Pipeline Visibility Isn't Enough: Turning IMB Metrics Into Action

Introduction

Your loan origination system knows everything about your pipeline. It records every milestone, every condition, every status change. What it won't do is tell you which loans are about to stall, which borrower hasn't uploaded a pay stub in nine days, or which file your processor forgot about.

Most IMB ops leaders track pull-through rate, cycle time, condition aging, fallout, and document fulfillment without a clear way to act on them. The numbers describe what already happened. They rarely trigger anyone to do anything. That gap between data and action is where loans slip, locks expire, and borrowers walk.

The 5 Metrics That Actually Drive IMB Pipeline Performance

Five numbers tell you whether your pipeline is healthy or quietly leaking money. Each one measures a different stage of the journey from application to funding, and each fails in a predictable way when nobody watches it.

Metric	What it measures	Why it matters
Pull-through rate	Funded loans as a share of applications	Predicts revenue and exposes leaky stages
Cycle time	Days from application to close	Drives borrower satisfaction and lock costs
Condition aging	Days an open condition sits unresolved	Early warning for stalled loans
Loan fallout rate	Loans that never fund	Quantifies wasted origination effort
Document fulfillment rate	Share of requested docs returned on time	Leading indicator for the other four

Pull-Through Rate

Pull-through rate divides funded loans by total applications over a period, and it tells you how much of your origination effort survives to closing. Most IMBs run somewhere between 70 and 80 percent, though purchase-heavy shops trend higher than refi-heavy ones.

Common causes of drag:

• Borrowers shopping rate after lock
• Appraisals coming in low
• Conditions that age out past rate-lock expiration

Cycle Time

Cycle time counts the days from application to close. Conventional purchase loans typically run 30 to 45 days, and every day above your average burns lock budget and borrower goodwill.

Watch for days lost in:

• Initial disclosure and intent-to-proceed delays
• Underwriting back-and-forth on incomplete files
• Conditions waiting on borrower documents

Condition Aging

Condition aging measures how long an underwriting condition sits open before someone clears it. A loan with three conditions aging past five days is stalled, even if the LOS still shows it green. Most LOs track conditions in their heads or on a sticky note, which is why these slip.

Watch for:

• Conditions assigned to nobody specific
• Borrower-dependent items with no follow-up cadence
• Stacked conditions waiting on a single missing document

Loan Fallout Rate

Fallout counts loans that enter your pipeline but never fund. Split it into two types, because the fix for each is different.

Type	Cause	Owner
Voluntary	Borrower walks, shops elsewhere, withdraws	Sales and follow-up
Involuntary	Denial, expired lock, failed appraisal	Operations and underwriting

Voluntary fallout usually traces back to slow communication. Involuntary fallout traces back to conditions and timing you could have caught earlier.

Document Fulfillment Rate

Document fulfillment rate measures the share of requested documents borrowers return within your target window. It sits upstream of condition aging and cycle time, so a low rate predicts trouble in both. When borrowers return docs slowly, conditions stack and days pile on.

Common causes:

• Vague or batched document requests
• No automated reminders between manual follow-ups
• Borrowers unsure which document satisfies which condition

Why Your LOS Surfaces Data But Not Insight

Your loan origination system is a system of record. It captures every milestone, every condition, and every document as the loan moves through origination. What it doesn't do is tell you which loan is about to slip or who should be working it right now.

That distinction is why ops leaders feel buried in data and still get surprised by fallout. The numbers exist. The signal that turns those numbers into action does not.

The Gap Between Reporting and Action

Open your LOS dashboard and you see a pipeline snapshot. You can sort by stage, age, or lock expiration and read exactly where every loan sits today.

The dashboard stops there. It won't escalate a loan stuck in conditions for nine days, and it won't tell a processor to pick up the phone.

• Reports show status, not the next action or its owner
• Stalled loans look identical to healthy ones until a deadline hits
• Nobody gets pinged when condition aging crosses a threshold

Manual Workarounds That Don't Scale

Most IMBs close the gap with human effort. A pipeline manager exports data into a spreadsheet, color-codes the at-risk loans, and runs a daily call to assign follow-up by hand.

This works until volume climbs. Every new loan adds rows to chase, calls to schedule, and borrowers to nudge for the same missing pay stub.

• Spreadsheets go stale the moment the LOS updates
• Pipeline calls eat hours and surface problems too late
• Condition chasing depends on whichever LO remembers to follow up

What Visibility Without Action Costs You

Seeing a stalled loan does nothing if you spot it the day before the lock expires. The cost of late action shows up in your numbers.

• Missed locks. A condition that ages quietly pushes closing past the lock date, and the loan blows the rate it was priced on.
• Re-lock fees. Extending or re-locking eats margin you already thinned to win the borrower.
• Voluntary fallout. A borrower who waits two weeks for a status update starts answering calls from the lender who didn't make them wait.
• Involuntary fallout. Documents that never get chased turn into denied or withdrawn files that counted as production right up until they didn't.

Each one traces back to acting after the signal mattered instead of when it appeared.

How Penny Connects Pipeline Visibility to Action

Penny sits on top of your LOS and reads the same pipeline data your reports already contain. The difference is what it does with that data. Instead of waiting for a pipeline call to surface a stalled loan, Penny watches milestone history, condition aging, and document gaps in real time and acts on them.

Capability	What Penny does
Stalled loan detection	Flags at-risk loans from milestone and condition aging signals
Condition chasing	Contacts borrowers by SMS, voice, and email to collect documents
Task routing	Assigns open items to the right person by loan stage and condition type

Flagging Stalled Loans in Real Time

Penny tracks how long each loan has sat at its current milestone and how long conditions have stayed open. When a file crosses your defined threshold, Penny surfaces it before the lock expires or the borrower starts shopping. You see the at-risk loan while you can still do something about it.

Chasing Conditions Without Manual Intervention

Penny reaches borrowers directly to collect outstanding documents without an LO making the call. It sends the request by SMS, follows up by email, and places a voice call when a condition stays open too long. The document lands in the file and the condition clears without manual chasing.

Routing Tasks to the Right Stakeholders

Penny reads the loan stage and condition type, then assigns each open item to the person who can clear it. A title condition routes to the closer. An income document routes to the processor. Nobody picks work off a shared list, so items stop falling between roles.

Getting Started: What to Measure First

Don't try to instrument every metric at once. Start with the two that move money fastest and give you the clearest signal on where loans stall.

1. Pull-through rate. This tells you whether your pipeline is converting or leaking. Track it by branch and by loan officer, and the soft spots show up immediately.
2. Condition aging. Open conditions are where most cycle time disappears. Watch how long conditions sit unaddressed, and you find the stalled loans before they fall out.

Penny watches both for you and acts on what it sees. Stop reading dashboards and start clearing your pipeline.