Mortgage origination companies are using technology to operate more efficiently and better serve their clients. Even with new technologies, loan origination costs have climbed precipitously while profitability has declined over the last decade. What gives?
One primary culprit for the climb in costs per loan is that most technology targets streamlining the front-end side of the business but neglects the back-end processes. There may also be a lack of prioritization resulting in inefficient processes or a lack of consistency in performance standards for employees. In any system, input equals output. So, what are your measures of success? How are you measuring performance? If you see anything out of balance, it can quickly lead to bottlenecks and costly delays in the loan process. Data scoring is a method for identifying bottlenecks and making targeted improvements across the entire origination process to increase efficiency.
Using Scoring to Measure Mortgage Operations
Processing a mortgage is an extremely complex and labor-intensive endeavor. Identifying the bottlenecks and inefficiencies in the process is problematic as well. Mortgage companies often have limited measurement of their mortgage operations performance or measure dozens of individual key performance indicators that must be manually analyzed. This process is confusing and time-consuming.
Manual data entry and collection on a monthly or quarterly basis from disconnected systems, not in real-time, results in stale data. Stale data does not provide the insights needed to take immediate action to improve efficiencies and, ultimately, profits. To take appropriate action, it’s essential to quickly view the current, historical, and projected health of mortgage operations in a simple, easy-to-understand method. Real-time scores provide an efficient and immediate solution to both processing and consuming insights from analytics.
How Does Scoring Work?
Scoring is the process of generating values from a trained machine learning model based on designated criteria. Raw data elements are extracted from mortgage operations systems, including loan origination, customer relationship management, accounting, and other sources - and stored in a data repository. Artificial intelligence engines then read the raw data and create certain derived data elements. Next, derived data is compared to key performance indicators, company benchmarks, and industry benchmarks, then presented in a singular score. The score and associated data may then be presented directly to end-users for interpretation and interaction.
How can Scoring Improve Operations?
Scoring provides a means of assessing current operations and provides opportunities for predictive analytics to inform your action plans for future operations. With real-time inputs, the scores inform your decisions to prioritize and make corrections to deliver increased efficiency. Additionally, the types of scores, and their meanings, can be modified and customized for your workflows to fit your unique goals and objectives. Let’s take a look at a couple of examples:
Focusing on the Right Applicants
Loan origination companies focus on qualifying applicants but place little to no emphasis on the probability of an applicant closing a loan. The qualification process is like a pass-fail test - if an applicant meets specific requirements, they will qualify with a good credit score, debt-to-income ratio, employment history, etc. A probability score, however, weighs the individual elements and compares them to historical data for loans that have closed, as well as those who have not, to present a complete picture.
One lender we worked with scored all loan applications on a 1-10 scale based on a number of contributing factors they could capture at the front end - ownership history, DTI ratios, previous relationship with them - overall, 23 distinct criteria. Scoring allowed them to triage the submissions and fast track the applications with the highest probability of closing. They could see how applications fell percentage-wise within different categories and modify marketing and processing components to drive higher quality submissions.
Making Targeted Process Improvements
There are numerous stages in the loan origination process, including loan processing, underwriting, closing, and other functions. These primary stages are similar from one company to another, but the individual elements can have variations. Scoring provides a normalized value that makes industry or internal comparison quick and easy. Target metrics for key mortgage benchmarks are established by normalizing the data and controlling for various factors. The metrics provide real-time alerts, such as when specific processes are taking longer than usual, are requiring greater workloads, or are incurring higher than average costs.
Increasing Employee Productivity
While many facets of the loan origination process have become automated, there is still a large human element involved. Before making new investments in technology, it makes sense to understand both the capacity and the productivity of your team members. Scoring provides a means for identifying when employees are underperforming relative to their past performance and their colleagues. Scores can also indicate when systems reach maximum capacity and when additional technical resources or personnel investments are required.
Optimize Processes and Achieve Results
Whether you have already implemented technology into your loan origination process or are assessing whether a new investment is beneficial, scoring provides a benchmark to quantify your investment. For example, scores can identify the maximum number of applications permitted related to the number of loans that can be processed. It makes little sense to improve technology in the loan application process if your processors and underwriters can’t keep up with the demand. Knowing when inputs (applications) exceed output capacity (underwriting) will help you see where and what technology you need to increase the overall throughput. A competent scoring strategy is a critical component to making insights action-oriented. Working with an experienced provider like Grind can help you leverage industry best practices and supercharge your results.
MODM from Grind Analytics
MODM, our comprehensive business intelligence solution, attaches directly to the mortgage operations platforms and collects and analyzes your data in real-time. Our proprietary Grind Score technology provides data analytics to help inform your decisions using algorithms to produce a singular score so you can quickly assess the health of mortgage operations. The inputs can align to meet your unique objectives, such as the timing of mortgage loan stages, work queue information, employee productivity, lead information, or dozens of other categories you can customize. The scores, along with summarized details regarding the most critical factors contributing to the score, are then displayed to end-users so they can take immediate action to improve efficiency and increase your bottom line.
The Grind Foundation:
Turn Data into Information. Turn Information into Insight. Turn Insight into Action.