Presented by Envestnet
Personalization-at-scale is a key strategy for fintechs to deliver hyper-relevant products and services. Learn how top fintechs are delighting customers and building strong relationships with AI-enabled platforms and data sources in this VB Spotlight.
“There’s a direct correlation between customers loving your products and revenue,” said Bala Chandrasekharan, VP of product management at Chime. “Customers are much more highly engaged and likely to recommend your product to others. Referrals are an incredibly powerful viral marketing channel compared to paid marketing.”
And to do that requires true personalization. Chandrasekharan, David Goodgame, COO at Tricolor Auto Group and Eric Jamison, head of product — banking & technology at Envestnet spoke about how personalization-at-scale offers fintechs a larger competitive advantage than ever, and how AI and analytics are changing the game, during a recent VB Spotlight event.
The case for personalization
For Tricolor Auto Group, an auto retailer and direct lender in the used car space, personalization means digging into the deep-down desire that lies behind a customer’s request.
“All of our efforts here, when it comes to marketing and even the inventory we select to put on our sales lots, are all focused on what we call jobs to be done,” Goodgame said. “We looked at our entire business and said, ‘Whenever a customer comes to us, what are they looking for us to provide? How can we, in our marketing efforts, in our dealings with them, in our customer service centers, make sure we’re addressing that?’”
A “job to be done” could be a customer being insecure about their credit, aspiring to an American dream kind of lifestyle or tackling big projects — and personalized ads sell that value proposition in the form of a car or loan.
Achieving that requires marrying the consumer relationship with the use case, which is where good data is crucial, Jamison said. As a B2B2C service provider, Envestnet comes in when a lender might need to fully understand a loan applicant outside of their credit report — or if they don’t have one. That data might include cash flow information, like earnings and expenses from a bank or another provider.
“It really helps to personalize that application for that consumer, to help that provider make a more informed decision and to help connect the dots that maybe don’t show up in a traditional manner for that consumer,” he explained. “It’s marrying our ability to do something with the consumer need and ensuring that those things are aligned. That’s going to drive the best outcome.”
How AI and machine learning change the game
“Our AI risk model is the secret sauce behind our company,” Goodman said. “What we believe is that if our customer goes anywhere else in America, they’re all thrown into one bucket. That one bucket is a very predatory-looking set of terms for that customer. It’s going to be the state max interest rate. It’s going to be an inferior product. Affordability for that customer is never going to even become part of the conversation.”
According to Goodman, about 90% of the applications the company gets don’t have information in any of the credit bureaus. But the large amount of data they collect, from a broad array of arenas, can identify what he called a more reliable scoring system than a FICO score, so that they’re able to offer low rates to someone with no credit data.
“Our risk model — it enables us to sell cars that have very low losses,” he explained. “We’re able to then lower our prices, which attracts more borrowers. We do more of this, and the flywheel effect begins to happen, because as we’re able to get more data and get more applicants, our model gets smarter. It gets tighter. We can reduce terms even more. We’re taking more and more risk out of the equation, so we’re able to offer better terms. As we offer better terms, we get more customers. That flywheel effect becomes real.”
And in that way, they’re helping to uplift an often-overlooked demographic, so they’re able to begin establishing a financial history and building credit.
Taking all the application data also helps them move the risk model higher in the funnel — and the higher in the funnel they can do that, the more personalized marketing can get. If a customer comes through a particular channel, their interests, needs and background can be identified to ensure the content they receive is relevant to them, thereby increasing conversion rates because they feel as if their needs — for a particular style of financing, price range, etc. — are being seen and met.
This holds true of Chime as well, which aims to offer accessible financial services for Americans who might have been denied traditional banking services.
“In that world, when you don’t have as much explicit public information available, AI and ML play a huge part,” Chandrasekharan said.
For example, it’s important to differentiate the negative marks on a customer’s records between irresponsible behavior and someone who encountered unlucky circumstances. The question becomes how to read a customer’s behavior pattern — how they have used the platform and products previously, what a negative event looks like, and what value the customer might bring.
“That’s where AI and ML play a huge part in trying to understand how we can separate the good from the bad,” he said. “In fact, what that enables is the flywheel effect previously discussed. You can drive an excellent, delightful member experience in that case when you know they’re a good customer. That can be a huge differentiator. These are the moments that matter for a customer. When you’re able to use AI and ML to get it right, that ends up transforming into a delightful experience, which means they’re likely to be loyal customers for a long time. They’re likely to refer your products to others.”
The power of data comes from identifying patterns, Jamison said, which requires as large a pool of data as possible. Envestnet works from a set of about 40 million consumers, and their regular transaction activity that allow the company’s data scientists to identify crucial behavioral similarities, he said.
It could be identifying ways to act in their own financial portfolio to save money or helping a financial advisor scale by bringing wealth management advice to the masses. It helps eliminate the danger of taking a one-size-fits-all approach, which means missing the bulk of your customers.
“We’re all individuals and we’re all unique, but our patterns typically align with someone else’s,” said Jamison. “We can start to align those intersections to help identify the next best actions. It can help that consumer achieve a better financial outcome. Our platform and the AI and machine learning they apply helps consumers throughout that life cycle. We can bring to bear the right solution at the right time for our client to help their customers. That’s really the power of data, helping understand consumers across these broad segments in a very targeted and specific way.”
To learn more about driving nuanced hyper-personalization at scale, overcoming data and privacy challenges and more, don’t miss this VB Spotlight.
- How FinTechs are using personalization at scale to gain a competitive advantage
- Various AI-enabled technologies to securely collect, enrich, and analyze financial data
- How advanced analytics and transactional data can deliver valuable customer insights
- Ways to identify customer acquisition, cross-selling, and upselling opportunities
- How to create personalized experiences that are relevant and emotionally “sticky”
- David Goodgame, COO, Tricolor
- Bala Chandrasekharan, VP of Product Management, Chime
- Eric Jamison, Head of D&A Product — Tech & Bank Product & Design, Envestnet
- Mark Kolakowski, Freelance Writer & Editor; Lecturer; Former Financial Services Professional (moderator)