Pull-through is the mortgage KPI most talked about and least improved. Here's how to actually move it.
Two numbers
App-to-fund and lead-to-fund tell different stories.
Top quartile: 25%+
Lead-to-fund benchmark for elite shops.
Denominator wins
Fixing top-of-funnel beats squeezing the back.
Move it in a quarter
Pre-qual delivers 5–10 point structural lift.
Pull-through rate is the ratio of funded loans to a chosen upstream count — most often applications taken, sometimes leads received. Always specify the numerator and denominator when quoting a pull-through figure; the same lender can honestly quote 75% and 18% simultaneously because they measure different things.
| Segment | App-to-fund | Lead-to-fund |
|---|---|---|
| Industry average | 70–75% | 15–20% |
| Top quartile | 80%+ | 25%+ |
| Bottom quartile | <60% | <10% |
| Pre-qual-enabled shops | 78–85% | 24–30% |
Almost every published "improve your pull-through" article lists 15+ levers. In practice, three of them explain most of the movement:
Doc-collection improvements move pull-through 1–3 points. Program routing moves it 3–5. Pre-qualification moves it 5–10 because it changes the denominator itself: fewer bad files become applications, so the pull-through math starts from a cleaner base. It's the only lever that changes the composition of what you're measuring.
One-line diagnostic
Pick your denominator
Use funded date, not app date
Segment by channel + product
Track LO-level
Industry averages hover around 70–75% for application-to-fund and 15–20% for lead-to-fund. Top-quartile shops hit 80%+ and 25%+ respectively.
Both are used. 'App-to-fund pull-through' is the standard back-office metric; 'lead-to-fund' is the marketing / acquisition metric. Always specify which one you're citing.
Rate spikes kill refi affordability and squeeze DTI on purchases. Files that penciled at application no longer pencil at lock, and pull-through drops. A pre-qual layer with live rate assumptions mitigates most of this.
Usually, yes. Scaling paid channels almost always dilutes lead quality unless you install a pre-qualification layer that keeps the top-of-funnel bar constant regardless of volume.
For unit economics, yes. Two shops with the same fund count but different pull-through have wildly different LO productivity and cost per funded loan.
Rarely, but a suspiciously high pull-through (95%+) usually means the lender is only taking applications from files already essentially closed elsewhere — a sign of underutilized capacity.
Structural levers (pre-qualification, program routing) can move pull-through 5–10 percentage points within a quarter. Ops levers (doc collection tightening) move it 1–3 points over the same period.
The MBA reports origination volume and mix but not pull-through directly; pull-through has to be reconstructed from application and closing counts, which is why so few shops track it consistently.
Schedule a 30-minute demo and we'll map the pre-qualification layer to your current LOS + CRM stack.