“I spend pretty much all my time thinking about where the auto lending business is going,” says Charles Sutherland, the Chief Strategy Officer of defi SOLUTIONS. “It’s been six months since defi merged with Sagent Auto to form one of the biggest software platforms for auto and personal lending, with capabilities in originations, servicing, and business process outsourcing for both direct and indirect lenders.”
We thought it would be a good time to check in with Sutherland, who is busy at their Dallas head office running corporate and product strategy for the combined company.
“A historic tagline of defi was ‘Built by Lenders for Lenders,’ and that holds true for both sides of the newly merged company,” says Sutherland. “We see ourselves as a provider of systems of record and core processing services. We’re building an ecosystem that makes the lending experience better for everyone.”
One of the biggest investments toward that end, says Sutherland, has been around robotic process automation (RPA) across the entire lending and leasing lifecycle. “Almost a quarter of the full-time employee capability we had at the beginning of 2018 has now been converted from human process operation to RPA.” In one example, defi eliminated enough manual hand-offs and approvals to shrink the time it takes to process bills between lessors and lessees from three and a half days to within an hour using process automation.
Sutherland says defi’s strategy is informed by a handful of industry trends. The first is the migration of lending software applications to the cloud, which Sutherland sees as a clear win for lenders. Over the last 18 months, defi has built a lot of native cloud applications that take advantage of the investments in tools, security, and stability of AWS and Microsoft. Cloud brings standardization, which confers benefits in efficiency and platform stability. “You avoid all those one-time security reviews and assessments, and you get a greater degree of comfort around how that hosting infrastructure is being operated and managed,” he says.
The second big shift is a decentralization of lending services. Lending origination (LOS) systems, he says, typically have been these monolithic, system-of-record applications that try to do everything. defi SOLUTIONS has turned that on its head by making the LOS a central component of a broader ecosystem tied together with application programming interfaces (APIs). “We’re making it easier for lenders to do more experimentation with new partners by moving away from hard-wired partnerships to a microservices model built around innovation hubs,” says Sutherland.
Lenders want to be able to experiment more easily with data, he says, and don’t want to go through a massive integration exercise only to find out that the correlation or efficacy of a particular data set is low and then have to pull it all apart again. They want to be able to run trials and test benches and get a quick sense of those new capabilities.
Sutherland cites the integration of Zest AI machine learning and defi SOLUTIONS as an example of what an ecosystem can provide, especially on the originations side. “Originations are moving from being a rules-based workflow to being powered by insights fed by alternative data and machine learning. No one party on the tech side can do that alone. You have to have an ecosystem of partners that can do the kinds of mashups of data and process capability like we’re doing with Zest AI, to realize benefits for clients.” (You can find out more about Zest’s integration with the defi platform in this recent webinar.)
“Lenders want to pick and choose the best services and components and know that those components have been tested and thought through, rather than being something they have to produce on their own. Ultimately the responsibility of extracting value out of a partner ecosystem will move from the lender to the LOS provider and the partners they’re working with.”
“The upshot of all of these trends is that you get a lower cost of innovation, which helps lenders differentiate and respond to macro and micro trends,” says Sutherland. “And that’s the best way to manage risk.”
While staying up on current trends, Sutherland’s job also requires him to see around corners, specifically at how vehicle ownership will change in the next five to ten years and how that will impact his lender clients. “Things are moving beyond retail stores, loans and leases to a world of subscriptions, ride sharing, car sharing, and connected-car models.” He points to services such as Fair, Maven, Care by Volvo, and dealer-based programs, which are all shifting driver attitudes about long-term ownership of a single car.
Sutherland sees machine learning playing a big role in helping lenders service these new ownership models. “While it’s easy today to know when to intervene or engage with fixed-length contracts, auto subscriptions are going to be a lot more variable,” he says. “To predict when drivers are going to want to terminate, you're going to have to collect as much data as you can to predict what’s going to happen to your investment in a subscription fleet. That’s a classic example of how machine learning capabilities will allow you to have a degree of insight in a way that wouldn't be possible before.”
Sutherland sees 2020 as the year the connected car market expands well beyond the early adopters to the big OEMs, as drivers seek to download more digital services and products to their vehicle. He’s got his eye on platforms launching in the next few years such as the Mustang Mach-E and the Porsche Taycan. “Tesla has shown that the sky's the limit for the sorts of things that can be placed into the vehicle. It creates a second life for auto financing,” he says. “But will these services be a per-usage or one-time charge? Or, in some cases, will they be add-on to your loan or lease? Lenders will have to support a range of options.”
The big question is what’s not going to change, and that’s the reliance on data. “Financial technology is still about ones and zeros flying back and forth, but the expectations around what you should be able to do with the data generated is changing greatly.” He says lenders will expect to do more sophisticated planning and modeling and consumers will get more insights-driven analysis about what they’re able to afford.
Auto lending, specifically, has for far too long been a back-and-forth model where lenders adjust offers a little bit here and there until the rates or terms are good enough. “I think we're moving to a world where the lender will be able to present all the possible combinations and matrices that may work for a consumer. That will enhance the borrower’s sense of empowerment while making the process more efficient for lenders – and those are really good outcomes for the industry."