A closed-deal count answers one question — how much revenue landed — and tells a brokerage almost nothing about where the process itself is losing time or losing files. A file that fell out at conditions looks the same in a closed-count report as a file that was never viable to begin with, and a submission that took three weeks to reach commitment shows up nowhere at all.
Four numbers fix that blind spot without turning operations review into a data project: pull-through rate, first-pass approval rate, time-to-commitment, and pipeline aging. Each one answers a different question about where the process is working and where it is not, and together they take about fifteen minutes a week to pull and review.
Step 1. Track these four numbers, not a dozen
A dashboard with twenty metrics gets checked once and then ignored. These four cover conversion (pull-through), quality of submission (first-pass approval), speed (time-to-commitment), and where files are stuck right now (pipeline aging) — between them, almost every operational problem shows up in one of the four.
The point of building this is not the numbers themselves — it is that they surface a problem while it is still small. A first-pass approval rate that slips five points in a week is worth a conversation immediately; the same slip discovered three months later, buried in a year-end review, is three months of avoidable resubmissions that already happened.
Step 2. Compute the pull-through rate
Pull-through rate is the share of applications that actually reach funding, out of all applications started in a given period. It is the single clearest signal of where the pipeline is leaking, and the fuller breakdown of why files fall out lives in the broker pull-through rate article.
- 01Count applications started in the period (the denominator).
- 02Count how many of those funded (the numerator).
- 03Divide funded by started, and multiply by 100 for a percentage.
- 04Segment by deal type (purchase, refinance, self-employed) if the overall number looks off — the average often hides a single deal type dragging it down.
Step 3. Compute the first-pass approval rate
First-pass approval rate is the share of submissions approved by the lender without a resubmission or a major condition tied to a preventable file error. It measures submission quality specifically, which is why it moves independently of pull-through — a brokerage can have a healthy pull-through rate while still resubmitting a third of its files. The first-pass approval rate article covers how to separate a preventable resubmission from an unavoidable one.
The number that flags a training gap: a consistently low first-pass rate on files from one team member usually points to a specific, fixable submission habit, not bad luck.
Step 4. Compute time-to-commitment
Time-to-commitment is the number of business days from a complete submission to a signed commitment. Track it as a median, not an average — one unusually slow file skews an average badly, while the median shows what a typical file actually experiences.
| Metric | This week | 4-week trend |
|---|---|---|
| Pull-through rate | — | — |
| First-pass approval rate | — | — |
| Time-to-commitment (median, business days) | — | — |
| Files aged 14+ days without movement | — | — |
Step 5. Compute pipeline aging
Pipeline aging buckets every open file by days since its last meaningful movement, not days since it was opened. A file that moved yesterday is not stuck even if it opened six weeks ago; a file that has not moved in two weeks is stuck regardless of how new it is.
- →0 to 6 days since last movement — normal, no action needed.
- →7 to 13 days — worth a status check.
- →14+ days — flag for the weekly review; something is blocking it.
Review the dashboard weekly, in fifteen minutes
Pull the four numbers at the same time every week, note anything that moved meaningfully in either direction, and spend the review time on the aged-file list specifically — that is where the dashboard turns into action rather than a report nobody acts on. Pairing this with a fixed weekly pipeline review structure keeps the two working together instead of as separate habits.
Resist the urge to add a fifth or sixth metric the first time one of the four looks fine for a few weeks in a row. A quiet metric is doing its job; the discipline that makes this dashboard useful is checking all four every single week, not expanding the list until nobody has time to look at any of them closely.
A brokerage that would rather have this dashboard running without building it from scratch can plug directly into Treadstone's fulfillment services, where pipeline visibility is already part of how files are run.