Insight

Mortgage modernization: Processing made intelligent

Four actions that help get digital mortgage lending right.

Teresa Blake

Teresa Blake

Advisory Principal, Financial Services Solutions, KPMG US

+1 704-747-8852

Michael Herman

Michael Herman

Principal, Advisory, C&O Financial Services, KPMG US

+1 212-954-3898

Braden Mark

Braden Mark

Partner, Advisory, Financial Due Diligence, KPMG US

+1 303-382-7980

An insight brief from KPMG Connected Enterprise for Banking.

A surge in home buying, coupled with dramatically low interest rates, has created a historic increase in mortgage demand. Add a global pandemic that requires social distancing to the mix, and you have the perfect storm for industry disruption.

Against a rapidly shifting landscape, lenders have responded to the growing need for lowering costs and improving service by adopting new technologies and addressing underlying processes.

Yet, despite notable progress, improvement remains incomplete. Much of the mortgage industry is still fragmented and labor-intensive and rarely does the customer have a seamless experience.

Next-level technology—from the customer’s point of view

Getting a mortgage is one of the most significant decisions made in a lifetime. It is also one of the most complex and bureaucratic. Customers are repeatedly asked the same questions, papers are shuffled back and forth between departments, and rigid conformity to redundancy creates never-ending delays.

In contrast, borrowers want an online loan application experience with user-friendly processes, transparent pricing, minimal information uploads, quick closure—and to be notified of their status at every step of the application process.

All technology improvements should be regarded in the effort to satisfy these customer expectations.

The good news is that with the advent of software technologies such as artificial intelligence (AI), machine learning (ML), and robotic process automation (RPA), a digitized mortgage process is made faster, smoother, cheaper, and more accurate and compliant—pleasing both the borrower and the lender.

Data tells the story

Data—the information captured within your four walls, including structured and unstructured data—as well as data from external sources—can be aggregated, and using AI and ML models, predict applicant behavior and flag at-risk loans. They can also help identify emerging risks, such as customers that are likely to switch to other providers for better rates or terms.

It’s often difficult for traditional mortgage lenders to see the entire customer journey because data is dispersed between multiple channels and touchpoints. Thus they lose the insights from all that data to drive a better customer experience.

But not every bank can invest in a centralized customer data platform. How do you prioritize when resources and time are limited?

Where to begin

Begin by prioritizing the customer experience. That means re-examining your processes and loan origination system (LOS). The LOS is the center of a bank’s ecosystem, and there are multiple third-party technology solutions to improve that system.

An end-to-end engagement model starts with a digital front end that makes it simple for clients to come in, authenticate themselves, upload and validate documents—and then connect with an agent either virtually or in person. At the back end is a Customer Service Center that helps them self-direct should any issues arise. Data is automatically fed into the center so that service agents know exactly why the individual is calling and they’re prepared with answers—and maybe even an opportunity to upsell or cross-sell.

In our experience, a thorough shadowing of existing systems and processes, viewed through the lens of the customer journey, often reveals redundancies—and opportunities to refresh the workflow using RPA or AI/ML. For example, we’ve recommended the elimination of underwriter involvement in multiple areas of the loan application, restricting their work to only those areas they influence. There’s no need for underwriters to be involved in 100 percent of the loan processing.

Four actions to focus on

  1. Target: Get hyperfocused on identifying potential customers—those most likely to close on a loan. ML-based models using multiple data sources will help predict conversion rates. Leads can be generated through your website, social media, referrals, community contacts—and by extending the relationship of your current clients, such as approaching those with existing auto loans to see if they would like to explore a home equity mortgage.
  2. Engage: Customers crave speed, online access, and a personalized approach to service. They want their lender to anticipate their needs within a frictionless, omnichannel journey. Customer experience is the primary differentiator for mortgage lenders. If you can deliver an exceptional, personalized experience, the result will be that your customers buy your brand, not your rate.
  3. Scoring: Improvements in credit scoring will reduce delinquencies and increase profitability. Collecting data on factors such as borrower credit history, wages, employment, assets, tax, and debt will transform data into insight into the projected loan performance. Previously hidden financial relationships are uncovered, and with growing volumes of data, predictions rapidly become more accurate. Better understanding of a qualified lead’s profile will continually refine your credit scoring engine.
  4. Distribution: Ensure that you’re funneling leads to the proper loan officer—the best skill set match and communicator for that lead so that you have the highest conversion rate. And make sure that the employee has all the information they need to take the next right action.

Lessons learned

The traditionally conservative mortgage industry has not been quick to modernize. But the global pandemic opened doors to new ways of lending and taught us that loans could be opened and closed remotely. Those that were ahead of the modernization curve have benefitted greatly.

The pressure to digitize is valid. Digitization transforms a cumbersome, time-consuming process into an automated, efficient one that increases the integrity and usefulness of data, leading to better-informed decisions and an expanded market for credit.

With the available new technologies, coupled with how deeply we understand the borrower and human behavior involved in the loan process, there’s never been a better time to be in the mortgage business.


Listen to the companion podcast