This part of mortgage operations doesn’t get much attention, so it makes sense to focus on the handoff process. Most public resources only talk about the ideal workflow: intake, verify, underwrite, and close. They usually leave out what happens when a file stalls but can still be saved. Here’s what actually happens in practice, based on how these organizations operate.
The main problem is that ownership gets divided, and this split is behind most failures.
When a file runs into a recoverable problem, such as missing income documents, a debt-to-income ratio that requires a compensating factor, credit events that need time, funds that must be sourced, or employment history that needs rebuilding, there is almost never a single person in charge. The file gets stuck between stages. Processing teams usually handle the process from intake to closing, but a stalled file is where this normal flow breaks down.
In practice, here’s where the file usually ends up:
- The file stays with the loan officer in an informal ‘pending’ bucket. This is most common in small and mid-sized shops. The loan officer keeps it in their pipeline, adds a note in the loan origination system, and works on it when they have time between active deals. Without deadlines or requirements, these files often get forgotten.
- Sometimes the file sits in the loan origination system with tasks or notes, but no one is working on it. Technically, it’s still ‘in process,’ but if you look at the five back office stages—document intake, verification, conditions management, underwriting submission, and clear to close—it’s stuck in a stage it can’t move past. The tasks are open, but no one is making progress.
- In some cases, the file is moved to a ‘second look’ or credit remediation track. Larger lenders and some correspondents use this setup. Usually, a specialist processor or junior underwriter manages a group of ‘recoverable’ files. This is the closest most organizations get to real remediation instead of just nurturing leads.
- Sometimes the file is moved to CRM nurture and removed from active operations. This is common in high-volume shops. The file leaves the loan origination system; the borrower is added to a drip marketing campaign, and reactivation occurs only if the borrower reaches out again. Remediation and lead nurture are both real, but most organizations do not clearly separate the two.
When organizations do make a distinction, it’s usually in one of three ways: a separate pipeline stage, such as ‘credit repair,’ ‘second look,’ or ‘pending conditions’; a dedicated queue or team; or a tagging system in the CRM that a manager reviews later. Most of the time, though, there’s no distinction, and the file is treated like a dead lead. The borrower keeps getting marketing emails, but their real underwriting issue is ignored. The ‘remediation’ and ‘cold lead’ files follow the same workflow, using the same templates and timing. That’s a structural mistake. A borrower who is 40 DTI points away from qualifying is in a very different situation than someone who just wasn’t ready to buy. They need a different approach.
Now, let’s talk about tracking the specific blocker, ongoing work, received documents, and unresolved items:
This is where the ‘PDF trap’ and unstructured data become a real problem. In older, document-driven loan origination systems, the file is just a stack of PDFs and some free-text notes. The blocker might only be in someone’s memory, in an email thread, or in a note like ‘waiting on 2023 K-1s.’ There’s no structured field to capture the real blocker, so tracking depends on whoever remembers.
Organizations that do this well use structured data and track each condition with a status, owner, and due date. The system flags stale conditions instead of relying on a processor to mark a package as complete without proper tracking. Deciding when a file goes back to the underwriter is still usually a human decision, not an automated one.
Your main operational question—can a successor reconstruct the file?—is important. The answer is usually: only partly, not very reliably, and only if the organization is very disciplined.
If the original loan officer or processor is gone, a replacement can usually figure out what the blocker was (it will be in the notes or open conditions), but almost never what work has been done since, what’s been received but not yet logged, or whether it’s actually ready for another look. This is because remediation work occurs between stages and systems, where friction is highest and documentation is weakest. Files often move backward when underwriting finds a gap, and these backward moves are the least tracked part of the process.
A successor who gets the file might find the blocker and the documents, but there’s a good chance they’ll repeat work that’s already been done or ask the borrower for documents they’ve already provided. This quickly destroys borrower trust.
Here’s what it looks like when an organization has solved this problem:
- There is a separate pipeline stage for files that are recoverable but stalled, separate from both active underwriting and cold nurture.
- Condition records are structured with owner, status, and age, rather than just free-text notes.
- There is a clear reconsideration trigger, such as a specific event, a score threshold, or a seasoned trade line, that automatically sends the file back to the processor or underwriter rather than waiting for someone to notice.
- There is a clear rule that remediation files are handled on a different schedule than lead nurture files.
- The system is set up so that if an employee leaves, a successor can rebuild their files in minutes instead of days. Organizations have the first and fourth features. Some have the second, and only a few have the third or fifth.
In today’s market, the ‘stalled but recoverable’ file is the least managed, least tracked, and least standardized part of the mortgage process. That’s why it’s so hard to answer your question with public data, and why the answer varies so much between lenders. If you’re working on solutions in this area, the gap is real and still unaddressed.
Let me know if you want more detail on any part. For example, I can explain how second-look or remediation teams are staffed and paid, or how structured-data loan origination systems handle condition aging.