The Hidden Cost Of Mortgage Tech: How Workflow Fragmentation Is Driving The “interoperability Tax”

The cost problem lenders can’t automate away

Mortgage lenders have spent years investing in automation, yet one metric continues to resist improvement: cost per loan.

New tools promise faster processing, better decisioning and improved borrower experiences. For many lenders, however, those gains have been offset by growing operational complexity that is harder to see but just as impactful. Workflows have become more fragmented, creating a hidden layer of cost that slows production, increases rework and limits the impact of automation.

According to Steve Butler, CEO of TRUE, that cost has a name: the interoperability tax.

Why mortgage workflows move backward instead of forward

In most industries, production follows a predictable path. Inputs move forward, decisions are made and outputs are delivered; however, mortgage lending doesn’t behave that way. Loans frequently move backward in the process, especially when underwriting identifies missing or inconsistent information. Instead of progressing toward a decision, files are sent back to processors or borrowers to correct or supplement data.

That dynamic fundamentally changes how efficiency works in mortgage operations. “Decisioning is what mortgage is all about,” Butler said. “And it only happens when all the data is in a good place.” When that data isn’t complete or consistent, the entire workflow stalls. What should be forward momentum becomes a cycle of correction.

The interoperability tax, explained through everyday workflows

The interoperability tax becomes most visible in routine tasks such as document handling.

Consider a bank statement. Once submitted, it does not move through a single system. Instead, it passes through multiple tools, each responsible for a specific function. One system captures and classifies the document. Another evaluates it for fraud. A third extracts and analyzes the data for underwriting purposes.

Each step is necessary, but they rarely function as a unified workflow. “You’ve got at least three tools,” Butler said. “They all need data… and they’re probably going to be specialized UIs.”

Because these systems are not seamlessly connected, employees must fill in the gaps. Data is re-entered, verified or reconciled across platforms. Over time, this creates inefficiencies that are difficult to eliminate and even harder to scale.

The result is not just slower processing, but an operational structure that depends on specialized knowledge of each tool.

Why more automation hasn’t solved the problem

Many lenders assume manual work exists because automation is missing. In reality, much of that work happens between automated systems, not within them. Even with multiple tools in place, workflows still rely on human intervention to connect processes, validate inconsistencies and manage handoffs. This is why technology investments often fail to reduce cost per loan.

“[Lenders] tell me, ‘We spent all this money on technology, and we’re not lowering our cost per loan,’” Butler said. “The issue is . . . they just have this new department of specialists.”

Instead of eliminating work, fragmented automation redistributes it — often in ways that are harder to scale.

The real gap: Getting the data right first

At the center of the issue is data integrity. When lenders push loans downstream before validating core data, they increase the likelihood of rework. Inconsistent borrower information, outdated documents or incomplete records force teams to revisit earlier steps, breaking the flow of the process. “Garbage in, garbage out,” Butler said.

A more effective approach is to ensure that data is accurate, consistent and aligned before it reaches underwriting. This includes confirming that borrower information matches across documents and that all required conditions have been met.  “When you have a 360-degree view of the data . . . the chances of going backwards are a lot less,” he said. By addressing data quality early, lenders can reduce the number of times a loan must be revisited later.

How offshore review adds friction to the interoperability tax

Another layer of the interoperability tax comes from how data is reviewed. In many workflows, documents are routed through offshore teams for validation and correction before moving forward. While intended to ensure accuracy, this introduces delays and breaks the continuity of the process. “Lenders have to go through the experience of the data having to be reviewed and corrected offshore,” Butler said. “There’s a lag then . .. and that’s part of the interoperability tax.”

When validation happens outside the core workflow, data is no longer updated in real time. New or conflicting information may not be flagged immediately, increasing the likelihood of rework later in the process. “You’ve lost the idea of continuous and in real time if it has to go somewhere for review and come back later,” he said. Keeping validation within the workflow allows lenders to maintain momentum and reduce delays that drive up the cost per loan.

How AI is changing the shape of mortgage work

Artificial intelligence is beginning to reshape how mortgage workflows operate by handling many of the tasks that traditionally create bottlenecks. Rather than replacing human roles, AI functions as a background layer that continuously processes documents, validates data and applies rules.

This allows processors and underwriters to focus on decision-making instead of data cleanup.  “They’re not getting replaced, but they’re getting hugely productive,” he said. With fewer inconsistencies and cleaner data entering each stage, loans require fewer touches. That reduction directly impacts both cycle times and cost per loan.

The workflow becomes more predictable, and teams spend less time reacting to issues that could have been prevented earlier.

From fragmented tools to unified workflows

Lenders that are seeing meaningful improvements are moving away from disconnected tools and toward unified workflow environments.

In these environments, data flows across systems without interruption, and users operate within a consistent interface. This reduces the need for manual handoffs and eliminates many inefficiencies caused by fragmented processes. “The productivity of the processor goes up three to four times,” Butler said.

That increase in productivity translates directly into cost savings. Many lenders see reductions of $150 to $250 per loan in early-stage processing alone, with additional improvements across the rest of the lifecycle.

As workflows become more connected, they begin to resemble a true production line, where each step builds on the last without unnecessary interruption.

Rethinking scale in a more efficient system

As workflows become more efficient, the role of loan originators and ops teams evolves. Rather than spending time on document management and error correction, teams can focus on higher-value activities such as borrower engagement and pipeline growth.

“I think you see top-line growth before you see staff reductions,” Butler said.

In this model, technology doesn’t replace people — it amplifies their capacity. Loan officers can handle more volume. Processors can move files faster. And organizations can scale without adding proportional overhead.

Efficiency starts with interoperability

The mortgage industry’s cost challenges are not rooted in a lack of technology. They stem from how that technology is connected. Fragmented systems, inconsistent data and disjointed workflows have created an environment in which automation alone cannot deliver its full value.

The next phase of efficiency will come from interoperability. Lenders that align their data, workflows and user experiences into a cohesive system will be better positioned to reduce costs and improve performance. “I’d look for a single platform… with a common data layer and a common UI,” Butler said.

In a market defined by tight margins and rising complexity, solving for workflow is no longer optional. It is a requirement for scale.

Click Here