VOIE… The Gap Between the File and the Truth

My previous piece argued that credit underwriting in mortgage is restructuring around cashflow-native frameworks that treat repayment capacity as a distribution rather than a score. The GSEs have moved, and the capital markets are following. The logic is compelling and the direction is clear.

A model is only as good as its inputs: input the wrong income or verify a job that is gone, and the shape of risk changes. The system that promised certainty now hides uncertainty. We cannot call underwriting modern until we answer one question: does the borrower’s economic reality match the file?

The Verification Problem Is Not What It Appears

We treat income and employment verification (VOIE) as a paperwork issue.  This frame feels comfortable, but is incorrect. Documentation looks backward by nature. It’s about last year's W-2, last month's paystub, or a current employer call. We’re not lending to what was, but to what it will be.

The present standards fit slow fulfillment processes, where time gaps didn't matter, when fraud was caught with paperwork and horse sense. Those days are over. 

Despite enormous investment in mortgage technology over the past decade, fulfillment productivity has barely moved. An employee closes about the same number of loans today as in 2018. This shows that solutions were built for the easy path. But exception handling is the real work in this industry. VOIE is still managed as documentation, confirmed too late. The gap between what the borrower claims and what the lender records gets too wide. This is a timing and workflow problem disguised as a documentation problem.

The Incumbent and the Market It Created

To understand how the industry arrived here, it helps to revisit TheWorkNumber (TWN). In TWN, Equifax has built a database with over 200 million active records covering 75%+ of U.S. nonfarm payroll workers. Genuinely valuable coverage. The problem is what Equifax did with the position.

To see how we got here, look at TheWorkNumber (TWN). Equifax built a database in TWN. It holds over 200 million active records. This covers more than 75% of U.S. nonfarm payroll workers. That coverage is truly valuable. The problem is how Equifax used this position. 

The price per transaction rose from under $20 a decade ago to over $65 now. Detailed pulls can cost $200. Equifax Workforce Solutions, where TWN lives, consistently hits 50%+ adjusted EBITDA margins. This margin shows their pricing power. The industry has no real alternative. 

This is a company that built a moat. It extracts rent from an industry that lacks options. TheWorkNumber confirms data exists. It does not promise you will get an answer.

The Single-Method Challengers

A wave of fintech challengers entered with a cleaner, technology-centered thesis: direct payroll connection beats a static database. Consumer-permissioned, real-time, lower cost.  Argyle, Truv, and Pinwheel used this to cut out the middleman and deliver fresher data at a lower price. Their critique of the old way was right, but their architecture is still incomplete. 

Banking consolidation made connection infrastructure, like Plaid’s, scalable. Employment moved the opposite way. Gig work, multiple streams, contracts, and self-employment layer onto W-2 jobs, resulting in a landscape diverging counterparties. No single connection can reliably sort out this complexity.

The industry chose to optimize for access. It tracks hit rates, coverage, and cost per pull. These statistics are useful, but they are disconnected from the one question that matters in underwriting: did a verified income picture arrive in time to sway the credit decision? A successful data retrieval is not the same thing as a resolved question. Late-arriving verification results only document the uncertainty that already exists.

Resolution Infrastructure

In Part 4, I argued that Plaid’s value was in workflow, not just data. It built a connection infrastructure that made bank data arrive exactly when needed. Downstream systems could use the data immediately. ‘

VOIE needs a similar shift. Its domain is harder and more fractured. Payroll processors, employers, tax authorities, and bank statements use different timelines and models. No single path consistently works. When the first method fails, the system routes. If that fails, it routes again, intelligently. VOIE needs a resolution infrastructure, a system must take responsibility for the outcome, not just for trying.

So far, one player, Truework, has come closest to building its architecture around this goal. The difference is fundamental as it shifts the value equation. Value does not come from owning a database or perfecting a single connection. It comes from lowering the chance that verification secretly derails the loan. 

The operational effects are clear, though often ignored. Intelligent routing speeds up turnaround times. It raises completion rates. It is not just faster; it puts verification earlier in the process. Clarity on income arrives before rate locks start to cost money. It arrives before deal momentum makes surprises expensive. This changes the decision instead of just confirming it. Every unresolved check costs money. Locks are extended. Processor time is wasted. Referral relationships get strained. The total effect is not just a higher cost per loan. It is a steeper fallout curve. 

Implications to the VOIE Landscape

The competitive edge will not be about having modern verification tools, but about using them. It separates those who move verification into the decision loop from those who keep it in the back office. 

When the GSEs introduced appraisal waivers, many lenders failed to extract value because their workflows had a slot for an appraisal and no slot for a waiver. Efficiency arrived, but the process was not built to receive it. VOIE is approaching a similar inflection. Early movers will not merely save cost; they will operate on a different risk curve.

The Proxy Problem, Resolved

The core underwriting question has always been the borrower’s ability to perform, but this remains the least directly observable fact. The result is a reliance on proxies. FICO stands for creditworthiness. Appraisals stand for value. Paystubs stand for income continuity. Each proxy has worked, until the environment changes.

A resolution infrastructure does not eliminate uncertainty. What it does is reduce proxy error and narrow the distance between reality and decision. In a business defined by thin margins and tight timelines, that distance shapes pricing, fallout, and retained risk.

This is why Truework matters. It is more than just an efficient operation. Its intelligent routing answers the core question of mortgage underwriting: Can we truly trust what we know about this borrower? 

Certainty Requires a Decision

Mortgage operators do not experience valuation, credit modeling, and verification as separate debates. They experience a fulfillment operation built around information that arrives after it would have mattered, and a team whose institutional muscle memory runs exactly that process, efficiently, every day.

The true obstacle is not with technology, but with the organization. As AI and automation accelerate the sequence, margins compress and cycle volatility increases, the cost of delayed clarity becomes more visible, and more attributable. The gap between what's possible and what's practiced can be seen in fallout rates, lock extension costs, and cost per funded loan.

The starting point is a deep review of the workflow. Map where each signal actually arrives in your current process against where it would need to arrive to change a decision rather than confirm one. That map will tell you more about your competitive exposure than any vendor comparison. The gaps it reveals are not technical failures. They are organizational choices that have hardened into defaults. 

Lenders who are closing these gaps are not waiting for a total overhaul. They pull one signal forward at a time. They rebuild the workflow to accept it. They let the earlier clarity dictate the next step. This is how muscle memory changes. They show that the new way works better than the old way. 

The infrastructure is ready. The signals are there. A decision remains. Where does the process confront uncertainty? Is that moment early enough to matter? In mortgage, the truth usually exists. The real question is how far it sits from the moment of decision.