Editor's note: This article appears in the March 2021 issue of DS News, available here.
As SVP of Servicing Product Strategy at ServiceLink, Yvette Gilmore is responsible for developing products and services that support strategic servicer client initiatives. She also supports ServiceLink's EXOS One Marketplace, the only AI-powered asset decisioning tool of its kind that uses predictive modeling to determine the optimal disposition strategy for properties in default. Gilmore has more than 20 years' experience leading servicer relationship and performance management efforts for leading Fortune 500 financial service organizations. Prior to joining ServiceLink, Gilmore spent more than a decade at Freddie Mac, where she held several leadership positions and most recently served as VP of Servicer Relationship and Performance Management. Prior to joining Freddie Mac, she led the loss mitigation departments at IndyMac and Washington Mutual.
Speaking to DS News, Gilmore discussed the current state of servicing, COVID-19’s impact, and how artificial intelligence and machine learning can be the ‘next best thing’ to having a crystal ball for servicers.
How has COVID-19 impacted the servicing industry and its operations?
While servicers key concerns have remained largely unchanged—like mitigating risk, streamlining operations, minimizing losses, reducing costs and elevating the borrower experience—today’s extraordinarily active marketplace is truly putting them to the test in myriad ways. Those whose portfolios are growing in number and/or variety of properties face the added challenge of managing this new influx without slowdowns or errors.
From a risk management standpoint, servicers want a rules-based approach to every aspect of their operation. They are seeking to not only facilitate their decision-making processes and comply with evolving regulations, but to also ensure a level of consistency and certainty for the customer. Customers rely on servicers for fast, transparent, well-informed decisions and for consistent answers and results; regardless of the company representative they may speak with.
As COVID-19 continues to impact the servicing industry, many are choosing to lean-in to technology to streamline their approach.
What is preventing some servicers from embraced technology to the extent they could?
We actually don’t find that servicers are reluctant to use technology. Typically, the issue is finding technology that is easy, cost-effective and attentive to the portfolios’ unique characteristics. Servicing is a complex business that can be heavily nuanced. To put it bluntly, a “one size fits all” tool that is so expensive and difficult to integrate that it defeats the purpose of the technology’s utilization in the first place makes adoption even more of a hurdle. Service providers that offer customizable solutions to fit the unique needs of the servicing industry are best positioned to disrupt the status quo of asset decisioning in 2021.
While servicers key concerns have remained largely unchanged—like mitigating risk, streamlining operations, minimizing losses, reducing costs and elevating the borrower experience—today’s extraordinarily active marketplace is truly putting them to the test in myriad ways.
How can AI/machine learning serve as a “crystal ball” for servicers?
Servicing on its face is simple: there’s an asset, and it needs to be managed. But there are a host of incremental decisions that could make consequential impacts on the ultimate outcome. Machine learning and analytics opens the door to more rapid and confident decision-making and can help make sense of an ever changing economic and mortgage industry landscape.
In the past, servicers needed to base their decisions on their own historical information and best practices. Today, they can rely on machine learning to take in and process massive amounts of data from a variety of sources—like historical operational data and leading industry databases—to produce a new level of intelligence. Machine learning can streamline the evaluation process by identifying patterns that enable it to predict outcomes under various sets of circumstances; for example, how a specific property might fare at auction or in the CWCOT second-chance program, or what level and cost of repairs a particular property might require if held for a certain period of time.
Leveraging technology and AI/machine learning can make a world of difference for time-strapped servicers, but it doesn’t absolve them from the decision-making process altogether. While it’s the closest thing we have to a “crystal ball” in servicing, machines and technology should always be a complement-to servicer’s decision-making, not a replacement-for.