Faster Underwriting Is a Pipeline Decision, Not a Headcount Decision

By Steve Iskander, Founder and CEO, Intrepid

In this article

• Two clocks, one gap
• Why hiring scales the wrong thing
• The measurement that changes the conversation
• Where to start

Every lender I talk to who wants better loan underwriting speed starts in the same place. We need to hire.

I understand the instinct. I ran sales teams for fifteen years and my first answer to more volume was always more people. It is the answer you can execute this quarter.

It is also the one that scales your problem along with your capacity.

Here is what I mean. If a deal takes twenty-two days and eighteen of them are assembly, chasing a missing statement, re-keying a debt schedule, waiting on a document already sent to someone else, then hiring a second underwriter gets you two people doing eighteen days of assembly in parallel. You bought throughput. You did not buy speed. And you added coordination cost that shows up later.

The industry data points the same direction. Analyses of automated underwriting workflows report approval times dropping by up to roughly 80% when the assembly layer is systematized, and the framing I keep seeing from practitioners is that the underwriting bottleneck is a workflow problem, not a staffing one. Roughly nine in ten financial institutions now describe AI as critical to the lending lifecycle, and most are raising budgets accordingly. Results vary by institution.

But notice what those gains actually measure. Not better credit judgment. Less time between a document arriving and that document being usable.

That distinction tells you where to spend. Judgment is what you hired for. Assembly is what you should never have been paying for.

The lenders pulling ahead right now did not out-hire anyone. They shortened the distance between a borrower’s raw information and a credit officer’s screen, then let the same team review more deals without lowering the bar.

That is the thesis behind how we built Intrepid. Financial data ingested in real time, structured, ready to underwrite against, with the decision still sitting where it belongs.

Speed is not the opposite of discipline. Slowness just feels like diligence because it costs something.

For the credit teams here: if you timed your last ten deals from first document to decision, how much of that clock was judgment?

Two clocks, one gap

Most lenders can quote their turnaround time, but it usually starts when the file is already complete. Complete means someone already collected, chased, formatted, and verified. The borrower’s clock started earlier, at the first document they sent. The gap between those two clocks is where deals are lost to a faster competitor and where good borrowers quietly conclude you are hard to work with. Measuring from the borrower’s first document, rather than from a complete file, is uncomfortable precisely because it reveals how much of the delay lives in assembly.

Why hiring scales the wrong thing

When a team is capacity constrained, the instinct is to hire. But if a large share of each underwriter’s day goes to assembly rather than judgment, adding people multiplies the assembly work along with the throughput, and adds coordination cost on top. You end up paying underwriting salaries for document logistics. Capacity is often a real need, but it is usually the second problem. The first is how much of the existing team’s time is spent on work that should not require an underwriter at all.

The measurement that changes loan underwriting speed

A single metric reframes the whole discussion: what share of an underwriter’s time is spent on credit judgment versus assembly. Teams consistently guess high, then track it for two weeks and come back with a smaller, more honest number. Once that number is visible, the path forward stops being about headcount and starts being about the pipeline that feeds it. Structure the inputs, and the same team reviews more files at the same standard, with the calendar shrinking rather than the rigor.

Where to start on loan underwriting speed

The first move is not to buy anything. It is to measure the real cycle time, from the borrower’s first document to the decision, on the last ten files. That single number reframes the conversation, because it exposes how much of the delay is assembly rather than analysis. The second move is to look at where documents change hands and formats, since those seams are where time and errors accumulate. From there, the highest-return fix is almost always structuring inputs on arrival, so the file is machine-ready before an underwriter opens it. Hiring can come later, once the process is clean and you are still genuinely capacity constrained. Done in the other order, new hires simply absorb the same assembly work, and the throughput gain is smaller than expected. Speed is bought upstream. The credit judgment at the end is not the slow part, and treating it as though it were leads teams to optimize the wrong thing. None of this requires a wholesale rebuild. It starts with one honest measurement and a decision to fix the seams where files stall. Teams that do this consistently find that speed and standards stop being a trade-off.

Related reading

→ Why AI is not the bottleneck in lending

→ The 80% of credit data you are not underwriting

Frequently asked questions

What is the main bottleneck in underwriting speed?

In most operations the bottleneck is assembly, not analysis: collecting documents, chasing missing items, and re-keying data into a usable form. The credit judgment itself is often a small share of total elapsed time, which is why practitioners increasingly describe the underwriting bottleneck as a workflow problem rather than a staffing one.

Does automated underwriting reduce approval times?

Industry analyses report that automating the data-assembly layer of underwriting can cut approval times substantially, with some estimates of up to roughly 80%, though results vary by lender, portfolio, and process. The gains come from making information usable faster, not from replacing credit judgment.

Should lenders hire more underwriters to increase throughput?

Hiring adds capacity, but if most of an underwriter’s time goes to assembly rather than judgment, adding people scales the assembly work along with the throughput. The more durable fix is to structure inputs so the existing team can review more files without lowering standards. Capacity is a real need, but usually the second problem, not the first.

How do you measure time-to-decision correctly?

Measure from the borrower’s first submitted document, not from the point the file is already complete. Most lenders start their clock after collection, chasing, and formatting have happened, which hides the largest part of the delay. Timing from the first document reveals the real cycle time a borrower experiences.

Is faster underwriting less careful underwriting?

Not when the speed comes from the data layer. Removing assembly time does not touch the rigor of the credit decision; it gives the decision cleaner inputs and more time. Slowness often feels like diligence because it costs something, but a long cycle is usually a sign of manual assembly, not deeper analysis.

Sources: Gartner

Intrepid structures loan inputs on arrival, improving loan underwriting speed without touching your standards. See how at intrepidfinance.io.

Published by Intrepid. Democratizing Access to Capital. intrepidfinance.io