, the primary goal for most advisors is to get to the point where they never have to do that again. They want to grow through referrals.

(37:12):
Most of us aren't hyper-niche enough to go cold-outbound to everyone at a specific clinic. From the marketing end, ChatGPT and Claude are great brainstorming buddies for figuring out a target client, but you might only do that once every few years. For content creation, it's great at turning blank pages into first drafts. I don't know if that will be an off-the-shelf solution or just people using ChatGPT to find their voice.

Suzanne Siracuse (41:19):
It goes back to what you've been saying: it's less about AI and more about what challenges you're trying to solve. Here's a big-picture philosophical question: Is AI going to help smaller advisors scale, or will it just build the case for large firms to dominate?

Michael Kitces (42:56):
This is the golden age of the solo advisor. Full stop. Firms that used to need three or four staff can now be run by one person on a beach with a laptop. In the past, the best tech was at the wirehouses. Now, independent software providers serve thousands of firms, meaning small advisors have better access to technology scale than ever before. When small firms struggle with scaling, it's typically a people problem, not a tech problem.

(45:41):
Revenue per employee hasn't changed dramatically in a long time. Most firms run with $250,000 to $350,000 in revenue per employee. When you get to a certain size, you have to add people. Many firms sell in the $200 million to $800 million range because the founders don't like managing seven people, not because the tech failed. AI just provides another march forward in productivity.

(47:16):
Also, AI is good at coding because code is a language. We're going to see an explosion of "homegrown" tools. Advisors will build apps to solve their own problems and then sell them to their friends. That's how Orion, Redtail, and iRebal started. My prediction? The tech map will have 750 logos in less than five years. AI makes it easier to build and ship software, so the scarcity is no longer the ability to build—it's the ability to distribute. The cycles will get faster.

Suzanne Siracuse (50:49):
What's your bold prediction to close us out?

Michael Kitces (51:04):
The dominant narrative is that technology makes us incrementally more efficient, but we have to watch out for fee compression. Twenty-five years ago, a three-producer firm might have had eight support staff. One person's entire job was filing statements. That job doesn't exist anymore—software automated it. If I look at a three-person firm today, they might have two support staff instead of eight.

(52:46):
But here is the interesting effect: the median profit margin then was 30%, and it's 30% today. We got rid of "Betty" the filer, but we bought portfolio management software that cost more than her salary. We repurposed the people to do deeper things. Twenty-five years ago, an advisor might have had 500 clients; today, they have 100.

(53:32):
After every technology wave, client loads don't go up; they go down. We get deeper. We value-add our way up. Clients get a much richer proposition—continuously monitored, tax-loss harvested portfolios. The future of this tech is that we will work with 20–30% fewer clients but deliver 50% more value. This will lead to an even worse talent shortage because we'll need more advisors to handle the same number of people. If history is any guide, the endpoint of amazing tech is going deeper with fewer, not doing more with more.

Suzanne Siracuse (55:13):
That is the headline right there. Michael, as always, fascinating and informative. Thank you all for your time.