Is the mortgage industry broken—or structurally misaligned?

There is an ongoing debate in mortgage lending about whether the industry is fundamentally broken or simply experiencing another cycle of technological evolution. Over the past decade, lenders have invested heavily in automation, OCR, verification services, title integrations and income validation tools. Artificial intelligence is now emerging as the next expected leap in efficiency.

Yet despite this sustained investment, core operating metrics have not moved in the direction many expected. The cost to originate a mortgage has increased from approximately $8,300 per loan to more than $11,300 in recent years. At the same time, loans continue to require multiple handoffs and repeated “touches” before reaching closing.

This raises a fundamental question: Has technology meaningfully improved mortgage manufacturing efficiency—or has it simply added layers of complexity to an unchanged operational model?

Efficiency gains or structural constraints?

Speed is often cited as a defining competitive advantage in mortgage lending. Faster cycle times, quicker underwriting decisions and accelerated closings are frequently positioned as evidence of operational improvement.

However, speed alone does not necessarily reflect efficiency. In many cases, faster processing simply shifts complexity downstream, where loans require additional effort to resolve data gaps, clear conditions and correct inconsistencies introduced earlier in the process.

A significant portion of operational capacity in many lending organizations is still dedicated to exception handling rather than true straight-through processing. When headcount is primarily used to “repair” loans instead of originating clean ones, the cost structure inevitably expands.

As margins continue to compress, this model becomes increasingly difficult to sustain. Efficiency gains achieved through front-end technology investment are often offset by back-end rework, limiting the industry’s ability to realize meaningful cost reduction.

The submit-to-close gap as a structural signal

One of the clearest indicators of this imbalance is the submit-to-close ratio. Historically, the industry operated near a 65% conversion benchmark. More recent data, including Home Mortgage Disclosure Act (HMDA) reporting, indicates that national averages have declined to approximately 53%.

While often attributed to market cycles or temporary conditions, this decline may also reflect a deeper structural issue in how loans are originated and qualified.

If nearly half of submitted loans do not reach closing, the inefficiency cannot be isolated to underwriting or secondary review. It suggests that a significant portion of breakdowns are occurring before the loan is fully validated for submission.

Despite this, many performance frameworks continue to prioritize application volume and pipeline growth over funded outcomes or loan quality. This creates a misalignment between activity-based metrics and actual operational performance.

Over time, this structure reinforces a “submit first, resolve later” mentality, increasing downstream workload, elevating costs and placing sustained pressure on operations teams.

The front-end problem in loan manufacturing

The most persistent inefficiencies in mortgage lending often originate at the earliest stage of the process: the borrower–loan officer interaction.

This initial engagement establishes the foundation for the entire loan lifecycle. When information is incomplete, inconsistently gathered or not fully validated, the impact compounds through underwriting, processing, and closing.

In many organizations, the emphasis on speed to submission reduces the depth of this initial analysis. Loan officers are often encouraged to secure commitments quickly and move files into the system, with the expectation that downstream teams will resolve any issues.

This approach does not eliminate complexity—it redistributes it.

Yet the loan officer remains the closest point of contact to both borrower intent and property-specific realities. As such, they are the most critical control point in determining whether a loan is structurally sound at inception.

Improving outcomes at this stage requires more than process discipline. It requires a consistent framework for borrower engagement, structured data capture and disciplined qualification standards applied before submission—not after.

Reframing mortgage lending around loan quality

Addressing these challenges does not require a complete reinvention of mortgage lending, but it does require a reorientation of where quality is established in the process.

The most meaningful operational improvements are unlikely to come from additional layers of back-office automation. Instead, they will come from improving the consistency, completeness and accuracy of loans at the point of origination.

When the front end of the process is strengthened, the downstream impact is immediate and measurable: fewer underwriting conditions, reduced file touches, lower operational cost per loan and improved pull-through performance. This shift also requires a change in how success is measured. Volume-driven metrics alone are insufficient if they are not paired with meaningful indicators of loan quality and funded performance.

Technology will continue to play an important role in this evolution, particularly in standardizing intake, supporting compliance and guiding decision-making at the point of sale. However, its effectiveness will depend less on sophistication and more on how directly it improves borrower engagement and data integrity at the earliest stage of the process.

Conclusion: Fix the source, not the symptoms

The mortgage industry does not necessarily need to be rebuilt—but it does need to be rebalanced.

For decades, operational models have focused on moving loans through the system rather than ensuring they are structurally complete when they enter it. As a result, inefficiencies have been managed rather than eliminated.

In a market defined by rising origination costs, compressed margins and declining pull-through rates, the most durable advantage will not come from faster processing. It will come from originating cleaner loans from the start.

The opportunity ahead is not to add more complexity to an already layered system—but to rethink how the system is designed, measured and executed.

Fix the front end, and the rest of the process becomes materially more efficient.

Randy Senzig is the CEO of The LANIS Group, LLC
This column does not necessarily reflect the opinion of HousingWire’s editorial department and its owners. To contact the editor responsible for this piece: zeb@hwmedia.com.