Chat windows have been taking their performance-enhancing Watson pills, and they’ve spawned a new generation of fast-learning “chatbots.”
Text-based virtual assistants are poised to play a big role in the further automation of the retirement plan participant communication and education experience. These bots can be annoying, especially when they appear unbidden. But they’re much cheaper than human phone reps, capable of “machine learning” and programmable never to violate ERISA regulations.
Grant Easterbrook wants to use them to reduce the cost of 401(k) advice. The twenty-something CEO of Dream Forward, a 10-person venture-backed startup, is marketing low-cost bundled 401(k) plans in the small-plan market. Proprietary artificial intelligence (AI) bots that can “coach” participants into better savings habits are the core competency of his firm.
Dream Forward will outsource other 401(k) chores to strategic partners. Recordkeeping tasks will be handled by software from Schwab Retirement Technologies. Barnum Financial Group, part of MSI Financial Services, Inc., (the former MetLife Securities, Inc., and now part of MassMutual) provides the human financial advice as needed. Expand Financial is the third-party investment fiduciary and MidAtlantic Capital Group is the custodian.
Like other next-gen chatbots, Dream Forward’s chatbot will be active. Through coding and trial-and-error, it can steadily “learn” about individual participants from their online financial activity and initiate an interaction via smartphone or computer. That’s the “artificial intelligence” or “machine learning” part of it. Chatbots are related to iPhone’s Siri and Amazon’s Alexa—but with communication based on textual analysis instead of voice analysis.
As Easterbrook (right) put it, our “401(k) product includes an AI technology that is designed to address the questions, concerns, or excuses that keep employees from saving enough money for retirement, such as ‘I’ll save more when I’m older,’ or ‘I’ll get serious about retirement after my kids graduate.’ Our technology provides instant help to keep employees on track and saving enough for retirement. The Barnum professionals will bring a human face and perspective to the process.”
Chatbots inspired a wave of excited chatter in tech blogs and venture cap circles beginning last April, after Facebook invited developers worldwide to develop chatbots for its new Messenger Platform. These interruptive droids send screen messages when they sense your presence through your computer or smartphone. When tied to big data, they’re capable of knowing your identity, shopping history, preferences and tastes. If your phone is turned on, they have the potential to anticipate what you might need or desire at any moment.
Cleverly intrusive, chatbots can insinuate themselves into your field of vision like Microsoft Word’s bygone (and much-hated) virtual assistant, Clippy. But, for firms facing cost pressure who want to reduce the headcount of their call centers, they could be a competitive necessity. “Chatbots will cause a near-term disruption in how businesses interact with consumers, and a long-term paradigm shift in how people will interact with machines,” wrote a technology writer in an article at VentureBeat.com article last spring.
Though perhaps hyped a bit by investors, the market for virtual assistants is expected to be huge. According to a report published last August and offered on the website, chatbots.org, “the global intelligent virtual assistant market is projected to expand at a remarkable CAGR (compound annual growth rate) of 32.8% from 2016 to 2024. The opportunity in the market was pegged at $627.7 million in 2015 and is expected to be worth $7.9 billion by 2024.”
Front-row seat on the future
Easterbrook is bringing this loquacious technology to the small-plan 401(k) world. The son of noted journalist Gregg Easterbrook, he graduated from Bowdoin College in 2011 with a degree in international relations then spent the next several years at Corporate Insight, the research firm that gathers competitive intelligence by maintaining accounts at hundreds of financial firms in order to lurk online as a kind of mystery shopper.
As a fintech analyst there, Easterbrook watched the robo revolution unfold. “I spent four or five years talking to robo-advice company founders, and I saw which ones succeeded or failed and what did and didn’t work,” Easterbrook told RIJ. “Robo is great but it can’t replicate the coaching and handholding that live advisors provide. We didn’t have an Aha! moment. It was more of a gradual coalescing around what the gap in the industry was. We’re the first to build AI in the area of replicating financial coaching.”
Networking with fellow Bowdoin graduates and others, Easterbrook started assembling a virtual bundled 401(k) provider. “I’m not a programmer. I’m the CEO. I provide the vision, the sales and marketing,” he said. One of his financial backers is Newark Venture Partners, a New Jersey firm that invests up to $1 million each in “seed and Series A” companies. NVP receives funding from Audible (an Amazon company), Dun & Bradstreet, Prudential Financial and other backers. He also got money from VC Fintech Accelerator, a unit of FIS Global that sponsors bootcamps for fintech entrepreneurs and provides seed capital for a select few.
“Our 401(k) plan includes an independent 3(38) advisor,” Easterbrook said in an interview, referring to the section of the ERISA statute that defines trusted pension investment managers. “And we’ve got the compliance expertise behind the scenes since we’re built on top of an existing recordkeeper.
“We don’t generally disclose our recordkeeping partners, but I will say we’re built off SRT—Schwab Retirement Technology. Our business model is that we charge an AUM fee. We charge 65-75 basis points to employee, which is less than half the cost that small plan participants would ordinarily play. We’re not reinventing the wheel. It’s a bundled 401(k) plan,” he added.
About ‘AI,’ aka machine learning
Easterbrook and his team have jumped on a trend that recently fielded its own journal—Chatbotsmagazine.com. According to the blog’s creator, Matt Schlicht, “A chatbot is a service, powered by rules and sometimes artificial intelligence, that you interact with via a chat interface. The service could be any number of things, ranging from functional to fun, and it could live in any major chat product (Facebook Messenger, Slack, Telegram, Text Messages, etc.)”
On his blog, Schlicht describes how a retail chatbot might work. If you’re buying shoes online, for instance, you might go to a retail website, find shoes you like, click and pay. In the future, a chatbot will greet you, ask intelligent questions, make suggestions and close the deal.
In the banking realm, a sophisticated new generation of chatbot apps has already arrived. In October, Bank of America introduced “Erica” (as in Am-Erica), a chatbot that according to CNBC “will use artificial intelligence, predictive analytics and cognitive messaging to help customers make payments, check balances, save money and pay down debt. She will also direct people to look up their FICO score and check out educational videos.”
As one tech consultant recently observed, “These new technologies now let banks reduce friction and provide a higher level of engagement at far lower operating cost.” Easterbrook envisions similar benefits in the world of participant engagement in the process of saving for retirement.
“Let’s say you lowered the amount of your contribution to your 401(k) account,” Easterbrook told RIJ. “You’d get a pop-message on your desktop that said, ‘What’s on your mind?’ If you say, ‘I can wait until I’m older to start saving for retirement,’ then AI will try to coach you through that.” When the chatbot reaches the limits of its capability, it will pass the participant on to a human advisor.
Easterbrook talks about an “AI-driven experience.” “The notion of ‘artificial intelligence’ is somewhat misleading. The AI function won’t be able to tell you the meaning of life, but it will recreate the experience of talking to a human being,” he said.
“Artificial intelligence is the next stage of robo-advice. A lot of the media coverage of AI has been pessimistic, like the coverage of self-driving cars. Dream Forward is a positive application of AI. It makes the human advisor better at his or her job. We’re not trying to mislead people; we’re trying to help them out when they’re about to make a bad choice. It’s not, ‘Hi, I’m Sara Smith.’ It’s not phone tree technology—that’s a separate issue. It’s just someone trying to answer your questions in plain English.”
But don’t FAQs webpages do that already? “FAQs don’t deal with the emotional aspect of the questions. Part of our vision is that we work with human advisors. They always have a role. When you think about how financial advisors currently work with 401(k) plans, they go onsite and hold seminars, but most people never show up at the seminar. With our system, we can tell the advisors what’s going on with the participants. We can say, ‘Forty people aren’t saving at all, and ten aren’t saving for college,’ etc. We have the technology to intervene, and the advisors go in to finish the job.”
Easterbrook has no immediate plans to take Dream Financial public or to be acquired by a large asset manager or insurer, as several other fintech and robo-advice companies have been. “We’re not building for acquisition,” he told RIJ. “We see a huge opportunity to be part of a next-gen solution that everyone uses. If you’re thinking, ‘Let’s just get to a certain point and then someone will buy us,’ then you’re making wishful decisions instead of rational business decisions. That leads to bad choices.”
© 2017 RIJ Publishing LLC. All rights reserved.