Top 3 ways AI helps mortgage servicers combat the high cost of borrower communications

The mortgage servicing industry is under immense financial pressure. With interest rates high and home prices still rising, new loan originations and home sales have declined sharply to their lowest level since 2010. With fewer loans and less servicing rights to compete for as a result, servicers are facing stagnant revenue growth and heightened pressure to manage costs.

At the same time, servicing costs are rising as well. High interest rates and inflation have reduced the percentage of financially healthy borrowers to just 41%—down from 52% in 2022. In fact, the number of borrowers in default and loss mitigation has surged by 20% year over year. According to the Mortgage Banker’s Association (MBA), the cost of servicing non-performing loans is 11 times higher than performing loans, largely due to the increased volume of communications required to guide borrowers through complex loss mitigation processes. Each additional letter, notice, or phone call increases servicing costs.

In this environment, mortgage servicers must look for new ways to increase operational efficiency to maintain profitability. Borrower communications represent an area of the business that is ripe for optimization. The processes of creating and managing communications are highly inefficient and costly.  A thoughtful and careful approach to applying AI to address some key elements of borrower communications management could lead to significant cost reductions and efficiency gains. Leveraging AI’s ability to analyze, optimize, and generate high-quality content at speeds far faster than humans can help streamline content management, improve borrower experiences, and accelerate digital transformation without introducing undue risk into the process. The question then becomes: how can servicers best leverage these capabilities?

Three key cost centers that are impacted by inefficiencies in borrower  communications management represent the biggest opportunities for cost reductions:

1) Lowering call center volumes with clearer communications

Research from SQM shows that 41% of call center inquiries come from customers who initially tried to resolve their inquiries through other channels but were unsuccessful. For mortgage servicers, a large share of these inquiries likely stems from borrowers seeking information that was already provided to them in a written communication. However, because borrower communications are often overly complex and filled with industry jargon and legal language, many borrowers either miss or misunderstand the content. Frustrated or confused, they then turn to the call center for help. Providing clearer, more accessible communications would result in fewer borrowers turning to the call center for clarification, which could easily translate into hundreds of thousands in annual savings for servicers.

AI can be used to improve clarity. Natural language algorithms can analyze communications in seconds to measure readability levels using benchmarks like Flesch-Kincaid scoring. Generative AI can then just as rapidly rewrite the communication to meet a desired readability level or apply plain language principles—simplifying terminology, breaking down complex sentences, and restructuring information to improve the likelihood the recipient will understand it.

Translation is another area where AI can support better borrower understanding. With 22% of Americans speaking a language other than English at home, providing communications in borrowers’ preferred language could significantly reduce call center volumes. Today’s highly sophisticated AI translation engines can now produce accurate translations up to 20X faster than humans. AI can also be used to validate translation accuracy, reducing the human-in-the-loop requirement to reviewing exceptions and issues caught by the algorithms. This gives servicers the ability to translate communications more cost-effectively than they do today and provide better support to diverse populations.

2) Ensuring compliance with regulatory updates

The cost of ensuring compliance across different types of borrower communications and states has never been higher. In particular, loss mitigation communications are among the most highly regulated types of borrower correspondence and should the Consumer Financial Protection Bureau (CFPB) find that a servicer has failed to meet requirements, they may issue hefty fines, often in the millions.

With hundreds or even thousands of templates in their libraries, servicers often struggle to adapt to these regulatory changes quickly and efficiently. It’s a time-consuming and costly task to manually review and update every document template, especially when servicers may not have visibility into all the letters currently in production.

AI can significantly streamline this process with its intelligent search capabilities. Using natural language processing (NLP) algorithms, AI can analyze your entire communications library and locate exact and semantically similar content matches. Instead of manually looking through every document to find all instances of a disclosure, servicers can find every occurrence of a disclosure in seconds and focus their efforts solely on updating the impacted communications. Advanced algorithms can even find matching content across different languages, ensuring consistent updates for multilingual communications and reducing compliance risks across the board.

3) Reducing print and postage costs

So many mortgage servicing communications are still printed and mailed despite increasing consumer preferences for digital communications. Yet, there are easy savings to be had. According to JD Power, 43% of the 52% of borrowers receiving paper statements actually review them on digital channels. Simply stopping the mailing of paper statements to this group could save servicers hundreds of thousands of dollars each year in postage costs alone.

In other cases, the work of creating content suitable for digital formats is required. Shorter communications such as email, SMS, or even portal and mobile app pages require reworking of content to fit the medium. Generative AI can streamline this process by automatically rewriting existing communications to suit different digital channels. In more advanced content management systems, content authors can develop a communication and ask AI to generate the appropriate formats for each channel—optimizing content length, style, and structure while ensuring its overall meaning is preserved. By automating these transformations, servicers can offer a richer, more engaging digital experience that encourages more borrowers to adopt digital delivery of their communications, without adding to the content management burden of their internal teams.

Getting the best results with AI

AI isn’t an end unto itself. Its adoption should only be employed when it can either improve the quality or speed of a business process or reinvent a process. While tools like ChatGPT and Claude are easy to access, using them in isolation from the systems that manage the communications may not provide the security mortgage servicers require or truly drive the efficiency servicers are looking for. Fortunately, some modern borrower communications platforms now incorporate AI directly into their systems, enabling servicers to leverage these capabilities within the same secure environment where they manage their content. These solutions can ensure data security, provide more consistent results, reduce manual content rework, and maximize the overall efficiency of the process.

Looking Ahead

When it comes to AI, these use cases are just the beginning. As generative AI solutions continue to develop and comfort with them increases, mortgage services will have both a deeper and wider range of use cases available to them to improve the efficiency of operations and effectiveness. By embracing AI now, servicers will not only be better equipped to handle current financial pressures, but gain valuable experience and a head start in leveraging new capabilities as the technology evolves. Starting early with AI will enable servicers to be well-positioned to stay competitive and agile, ready to respond to market shifts and deliver better experiences as AI continues to reshape borrower communications.