I recently wrote that the mortgage industry has been suffering from shiny object syndrome. Our quest to build a better borrower experience, I argued, has been hampered by our tendency to fixate on one small piece of the puzzle — the borrower's application experience — instead of considering the back-end lender experience and all the other pieces that make up a truly digital mortgage. For best results, we should be tackling mortgage inefficiency from soup to nuts.

In my last article, I described this problem using a metaphor that imagines the mortgage lender as a fine-dining restaurant. In this scenario, rolling out a shiny new borrower app is akin to sprucing up the front of the house with new tablecloths and a fresh-faced hostess. Such enhancements can make a big difference, and they may attract new customers initially, but it's what happens behind the scenes — whether in the kitchen or the mortgage back-office — that determines long-term business success.

So, how do you fix the kitchen?

Professional chefs the world over are taught that organization is the key to working efficiently in the kitchen. This concept is embodied by the French phrase mise en place, which literally means to "put in place" everything that is needed for the day's work.

Professional kitchen
Adobe Stock

Similarly, to succeed in an era of increasingly narrow margins and broad competition, lenders must take a methodical approach to loan fulfillment. Improving the mortgage experience for loan officers, underwriters and processors doesn't just require slick mobile apps and friendly graphics. It's time for lenders to add intelligence to their back-office technology.

The ideal enterprise interface is straightforward in its presentation of information and takes a "less is more" approach, where each party to the loan sees exactly the information required for the next step in the loan process — no more, no less.

The key to this kind of dynamic, optimized workflow is technology that relies on data, not documents. In this day and age, there is simply no reason loan officers and back office staff should be wasting their time scrolling through scanned documents, toggling between screens to find the right calculation or doing the infamous "stare and compare" across multiple documents. An intelligent system can do this automatically.

Technology also offers the advantage of greater objectivity, which translates to reduced risk. There's an old joke that if you give five underwriters the same loan, they'll return six different underwriting decisions. By leveraging intelligent technology to drive the day-to-day origination workflow, lenders can manufacture loans more consistently and bring in human expertise where needed to handle exceptions.

By tailoring the information shown on-screen to just the data points required for exceptions or other loan actions that require human intervention, modern enterprise technologies enable fulfillment teams to maintain a laser focus on the task at hand.

And enough with the one-size-fits-all solutions. It's a well-known saying in the culinary world that the same well-written recipe prepared by 10 different chefs will produce 10 different dishes. It's also a well-known adage in mortgage lending that no loan is simple and no two loans are the same — yet many legacy systems walk every loan through the exact same workflow.

Loan officers know that even when everything is perfect on paper — take, for instance, a borrower with a 750 FICO, a loan-to-value of only 70% and a sizable down payment — some aspect of the loan will be unique. Consequently, in a system with a static workflow, nearly every loan becomes an exception case. It's a major reason the average mortgage loan takes well north of 40 days to close and costs upwards of $7,770 in total production expenses.

Experienced chefs use recipes as a guide and adjust as needed based on customer requests, available ingredients and environmental factors in the kitchen. Origination technology should leverage rules- and event-based logic, machine learning and predictive analytics to adjust the workflow on the fly in accordance with the distinct needs of each borrower, loan program and loan product. With basic intelligence and a data-driven workflow, lenders can significantly reduce production costs for the vast majority of loans.

Reducing the number of loans that are treated as exceptions doesn't just save lenders time and money — it also limits the number of human touches per loan, which in turn reduces mortgage risk. Try to get that ROI from just pretty graphics.

Here's one more way mortgage lending is like fine dining: in both industries, margins are slim for all but the best-run enterprises. Yet when a kitchen operates like a well-oiled machine, its staff can crank out exceptional products at astounding speeds — and turn a tidy profit doing so. That's exactly what lenders can achieve by using intelligence to automate the back office.

I want to be explicit about automation. It is not about eliminating personnel; on the contrary, it's about supercharging the productivity of mortgage companies by freeing their teams up to complete more loans per capita.

As rates increase, margins continue to compress and the market continues its shift from refi to purchase, lenders are finding they must offer an ever more complex array of products that goes far beyond the conventional 30-year fixed. If left unaddressed, the back-office inefficiencies that have crippled loan operations throughout the last market cycle could prove fatal. It's up to lenders whether they will cook up a recipe for disaster, or one for success.

Kyle Kamrooz

Kyle Kamrooz

Kyle Kamrooz is the COO and co-founder of Cloudvirga in Irvine, Calif. He also serves on Fannie Mae's Technology Advisory Board.